Home

Disk/ATE™ User`s Manual Manual Rev. 1.1 By Jim

1. 2 prints on the terminal the file name including disk drive designator for the most recently referenced file N no argument New file Creates a new empty source file at the end of the source area and makes it current The target character becomes gt DISKATE User s Manual 139 the at the end of this file REF lt Interval Argument string up to 24 characters gt REF REF DEF _ DEF DEF F lt Interval Argument gt Reference interval The string in brackets is saved in a special location For all subsequent commands which take lt F gt as the initial reference interval the Interval Argument given by this string is evaluated and becomes the initial reference interval Commands taking lt S gt as the initial reference interval are unaffected Discontinue use of the REF feature no argument Print on the terminal the last string saved as the argu ment for a REF command lt command string up to 24 characters gt Default command The string in brackets is saved in a special location Subsequently whenever a command is expected and a carriage return only is typed with an empty command line the string saved will be executed as a default command Discontinue use of the DEF feature no argument Print on the terminal the last string saved as the argu ment for a DEF command Bi1e Establishes as the current file the file containing the lower ad
2. lt GIVES THE LOWER ADDRESS OF THE LAST INTERVAL COMPUTED DISKATE User s Manual reference interval to the last interval computed so the whole argument means the first line of the last interval computed But the last interval computed did n t contain an entire line so this command will give an error The operation allows you to specify an interval with a given interval but there can t be any overlap With lt and gt you can see intervals that do overlap the last interval computed The command i oe 3_ will show you three lines beginning with the first char acter of the last interval computed If that character happpens to be a carriage return you will appear to see only two lines It doesn t matter one way or the other whether the previous interval computed contained three lines or not l Care must be taken in using lt in complex arguments since its value will change as the argument is evaluat ed lt is ASSIGNED A VALUE AS SOON AS THE LOWER ADDRESS OF THE INTERVAL IS DETERMINED So if you wanted to see everything in the current file up to the first character of the last interval computed you could NOT use the command ES since as soon as DISKATE sees the it knows the lower address of the interval to be computed will be the beginning of the current file lt is then assigned this value so the argument lt specifes only the first char acter of the current file The value of lt can
3. Memory Files Files in the memory are located within an area called the source area denoted by lt S gt The lower address of lt S gt is given by the 16 bit value stored at BOSAP and the upper DISKATE User s Manual 127 address by the 16 bit value stored at EOSAP where BOSAP and EOSAP are symbols from ATETBL S deontes the same value as 1 the contents of EOSAP Each file is a sequence of bytes bounded by s One such file is the current file denoted by lt F gt The lower address of lt F gt is given by the 16 bit value stored at BOFP and the upper address by the 16 bit value stored at EOFP where BOFP and EOFP are symbols from ATETBL lt F gt does not include its bounding s lt S gt includes the giving the lower bound of the lowest file and the giving the upper bound of the highest file Symbol Table The symbol table is denoted by lt T gt The lower address of lt T gt is given by the 16 bit value stored at SYMTB and the upper address by the 16 bit value stored at TABA where SYMTB and TABA are symbols from ATETBL Each entry in the symbol table consists of the characters of the symbol with the last character having its high order bit set followed by the 16 bit value of the symbol A occurs at the end of the table The symbol table includes both variables and assembler labels When a value is assigned to a symbol the symbol table is searched until the terminal regardless of the value of TABA a
4. source area is the interval from the contents of BOSAP to the contents of EOSAP and the current file is the interval from the contents of BOFP to the contents of EOFP for instance You will need such information to supply an initial reference interval if you call VALUS Note that reading the contents of internal DISKATE loca tions in this way is a second method of passing informa tion from DISKATE to a machine language subroutine If you use this method be careful to avoid mistaking an address like BOSAP which is the address WHERE THE LOWER ADDRESS of the source area IS STORED for the contents of the 16 bit area beginning at this address A second method of invoking a machine language program from DISKATE is by using a special feature which we haven t discussed before the USER COMMAND TABLE To execute a machine language program by the method above you have to give a command in the form X addr or GO filename optionally followed by an argument By using the user command table you can invoke a command precisely as if it were a DISKATE command When DISKATE is powered up it contains a table which it consults whenever a command is given The format of this table is identical to the format of the symbol table which weve already seen However the symbols in this table are not labels or variables but are the names of commands created by the user The values for these symbols are the addresses of the machine language programs which ar
5. Address Origin The assembler program counter amp is set to the value of the operand SORG Storage Origin _ The assembler storage counter is set to the value of the operand ORG Origin The assembler program counter is set to the value of the operand The assembler storage counter is set to its current value the new value of the program counter the previous value of the program counter DB Define Byte The operand is defined as a one byte constant Multiple _ operands separated by commas may be given in which case a one byte constant is defined for each operand DW Define Word Same as above except that constants defined are 16 bit DISKATE Users Manual 141 DS ASC ASCX g EQU END INTE PROS IF words Define Storage A block of storage is defined whose length is the value of the operand The storage is not initialized Define ASCII string Generates a string constant consisting of the ASCII char acters comprising the operand field where x is any printing non alphabetic or digit character Same as ASC except that in the string generated each occurrence of x is replaced by a blank Equate The label of the EQU statement is assigned the value of the operand A label must be present and the same label cannot be given a value by more than one EQU statement Indicates the end of the source code In the absence of an END pseudo the end of fil
6. As mentioned in the introduction there is consi derable overlap between this Section and the previous one Most of the commands we will discuss have already been introduced In using DISKATE as a text editor the commands you give will usually apply to the current file within the source area or sometimes the whole source area In using DISKATE as a monitor the commands you give may apply anywhere in the computer s memory Some of these commands WILL FUNCTION DIFFERENTLY IF THE TARGET IS OUTSIDE THE SOURCE AREA Naturally all such differences will be pointed out In addition to using ASCII codes when using DISKATE as a monitor you will also want to deal with the codes directly probably in hexadecimal or octal When you bring up DISKATE there will be an initially estab lished CURRENT BASE The current base represents the base in which numbers from DISKATE are OUTPUT When you input numbers to DISKATE THE BASE MUST ALWAYS BE GIVEN EXPLICITLY If no base is specified DISKATE assumes that the base of any number typed in is DECIMAL There are several options for specifying numbers to DISKATE First of all a number may either be SPLIT or not A number in split form is specified by giving two numbers separated only by a colon The first number gives the value of the high order byte of a 16 bit value and the second number gives the value of the low order byte Por instance suppose you are reserving the first page i e the first 256
7. DISKATE User s Manual would any other file Now suppose you want to leave DISKATE for another program such as a DOS which will be co resident with DISKATE and then return to DISKATE What entry point should you use There are two If you give an I command for the disk containing DISKATE the address for the file ATE will give you the entry point which causes initialization You can jump to this address from a front panel DOS or a piece of software To re enter DISKATE without full initialization the entry point RENT as given in ATETBL can be used To find out its value mount the disk provided and type the command L ATETBL lt T gt KEEP RENT In DISKATE release 1 this entry point is 12 bytes beyond the beginning of ATE though this might change in the future Jumping to this entry point WILL initialize the stack but will not change the pointers to lt S gt lt F gt etc In Part II we ll discuss how a machine language program can call DISKATE as a Subroutine and have DISKATE return to it If you jump to RENT in such a situation the stack in the program from which you jump will be lost You might want to reload DISKATE from the disk in case something serious has gone wrong IO will write at memory location 3 a jump to IO Since IO begins with the loader for ATE you can reload DISKATE by jumping to memory location 3 so long as the area of memory occu pied by IO is intact To do this from within DISKATE you can giv
8. SHOULD ALLOCATE lt T gt BIG ENOUGH TO HOLD ATETBL AT LEAST DISKATE User s Manual II 3 DISKATE provides no less than three different methods for interfacing machine language programs can either be called from DISKATE or they can call DISKATE and execute it as a Subroutine We ll discuss these in order The easiest method of invoking a machine language program is with the X command which we ve already discussed Writing a program to be invoked in this way gives great flexibility since the program need not actually be resident in memory at the time it is invoked it can just as well be invoked by a GO command Of course when you issue a GO command to a machine language program you must be sure it will not overlay a part of the memory needed for something else In this section we ll describe how such a program can have access to DISKATE information There are 3 subroutines within DISKATE whose addresses are given in ATETBL that can be used in a linkage procedure whereby the same kind of interval arguments used by DISKATE commands can be passed to a machine language program The structure of the process is this the machine language program calls a subrou tine within DISKATE to determine if an argument is present If so another subroutine can be called to evaluate it The easiest way to link the source code of your machine language program to the addresses of these subroutines within DISKATE is to set the symbol table to
9. SIMPLEST COMMAND TO PUT AT END OF FILE M COMMAND MOVES INTERVAL GIVEN AS ARGUMENT AND INSERTS PRIOR TO TARGET CHAR DISKATE User s Manual that it is the third line instead of the fourth line The first step is to set the entry pointer to the right place In trying to determine where to set the entry pointer the key is always to determine which two char acters you want an insertion to go between Since we want the line with the address to become the third line we want it to go between the second line and what is currently the third line In other words we want to move what is now the fourth line to the location between the first character of what is now the third line and the preceding character This is achieved by 3 or alternately 3 Now to actually execute the move we use the command M4 g or M4 The file will now look like NAME JOHN DOE OCCUPATION DEER ADDRESS THE WOODS SEX MALE Notice that the whole job could have been done with a single command say SZL M4 However while you may find yourself using such commands often when you are accustomed to DISKATE it s always wise at first to make sure the entry pointer is in the right place before making any change A more cautious sequence would be 3 to see the position of the entry pointer followed by M4 _ The move command is one of the most powerful in the DISKATE text editor and there is no limit to the siz
10. 29 WITH ENTERED TEXT TO RESUME TEXT ENTRY AFTER MAKING CHANGES ENTRY POINTER MUST BE RESTORED IIIN e ooe o o VARIABLE NAMES LETTERS amp DIGITS STARTING WITH LETTER VARIABLES USED BY ASSEMBLER COMMAND VAR ARG SETS VARIABLE TO LOWER ADDRESS OF ARG INTERVAL DISKATE User s Manual 28 ENTRY and then when you want to restore the value of the entry pointer give the command ENTRY Note that the command does not use an and automatically assigns the value of the lower address of the argument to the entry pointer In fact ENTRY would give an error message since is not an ordinary variable The choice of ENTRY as the name of the variable is of course arbitrary and you may want to use shorter names Be careful though about using single letter names If we used the variable E instead of ENTRY this could cause real problems since E is one of the variables used by the assembler Whenever you are ready to begin entering text after having made some changes it s always a good idea to give a or com mand to make sure the entry pointer is where you want it to be If you make a change fail to restore the entry pointer and then resume text entry you may find your text looking much more botched up than it was before you made the change While a simple way to save the contents of the entry pointer is to set a variable and then vice versa aS we just saw
11. Actually this IS CHAR doesn t all happen character by character as you enter POINTED the E command but conceptually you can visualize it TO BY that way The E command thus inserts the typed charac ENTRY ters into the text without overwriting To simplify POINTER things let s think of the entry pointer as pointing to a character in the file This character will be called the TARGET CHARACTER When you enter text with the E command the text is INSERTED into the file BETWEEN THE CHARACTER PRIOR TO THE TARGET CHARACTER AND THE TARGET CHARACTER ios Let s look at an example Suppose the current file E consists entirely of the characters COMMAND Oo INSERTS My computer is well fed BETWEEN 3 7 TARGET with the target character being the w in well CHAR We ll leave aside for the moment the question of how it AND got that way Now suppose we give the DISKATE PREVIOUS command w CHAR E very The file will now consist of the characters _ My computer is verywell fed and the w in well will remain the target character though it will be in a different place in the compu ter s memory Because the w is the target character the very goes between the w and the preceeding character If I now enter the command the file will consist of the characters DISKATE User s Manual My computer is very well fed The text entered with an E command can be any number of lines long so lon
12. All subsequent P commands and Assembler commands will use the Processor Technology Format until an INTE is encountered P lt Interval Argument gt Print in Assembler Format The interval given as argument is printed on the terminal with line numbers in assembler format with label fields opcode fields operand fields and comment fields in col umns as determined by the TAB settings The choice of DISKATE User s Manual 2139 Intel or Processor Technology format is unaffected by any INTE or PROS psuedo ops within the source code printed but is affected by INTE or PROS commands as well as INTE or PROS pseudo ops executed by the assembler during Pass a TAB no argument list TAB settings The current TAB settings are printed on the terminal in the following order l relative column for the label field 2 R opcode field 3 es operand field 4 i comment field 5 source code offset The columns for items 1 4 above are relative to for P command listings and relative to item 5 for assembler listings TAB lt list of up to 5 arithmetic expressions Separated by blanks gt The TAB setting items listed above are replaced by the value of the corresponding expressions in the argument list If there are less than 5 expressions in the list items for which no expression is given are unchanged The TAB is set to the low order byte of the value of the expression V lt Interval Argument gt Ev
13. Another way is to set up the macro to keep entering the lines until it determines that no more can be added We can do this with the QF command There is only one way that QF can produce a matching failure if the interval is given in the form expressionl expression2 where the two expressions do not involve matching in the case that the value of expressionl is greater than the value of _expression2 Using this fact we can modify the commands above as follows N E LINE R8 EL N E FILL 6 ZERO THE INTERVAL GIVEN AS ARGUMENT DISKATE User s Manual 10008 B16 R 1 QF 82 2000H D LINE E D Ad J This macro will keep on entering the lines until the entry pointer comes within 82 bytes of the area we want to leave alone in which case it will quit Using techniques such as this there is no limit to the flexibility that DISKATE will allow Again as so often in this manual it must be emphasized very emphatically that after using DISKATE as a monitor if you want to resume editing text already established in the source area you must be absolutely sure the entry pointer has been restored in the proper place To conclude this section we introduce a command by which you can call machine language Subroutines from DISKATE The X command for eXecute will issue a CALL to the lower address of the argument given This can of course be either an explicit address or a more complex expression If the machine lang
14. BYTE IN CURRENT BASE E codes ENTERS DISKATE User s Manual codes as well as ASCII characters When entering text using the E command the ASCII characters comprising the text are surrounded by brackets To enter numeric codes using the E command the codes are surrounded by the character and separated by blanks In this case the base must NOT be given the base is automatically assumed to be the current base A string of codes sur rounded by the symbol can also be concatenated with other valid arguments of an E command to form a compo site argument The E command is one of those that works differently if the target character is outside of the source area When the E command was introduced it was explained that it will INSERT the entered text into the current file between the target character and the previous character However if an E command is given and THE TARGET CHARACTER IS OUTSIDE THE SOURCE AREA the bytes entered WILL OVERWRITE whatever is there and there will be no insertion Let s look at an example Suppose location 209DH contains a jump to an output routine for your terminal but you want to change the jump to a jump to location 307AH First we can examine the place we want to change using the command 200DH lt 2 This will show us three bytes beginning with 2 8DH We want to look at 3 bytes because the jump instruction is three bytes long We ll assume that the current base is 16 In resp
15. FILE SETS INITIAL REF INTERVAL FOR REST OF ARG TO PART OF ARG TO LEFT OF DISKATE User s Manual place for a change In the same way that we saw before issuing JOAN A or even JOAN will not tell us we are making the change in the right place because it will not tell us which occurrence of JOAN we are looking at Let s suppose that we have a file in which we want to make exactly this same change deleting the A from an occurrence of JOAN but there are many occurrences of JOAN in the file and we want to be very careful to de lete only the right one The command JOAN will print everything in the current file up to the first occurrence of JOAN Suppose this is not the right one and we want to print the rest of the file up to the next JOAN One way of doing it would be JOAN 2 JOAN There s nothing wrong with this but if this is not the right JOAN either we would have to be keeping track of the occurrence number Here is a command that you can keep on issuing successive times to see the intervals of text between successive JOAN s gt e JOAN This command uses a feature we haven t seen before So far the only time we have encountered the character gt is as the prompt character DISKATE recognizes this symbol as a special argument which does not involve matching gt denotes the upper address of the LAST IN TERVAL COMPUTED BY DISKATE Since we just got through with the c
16. TOO SHORT COMPACTING ON DRIVE facies designator DISK l WILL BE will be printed while the compacting is going on In COM this way DISKATE handles all disk space allocation for PACTED you and you needn t be concerned about allocating lengths for files Of course the situation can always arise that you run out of space on a diskette In this case DISKATE will print an error message to this effect on the terminal It can also happen that the directory becomes full if there are 64 files on the diskette An S command for a new file will also cause an error in this situation The error messages will tell you that either the disk or the directory is too full and will tell you which drive A similar error message will be printed in case of hard disk errors Because DISKATE will automatically compact the disk if an existing file is too short if you have a DOS or NO OVER other piece of software that can directly manipulate the LAPPING directory you MUST NEVER CREATE OVERLAPPING FILES as FILES the DISKATE compacting routine will not work if files ALLOWED overlap Of course if all of the files on a diskette were created using DISKATE this will be no problem Any interval can be given as the second argument of DISKATE User s Manual an S command The general form of the S command is S filename interval Here the interval MUST be given you can t use the de fault argument for the disk commands to prevent errors
17. a Control C In any event if the panic detect routine returns to DISKATE signalling the condition that you desire a process to be stopped DISKATE will STOP AND WAIT FOR THE NEXT CHARACTER TO BE INPUT FROM THE KEYBOARD If this character is ESC 1BH then the ongoing DISKATE process will be TERMINATED A question mark will be printed on the screen to echo the abort If you type S instead of ESC S for Single step the next character of the output will be printed but the panic state will remain in effect Thus a Single character will be printed but prior to the output of the next character DISKATE will automatically go into the panic state as if you had again interrupted it Thus if you want to slow down a printout interrupt it by invoking the panic detect routine and then repeatedly type S or if your terminal will transmit a character repeatedly transmit the S This way the characters will be printed Slowly enough for you to be able to read them as they are being printed If you are in a panic state and you type any character other than S or ESC the process will continue full steam Recall that you can force entry to the panic state by executing the PAUSE command If a printout from a command is interrupted by invoking the panic detect and then aborted there is a way to continue the printout where you left off The 66 ECHO COMMAND PRINTS ARGUMENT ON TER MINAL PANIC DETECT PROVIDED RESPONDS BEST TO BREAK U
18. instance but what if you want to make the change in several places It frequently happens in text editing that you have to change EVERY occurrence of a part of the text to something else For instance you might discover that you had systematically misspelled a word every time you used it or in editing a program you might be forced to change the name of a variable which would mean changing it everywhere it occurred DISKATE provides very powerful commands for performing this kind of task Of course this power if used erroneously can leave you with a monster job of cleaning up so great care must be used to make sure these powerful commands do not themselves contain errors B Let s start with a simple example Suppose that we have a file which is a list of entries each one having the same format as our dog eared JOAN DOE file After DISKATE User s Manual 36 numerous complaints from the clientele we are pursuaded that it isn t good form at all to keep maintaining a line in each entry for the sex of the creature described This is the seventies after all So every line in the entire file containing the pattern SEX must go Before showing how it all can be done with a single command let s see how we would get rid of these lines one at a time The first such line in the file could be eliminated by the command K SEX Now what about the next such line Remember that a pattern matching argument such as SEX wi
19. s Manual the value of a variable MEMTEST and which takes as an argument the interval of memory to be tested To test the block of memory from 3900H to 31FFH the program could be invoked by the command X MEMTEST 3000H 31FFH Be sure to leave a blank between the first argument of the X command which gives the address of the subroutine to be executed and the second argument which gives the subroutine s argument The subroutine can then obtain the value of the argument by calling CVALS CVALS will put the lower address of the argument in the HL register pair and the upper address in DE What if you want to evaluate an argument and have control passed back to your machine language program in case of an error In this case you can call VALUS in stead of CVALS This routine works like CVALS but will return with the Z flag OFF if there is an error in evaluating the argument There is one important other difference You must supply to VALUS the addresses of the initial reference interval yourself the lower address in HL and the upper address in DE Note that this can be used to have the machine language program determine the initial reference interval and the argu ment within it passed from the terminal or a macro You don t need to supply the initial reference interval if you can be certain that an argument will not contain any matching symbols but this is risky since if the argu ment contained matching symbols by mistake yo
20. simpler command l 20 a Let s see why this works If you think of asan operator connecting two intervals then the argument AS 20 would appear to be incomplete the second MATCHING interval is missing Instead of viewing as an ARG operator it can be viewed as a MATCHING ARGUMENT WHICH MATCHES MATCHES ANYTHING The argument 2 can be read as ANYTHING the line containing the 20th carriage return back from the end of the file followed by PLL Ag Oe ae eo ei Because DISKATE is eziing with the current file the term anything is restricted in scope to the current file Notice also there is a construct here which we haven t seen before a matching argument consisting of one pattern followed by another pattern This is called CONCAT CONCATENATION Writing one pattern directly after ENATION another specifies the pattern obtained by joining the _ two patterns OIDE ee The pattern wean DOE is PERDE equivalent to the patteri DISKATE User s Manual 2 JOAN DOE Concatenating two patterns of the form textl text2 is not very useful because instead you could just use the pattern textltext2 Concatenation is very important however when the two patterns are specified by differ ent means A pattern of the form text specifies the pattern consisting of text followed by a carriage return There is no way to specify this pattern directly in the form text because a pattern of t
21. 116 144 V Command 119 139 VALUS 115 143 Variables 27 898 122 VCHK 114 143 W Command 77 136 WHAT 143 WID Command 65 134 X Command 63 132 Y Command 65 134 Z Command 88 138 Z gt Command 88 138 suffix 13 124 as file name 76 126 Command 5 127 Command 54 129 codes 54 56 123 argument 84 124 Command 84 138 suffix 9 125 argument 83 124 Command 84 138 Command 5 127 separating commands on one line 25 1M RM denM 29 122 Command 45 1321 29 122 gt 29 122r l 14 See wee 6 11 123 125 29 122 53 68 122 126 lt 21 115 124 lt F gt 31 124 149 DISKATE User s Manual lt R gt 75 124 lt S gt 42 124 lt T gt 79 See Symbol Table Command 27 133 gt prompt character 1 Upper address of last interval 19 22 31 115 124 argument symbol 46 66 124 Command 57 129 2 Command 71 129 as file name 88 126 matching symbol 23 123 text matching symbol 6 123 echoed by ESC 4 argument 25 124 Command 24 129 in ASC Pseudo operand 87 in command output 5 7 8 123 T 18 22 125 gt gt gt FO 158
22. 29 In addition the entire source code field is offset with respect to the left margin to allow room for the object code This offset is normally 21 These values can be changed by giving the TAB command If the TAB command is given with no argument the current tab settings will be printed on the terminal first the four settings for the four fields above and last the source code offset To change the settings give the TAB command with an argument consisting of a list of numbers separated by blanks the order of the numbers again being the four fields above and lastly the offset that the tab stops for the four fields are each WITH RESPECT TO THE MARGIN THE SOURCE CODE OFFSET and not with respect to the previous tab stop In the list of numbers given as argument to a TAB command the five numbers need not all be given Any that are omitted will keep their old value In editing assembler source programs you may want to keep the source code in as compressed a form as possible to save storage Note that the Intel format is usually more frugal with storage than the Processor Technology format since you don t need to begin a line which has no label with a blank DISKATE provides an editing command which will print an interval in the same format as that used by the assembler except there is no object code present so that the tab stops are relative to the left hand margin rather than the source code offset The line numbers will also
23. AFTER DISK COMMANDS DISKATE User s Manual 121 Part III System Reference Summary Special input characters ESC causes current line to be ingnored echoed ESC causes escape from the panic state BACKSPACE causes just typed character to be ignored Backspace space backspace is echoed S in the panic state causes one character to be printed and the panic state remains in effect Panic State A panic detect routine whose address is given by the jump at I0 12 is called periodically including prior to the output of each character to determine if a process is to be interrupted The routine provided will cause a process to be interrupted if any key is pressed provided it can be detected The Break key gives the most easily detected signal A user provided panic detect estab lishes its own conditions for determining when a process is to be interrupted If the process is to be interrupted the panic state is entered The panic state is also entered when a PAUSE command is executed or is encountered as the file name in a file name list for assembler commands When the panic state is entered the system will wait for an input If this input is the character S the process continues but the panic state is automatically still in effect the next time the panic detect routine is to be consulted Thus S will allow one character to be output during listings If the character input is ESC the process is aborted If the character
24. But now there is a problem We want to kill the A in the JOAN in the interval we just saw but how do we specify it There are different possible approaches but here is a simple one which will show another feature of DISKATE we haven t seen before In the interval that we just computed we can be absolutely sure that there is only one occurrence of JOAN since we set things up so that the command gt JOAN is showing us the interval of text until the NEXT occurrence of JOAN What we want to do is delete the A in the JOAN and we know there is only one in the interval just computed The follow ing command will do exactly that DISKATE User s Manual K JOAN A Let s take this carefully The first thing we see in this command is K which is the symbol for the Kill command The very next character in the command is a We have said that causes the part of the argument already given to become the initial reference interval But here we haven t given any argument yet at all What we re seeing is an example of an extremely useful DISKATE feature IF A COMMAND REQUIRES AN INTERVAL AS ARGUMENT AND NO ARGUMENT AT ALL IS GIVEN THE ARGUMENT USED IS THE PREVIOUS INTERVAL COMPUTED So The letter K says that this is a kill command and what follows is to be the argument The says to take that part of the argument already computed and use it as the initial reference interval Since no argument thus far has been given
25. COMMAND AT GROUND LEVEL RETURNS TO CALLING PROGRAM IF THERE IS NONE CAUSES DISKATE TO BE RELOADED FROM DISK RENT LOSES RETURN ADDRESS DISKATE User s Manual Part II Installation and Maintenance II 1 Bringing up DISKATE The version of DISKATE supplied with the Disk Jockey controller comes ready to run with a serial ter minal attached to the Disk Jockey serial I O port If you have a serial terminal attach it to this serial port and set the terminal to run at at 1200 baud Then all you have to do is bootstrap the DISKATE diskette and you should be up and running If you have a serial terminal and want to run it from another I O board best thing to do is bring up DISKATE from the serial terminal on the Disk Jocky then use DISKATE itself to customize your I O routines If your terminal hardware is not simply a serial terminal or if it is a serial terminal but wont run at 129 baud then you will have to make some minor patches Before describing the patch ing procedure in detail here is the general structure of how DISKATE is brought up As mentioned in Part I DISKATE consists of two independent but linked modules IO and ATE IO contains all of the I O drivers includ ing those for the disk It also contains a loader which looks in the directory for ATE which IT ASSUMES IS THE SECOND ENTRY IN THE DIRECTORY and loads it IO writes a jump to the beginning of IO at memory location 3 jump is the only link ATE has
26. DEVICE as the I O multiple input or output devices your terminal routines can simply ignore the device num ber in the A register and this discussion won t apply You can execute the terminal initialization routine at any time by giving the Y command Y for Wipe which takes no argument A device number is presented to the terminal initialization routine as for any of the other terminal I O routines If you have multiple devices you may want your terminal initialization routine to initialize all of them at once or you may want it to intialize only the device whose number is passed to it If you do it this second way then when ever you use a new device by means of an IO command you will have to remember to give a Y command unless the device needs no initializing All this is discussed more fully in II 1 Another command pertaining to the terminal allows you to set the terminal width This is the WID command If you give the WID command with no argument it will print the current terminal width while if you supply an argument the low order byte of the value of the argu Ment will be set as the terminal width The widest line which the line buffer can hold is 1398 characters though if a WID command is given for a value greater than this the value will be set as the width of the terminal for output purposes This width determines how many charac ters are output before DISKATE automatically inserts a carriage
27. If the statement takes an operand the operand field is separated from the opcode field by one or more blanks d The comment field is separated from the previous field by one or more blanks e The entire line can be designated as a comment by giving as the first character on the line The INTEL format differs from this in the following respects a A label is terminated with a colon If a statement is not to be given a label the opcode may begin the line since the absence of the colon following opcode designates it as the opcode field b The comment field begins with a semicolon In both formats the mnemonics for the machine instruc tion opcodes are the same which are referred to as INTEL MNEMONICS On power up DISKATE will recognize the Processor Technology format unless you have customized your system so that it comes up recognizing the Intel format The method for doing this is described in Part II You can change the format for the assembler using INTE for Intel format and PROS for Processor Technology These can either be given from the terminal as commands with no argument or as pseudo ops in a Source program Thus a program can be a mosaic of segments in both formats When the assembler begins Pass 1 it will be going ae through your source statements counting where the bytes of object code will go but without actually putting any thing anywhere The assembler needs to know where the In DISKA
28. SEX This pattern would also match a line containing the name THOMAS WESSEX for instance which is not one of the lines we want to delete Boundary characters such as spaces punctuation and carriage returns are often very important in specifying patterns that match the right thing DISKATE User s Manual An obvious question here is what to do when you don t Know how many times the command will have to be repeated Actually you can program DISKATE to count the number of times a pattern occurs and we ll see below how to do that all that work pattern is to give a repetition number which you can be Sure is greater than the number of times the pattern will occur If we know for instance that there are not more than a few hundred entries in the file but don t know exactly how many we could use the command R9999 K SEX Assuming we re correct that there are not nearly this many occurrences of the pattern SEX in the file the command will be repeated untill all of the occurrences of the pattern have been deleted Then what The pattern SEX will fail to match anything so an error message will occur Though this method will work perfectly well the obvious problem is that we have to assume that the error message is given because DISKATE ran out of occurrences of the pattern to match rather than because of some error that should really concern us It s better not to have to rely On an error to terminate a proce
29. SORG statements By storing the object code in a file called PRNTSYM the symbol table can be printed by giving the DISKATE command GO PRNTSYM which will load the object code and then execute the equivalent to the command X lt R gt The RET at the end of the program will return control to DISKATE After the assembly the object code can be stored by the command S PRNTSYM PRNTSYM SYMEND Notice that in this command the first time PRNTSYM is used it is a file name and the second time it is used it is a symbol in the symbol table There is no ambiguity here but if you find it confusing you can use a different name for the file containing the program and an assembler label for the program itself Here is the listing of the program exactly as gen erated by the assembler The source code was assembled directly from the disk file PRNTSYMS and the tab set tings were adjusted by changing the last setting i e the source code offset to 15 gt A PRNTSYMS PRNTSYMS PRNTSYMS 0 0 ql INTE 2 4 PROGRAM TO PRINT 5 SYMBOL TABLE ee Se oo go Z 7FOO ee ro ORG 7F 99H 7F60 2A 5B 45 9 PRNTSYM LHLD SYMTB SET POINTER PE EA 10 7F93 7E ill SYMBEGIN Mov AMO GET SYMBOL opg BT eng C ORA A SET FLAGS 7F 5 CA 38 7F 133 JZ SYMEND QUIT IF i ee EEE ee i i _7F 8 7E 15 SYMLOOP MOV A M GET SYMBOL LETTER 7FQ9 47 l6 MOV B A SAVE COPY 7FOA E6 86 17 gt gt ANI 80H CHECK HIGH mi
30. SPEED CONSTANT LOAD IN TERMINAL INITIAL IZATION ROUTINE ADDRESS OF THE ROUTINE AT I0 19 DEVICE NUMBER IN A REGISTER CHAR OUT PRINTS DISKATE User s Manual 2 The character output routine should print on the terminal the character supplied in the B register Be very careful not to make the mistake of printing the character in the A register remember this is the device number 3 The character input routine should return the input character in the A register replacing the device num ber 4 The terminal initialization routine is executed only in response to the Y command and at power up If you have multiple I O devices you may want to initialize all of them in response to any call to the initialization routine Or you may want to initialize only the device whose number is in the accumulator If you have this routine initialize only the device whose number is in the accumulator then you will have to remember to issue a Y command before using any devices other than The IO command will issue a carriage return and line feed to the device given as argument Thus if you initialize only the device whose number is in the A register the first time you give an IO command for a non zero device this carriage return and line feed will be passed to the character out driver for a device which has not been initialized This will not cause a problem if the driver for this device simply loses the carriage return and l
31. The machine language program can pass a string of commands for DISKATE to execute and if these commands include an ATE command further commands can be entered from the terminal just as if DISKATE were being executed from the power up state To use this method you must NOT call the entry point ATE since this will initialize the stack and lose the return address Rather you call a Special entry point called ATECOMS This entry point assumes that the HL register pair contains the beginning address of a string of DISKATE commands It will execute the string then return back to the calling program If the command string results in an error the error message will be printed on the terminal but con trol will still return to the calling program The one thing you must be careful of is that the command string cannot include a RENT command nor should this be given from the keyboard during an invokation of DISKATE called by a machine language program since it will result in the return address being lost To make it easier to use this feature when DISKATE is powered up a jump to ATECOMS is written at memory location Thus DISKATE can be called by a machine language program by the following procedure 1 Load HL with the address of a string of DISKATE commands ending with a The string can include sever al command lines 2 Execute the instruction RST DISKATE will be called and will execute the command string Note that it will
32. _ i ne 1ST OCCUR where in this case text is NOT longer than one line RENCE will match the first occurrence in the current file OF CHARS of the characters comprising text OF TEXT For example Suppose the current file consists of the characters oo NAME JOAN DOE OCCUPATION DEER SEX FEMALE ADDRESS THE WOODS The argument NAME denotes the interval in the memory occupied by the characters N A M E in the first line of the file wherever that happens to be Note that this form of specifying an interval is CONTENT ORIENTED in contrast to specifying the addresses bounding the inter val explicitly The argument NAME will still denote the first occurrence of N A M E in the file even if the file is changed so that something else occupies the memory locations that N A M E once occupied An argument like NAME will do us little good as an argument for the command since the command NAME DISKATE User s Manual will simply print NAME on the terminal Simple argu ments of this form are very useful with other commands however aS we ll see What will be very useful with the command as well as others is combining simple patterns such as NAME using the operations DISKATE recognizes into much more complex ones One powerful method is to use the construct which we ve seen in connection with explicit addresses The argument NAME ADDRESS refers to the characters from the first occurrence in t
33. address Because each entry in the directory is 16 bytes long the directory has space for 64 files Thus each disk is limited to 64 DISKATE files i As with any floppy disk system a program which goes out of control and manages to dump garbage on the directory of a disk MAY RENDER THE ENTIRE DISKETTE UNREADABLE For this reason you are emphatically advised to make backups of all disks containing impor tant information If you have a dual disk system backups can be made by copying entire diskettes DISKATE has a command for doing this which we ll see below If you have only a single drive you can keep consecutive backups by STORING A FILE ON A BACKUP DISKETTE EVERY TIME YOU STORE IT ON THE WORKING DISKETTE If you consider this to be too much of a nuisance to be worth the bother it is a safe bet you will change your mind the first time you lose a whole diskette worth of information Now let s begin the discussion of the file handling commands Suppose you are working on a file which is brand new you started with an empty source area and have created the file for the first time using the editor commands and now it must be saved on the disk Let s Say you want to save it under the name SPECS This can pedone With The command S SPECS lt P ar aia Terange Bo The DISKATE file handling commands will do a lot of work automatically so we had better be careful to understand just how the commands work First of all sinc
34. address of its argument F lt S gt will make the first file in the source area current NO Now suppose we want to see the next file in the MATCHING source area As usual with DISKATE there are many ways of doing this We saw above that the command lt F gt gt 1 will set the entry pointer so that the target character is the at the end of the file That means that the argument lt F gt gt 1 will specify the at the end of current file But if another file follows the current file in the source area this will also be the which forms the lower bound of the file which follows So if we give the commana DISKATE User s Manual 43 F lt F gt gt 1 it will set the current file to be the file following what had been the current file Don t be confused about F appearing in both the command and the argument this is just the same as the principle behind a statement like X X 1 Or we could just as well have given the command F 1 There is just one problem with this method Sup pose you just don t remember which is the last file in the source area so that you aren t able to recognize the last file when you see it The command F lt F gt gt 1 calls for no matching and neither does F 1 so that if the current file happens to be the last one in the source area and you execute this command anyway it may not return an error message but instead will give you a file consisting of garbage This could g
35. again up to the point you gave the load command but this is a lot of work and anyway you may not be able to reconstruct how it was done DISKATE has a special symbol which will help in this case lt R gt is a symbol analogous to lt S gt and denotes the INTERVAL IN MEMORY OCCUPIED BY THE LAST RECORD I E FILE READ IN FROM DISK If you are combining source files by the technique above and realize that the last L command given was in error that part of the current file that was put there by the last L command can be deleted by the command K lt R gt Because the K command will position the entry pointer to the end of the deleted material you can immediately follow it with the correct L command and the mistake will be corrected There are several disk commands which can be used to copy information or change the status of files If you want to CHANGE THE CURRENT DRIVE the command CD which takes as its argument a disk drive designator will establish the drive indicated by the argument as the current drive In Part II we will see how to customize your system so that DISKATE knows which drive designators are valid It is important that this be done correctly or else a CD command to a non existent drive will cause very bad mischief If you have followed the customization procedure outlined in Part II you will not have problems but here is what will happen if you give a CD command by mistake for a non existent drive DISKATE wi
36. allowed Intervals An interval is a pair of addresses with the first address lt the second address It specifies the interval in the computers memory from the lower address to the upper address inclusive A Single number is identified with an interval whose lower and upper DISKATE User s Manual 123 addresses are equal Initial Reference Interval The initial reference interval is determined by l The command and 2 The argument of the last REF command if one has been given The initial reference interval for D G and F commands is the source area For all others the initial reference interval is the current file if no REF command was given or the argument of the last REF command was empty otherwise the initial reference interval is the argument of the last REF command given as applied to the current file Matching Interval Symbols text matches the first occurrence of text within the initial reference interval text may contain any ASCII characters including non printing control characters sequence of numbers separated by blanks matches the first occurrence in the initial reference interval of the sequence of bytes whose codes in the current base are given by the numbers The numbers must be positive lt 255 and should NOT have a Qor H base suffix matches the first carriage return in the initial reference interval matches the first character in the initial reference interval
37. as 65535 which is the highest address in the memory So the argument S 1 3 means everything up to the 3rd carriage return within the interval starting just beyond the source area and extending to the end of memory The deleted text can be restored by s 1l1 3 In this case the is given as a precaution to 69 M COMMAND WITH NO ARG JUST AFTER K COMMAND RESTORES DELETED TEXT s GIVES FIRST BYTE BEYOND SOURCE AREA COMMANDS TO RECOVER 3 DELETED LINES DISKATE User s Manual make sure we re getting what we want and saving the entry pointer was a precaution which wasn t really necessary However in general you may not be sure how to specify the deleted text In this case you may want to give or commands to try to pin down exactly where the upper bound is of the material that was deleted and being able to set the entry pointer might be useful In general if you can figure out a pattern which will match the ending of the interval you want back the commands S l pattern M will put you back on the track Note however that you must set the initial reference interval by an argument such as S l since matching will always fail outside the source area unless the initial reference interval has been set There may be instances where you want to perform a block move but you don t want DISKATE to know about it For instance it may be useful to have a separate copy i
38. at 1l1E9 H and want to move it to 2C9 H This can be done with the command 2CQ0 H M lt T gt The symbol lt T gt can also be used with the L and S commands to save and restore symbol tables This is a useful technique for assembling one program to go with another without having to assemble the two of them together For instance the command S TBLSAVE lt T gt will save the symbol table in a file called TBLSAVE Now suppose you want to restore the symbol table to the way it was just before you saved TBLSAVE There are two things that have to be done First we give the command L TBLSAVE lt P gt This will OVERWRITE the current symbol table with the contents of the file TBLSAVE Now there is one problem DISKATE will still have the old value stored internally for the upper bound of the symbol table It is unlikely that this will be the same as the upper bound for the table saved in TBLSAVE To set the upper bound of a symbol table just read in from disk the easiest method is to SET A VARIABLE NOT IN THE SYMBOL TABLE EQUAL TO So if we can be sure the variable JUNK was not in the symbol table saved in TBLSAVE the command JUNK 9 will establish the upper bound of the symbol table The reason this works is that when an command is given DISKATE searches through lt T gt from the beginning for the symbol and when it finds it assigns it the value It defines the symbol if it reaches the end of lt T gt befo
39. be set to the first character after the text loaded from the disk file There are two important considerations here First if the target character happens to be outside the source area then the file will be loaded at the location given by the entry pointer OVERWRITING WHATEVER IS THERE The source area remains unchanged Because an L command can affect a very large number of bytes an error of this kind can be catastrophic If you have been using DISKATE as a monitor to work with memory outside the source area then give an L command with no second argument for a gt source file but have forgotten to restore the entry pointer this could overlay a substantial portion of the memory and possibly clobber part of the resident soft war resulting in a crash Second even if the entry pointer is in the proper place if the file is a long AND 736 TION AS I COMMAND U COMMAND DELETES FILE FROM DISK L COMMAND LOADS FILE GIVEN AS 1ST ARGU MENT TO MEMORY LOCATION GIVEN AS 2ND ARG IF NO 2ND ARG GIVEN USES MEMORY ADDRESS FROM DIREC TORY IF NOT SOURCE INSERTS AT TARGET CHAR IF SOURCE WITHIN SOURCE AREA ELSE OVER WRITES DISKATE User s Manual one it will expand the source area considerably You should always check to make sure that the expanded source area will still fit in the area of your memory that is safe for the source area to occupy If you want to work with on
40. byte segment in the compu ter s memory for special drivers or interrupt handling or the like The first address that you might have a program occupy then could be specified as 1 04 This means the number 1 256 Note that since no base was specified DISKATE assumes decimal Of course numbers in hexadecimal do not need to be split since the hexadecimal form is naturally split already Numbers in hex are indicated by adding the character asa suffix to the number to H LLN The number 1 8 is equivalent 1008 in hee To specify a DonBEr in setal or base 8 the suffix Q is added to ene end of the MORDER e So for example the numgSrSs 53 CURRENT BASE IS BASE FOR OUT PUT OF NUMBERS FOR NUMBER INPUT BASE MUST BE GIVEN DEFAULT IS DECIMAL GIVES n 256 m SPLIT FORM SUFFIX H FOR HEX Q FOR must be used DISKATE User s Manual 2680H 40 99980 32 0 all specify the same number which is 32 256 There is one important point about hex numbers A HEXADECIMAL NUMBER MUST BEGIN WITH A DECIMAL DIGIT Otherwise there would be difficulty distinguishing it from a variable name Thus FFFFH is not a valid way to specify a number for DISKATE you would have to use FFFFH The current base can be changed at any time by using the B command For instance to change from hexadecimal to octal you would use the command B8 When the current base is not hexadecimal
41. can be interfaced to DISKATE The details of how to do this are discussed in Part II of this manual For the moment all we need to know is the general struc ture of how DISKATE handles the I O calls The terminal I O routines involve calls to the following basic drivers l1 A routine to initialize the terminal 2 A routine get one character of input presumably from the keyboard and return it in the A register 3 A routine to output one character from the B register to the terminal 4 A panic detect routine This routine queries the terminal to see if a key has been pressed If not it returns with a flag set to indicate nothing is doing Otherwise if a key has been pressed it inspects the byte corresponding to the key and if it meets the conditions set by the user to indicate that a process should be stopped it returns with a flag set indicating this Otherwise it returns with the flag set as before indicating nothing doing The basic terminal input and output routines can handle multiple I O devices When the routines are called a device number is supplied in the A register The routine can use this number to vector the call to one of several devices Normally calls are made to device i e is in the A register when the terminal routines are called The device number can be IO changed by giving an IO command The command COMMAND ARGUMENT will cause n to be henceforth passed in the A register TO BE to th
42. change several times as a single argument is evlauated For instance in the argument JOAN A lt is first set to the address of the J in the first occurrence of JOAN in the file and then set to the address of the A within that An argument with several instances of may cause lt to be reevaluated several times during a single argument A value is not assigned to gt until either the whole argument is evaluated or a is encountered So the argument oo JOAN gt specifies the N in the first occurrence of JOAN in the current file If you are in doubt about what lt or gt should mean give DISKATE a simple test case using a command to see the result Mio ky Pm lt GIVEN VALUE AS SOON AS LOWER ADDRESS OF INTERVAL IS DETER MINED gt GIVEN VALUE WHEN WHOLE ARG EVAL UATED OR AT DISKATE User s Manual 23 Above we saw how we could see successive intervals of text ending in the next occurrence of JOAN but men tioned that by using the command gt JOAN the first character printed would be the N in JOAN It certainly does no harm to keep having this N printed since this is only a command but it would be nice to know how to get rid of it It s always good practise to be able to specify precisely the interval we want because a slight nuisance such as this initial N could turn into a signi ficant error with other commands The interval gt JOAN includes everything we want
43. command Then when you give the BYE command DISKATE returns to the process that invoked the ATE command As with COM ATE takes as argument a string enclosed in brackets which it prints on the terminal as a prompt The best way equal to the The ATE command can be used to create macros which han dle various functions automatically but where you want to leave some holes for doing work with an indefinite number of DISKATE commands yourself Here s an example of how you might use this Consider the following macro Oe at es R 1 lt F gt gt 1 ATE POINTER IS NOW AT END OF FILE S Suppose you begin your editing work by loading a file and then invoking this macro The macro will make certain that you start to work with the entry pointer at the end of the file DISKATE will issue the prompt and then you can give any number of editing commands Then 99 RENT COMMAND JUMPS TO REENTRY POINT GROUND LEVEL INITIAL IZES STACK ONLY ATE COMMAND PRINTS GIVEN AS ARGUMENT ACCEPTS ANY NUMBER OF _ COMMANDS RETURNS WHEN GIVEN BYE COMMAND DISKATE User s Manual you give the BYE command to resume with the macro This will automatically store the file on the disk restore the entry pointer to the end of the file and then issue the prompt and wait for more commands As long as none of the editing commands you type in response to the ATE command changes the latest file refer
44. command for deleting the A in JOAN in this situation which is even simpler When we used the command to set the target character to be the J in successive JOAN s we also set the last interval computed to be the whole JOAN So by the time we ve found the right JOAN we will know that the last interval computed will be this JOAN All we need to say then is K A The before any argument is given sets the initial reference interval to the last interval computed which is the correct JOAN and within this interval the first A the only A in fact is just what we want to delete Note that if we use this command for the dele tion the main function of having used the command was to be able to see where we were using the command It is often not a bad idea to use the command this way even when you are not sure whether you will need the target character to be set to the first character of the MULTIPLE argument in question COMMANDS Perhaps this is the eine to mention an important ON ONE feature of DISKATE So far we have been giving exam LINE ples of DISKATE commands one by one Actually you can SEP give DISKATE an ENTIRE LINE OF COMMANDS at once ARATED SEPARATED BY COMMAS Rather that giving two separate BY DISKATE User s Manual 26 commands COMMAS ARE gt JOAN ALLOWED it s easier to put both commands on a single line gt JOAN By now it should be clear that the
45. course this file might be empty The O command can also be given without any argu ment In this case rather than following the standard procedure of using the last argument computed an empty source area will be established beginning at the same place that the current source area begins useful for scrubbing the work you ve already done and Starting with a clean slate to bring in a new file from the disk Using the O command you can maintain several dif ferent source areas in the memory Suppose for instance you want to create a new source area at a special location which will be used by a piece of software that requires a file in a particular place If you want to be able to go back to the current source area you can save its location by f a This is very 58 COUNTING NUMBER OF TIMES A PATTERN OCCURS O COMMAND ORIGIN ATES NEW SOURCE AREA AT INTERVAL GIVEN WITH NO ARGUMENT CREATES EMPTY SOURCE AREA STARTING SAME PLACE AS CURRENT lt S gt DISKATE User s Manual OLDSOURCE1 lt S gt OLDSOURCE2 gt The new source area can be set up using the O command and if you want to go back to editing the old source area the command O OLDSOURCE1 OLDSOURCE2 will reestablish it as the source area The O command is what you want if you want to cre ate a fresh source area somewhere in memory other than where DISKATE wants lt S gt to be or if you want to make a given interval in
46. directory regarding a disk file it can be obtained by consulting the DISKATE disk buffer This buffer begins at the address DISKBUF where DISKBUF is again a symbol from ATETBL After an I command or any other disk command a copy of the entire directory of the accessed drive will occupy 1824 bytes beginning with DISKBUF DISKATE consults the disk resident copy of the directory for any disk commands this copy of the directory is left in memory only for use by user programs A machine language program can consult the directory by statements such as LXI H DIRCOM RST DIRCOM ASC L ATETBL lt T gt ASC KEEP 9 ASC I drive designator if file not on current drive ASC VDISKBUF DB Upon returning to the machine language program after the RST 6 HL will point to a copy of the directory of the current drive If the directory is needed for a differ ent drive than the current drive the command string at DIRCOM above should include an I command for the proper drive The program can page through the directory at DISKBUF entry by entry testing the first 8 bytes of each 16 byte entry for the desired file name then reading from memory the desired information Note that you can also consult the directory directly from DISKATE For instance to get a core dump of the directory type L ATETBL lt T gt KEEP 9 I n DISKBUF lt 1023 i 128 COPY OF DIREC TORY LEFT AT IST 1824 BYTES OF DISKBUF
47. e tt aneii gemet neal Aon near es O ORDER BIT 7F6C C217 7F 19 JNZ ENDSYM JUMP IF ON DISKATE User s Manual 94 7FOF 78 20 MOV A B RESTORE LETTER 7F19 CD 38 35 21 CALL OUT s PRINT LETTER 7F13 23 22 INX H INCREMENT POINTER 7F14 C3 8 7F 23 JMP SYMLOOP GET NEXT LETTER 24 25 7F17 78 26 ENDSYM MOV A B s RESTORE LETTER 7F18 E6 7F 27 ANI 7FH STRIP OFF BIT 7FIA CD 38 35 28 CALL OUT PRINT OUT LETTER 7F1D 3E 26 29 MVI A PRINT OUT A 7FIF CD 38 35 38 CALL OUT SPACE 7F22 CD 38 35 31 CALL OUT gt AND ANOTHER 7F25 23 32 INX H INCREMENT POINTER 7F26 5E 33 MOV E M LOW ORDER 34 SYMBOL VALUE 7F27 23 35 INX H INCREMENT POINTER 7F28 56 36 MOV D M 7 HIGH ORDER 7F29 23 37 INX H A INCREMENT POINTER 7F2A E5 38 PUSH H 7 SAVE POINTER 7F2B EB 39 XCHG HL CONTAINS 40 SYMBOL VALUE 7F2C CD E2 34 4i CALL PHLSB s PRINT VALUE 7F2F 3E 6D 42 MVI A 13 CARRIAGE RETURN 7F31 CD 38 35 43 CALL OUT PRINT CR LF 7834 El 44 POP H RESTORE POINTER 7F35 C3 63 7F 45 JMP SYMBEGIN GET NEXT SYMBOL 46 i 47 7F38 c9 48 SYMEND RET s RETURN TO DISKATE 49 Notice that DISKATE will generate line numbers The reason for having line numbers is to correlate with the error messages A line number MUST NOT begin a line in the original assembler source code If you want to use line numbers for your own reference in the source code and the P command will not suffice the line num bers must be part of the comment f
48. edit macros Before we execute these we want to look over the file with the letter one more time to make sure it is O K At this point the source area consists of two files the file with the letter followed by the file with the DISKATE commands The file with the commands is current The command F George will make the file with the letter current and set the target character to the end of the file Remember that the initial reference interval for the F command is the whole source area so in executing the F command DISKATE will begin searching for the pattern George at the beginning of the source area which is the file with the letter Suppose we had written three letters one each to Tom Dick and Harry and following that had a file with commands If we want to see the file with the letter to Harry we can use the command F Dear Harry It might not be a good idea to use the command F Harry since the letter to Tom might talk about Harry There is also a simple way you can page through the files without knowing what any of their contents are Suppose you want the current file to be the first file in the source area whatever that is The command lt S gt F lt S gt i a i at hee au eins aeeuMENT will do the trick lt S gt is a symbol we haven t seen DENOTING before It is analogous to lt F gt and denotes the source SOURCE area It does not call for matching Since the F com AREA mand only uses the lower
49. entered with the E command will go at the end of the file Future shipments of DISKATE will have a special symbol F which will denote the position where text should be entered to be appended to the end of the file This will make a command to do the job above much simpler We ll also see other methods for doing this same thing below There is a subtle point here which may cause some confusion In all of the previous examples we ve stressed that allows us to specify in essence an argument WITHIN an argument Here we seem to have violated this In the argument gt 1l the part gt 1 Specifies a location which is outside of the part to the left of the How can this work The answer is that gt is evaluated directly by consulting an internal DISKATE location WITHOUT ANY MATCHING When DISKATE has evaluated the argument as far as e l here is what has happened among other things lI The initial reference interval is still the current ile l 2 lt hae been assigned the address of the first charac ter in the file 3 gt has been ter in the assigned the address of the last charac file After these have taken place the value of gt 1 can be calculated directly It doesn t matter that this value lies outside the initial reference interval any more than it matters that an explicit address like 2A9 H may lie outside the initial reference interval There is an alternative symbol which can be subs
50. for other use The files are processed through a special disk buffer within the internal area occupied by DISKATE In Part II we 1ll discuss where this buffer is The file list can be given to any of the assembler commands A Al or A2 You can also include ina file list the symbol in place of a file name When the assembler encounters this it will execute a PAUSE com mand and you can use this pause to change diskettes in case a source program is too long to fit on one disk ette a Until you are used to the assembler BEFORE giving a command to invoke the assembler you should make sure you have taken care of the following points l Check to make sure amp and will be set properly either from the terminal or preferably from within the source code by ORG or AORG and SORG 2 I this is not the first assembly during the run you will have to give a Z or Z gt command to prevent labels from being doubly defined 88 SION HAS VALUE 9 END PSEUDO ENDS ASSEMBLY A COMMAND INVOKES BOTH ASSEM BLER PASSES Al PASSI ONLY A2 PASS2 ONLY NO ARG USES lt F gt ARG IS LIST OF FILES SEP BY BLANKS AS FILE NAME CAUSES PAUSE DISKATE User s Manual 3 Make sure there will be enough room in for the symbol table the memory is a simple is used in If you are assembling from disk there method for allocating memory This method source programs supplied with DISKATE and at first s
51. in preference to lt F gt gt 1 be very sure you don t confuse for There are simpler commands which will set the entry pointer to the end of the current file and we ll see them later but they require you to be certain that particular circumstances obtain and so can lead to errors Probably the safest command to use is either lt F gt gt l or 1 Once again it cannot be emphasized strongly enough that you MUST absolutely restore the entry pointer before resuming text entry if you make changes in text already entered Failing to do this will mean the new text you enter will be going in the wrong place But suppose the worst does happen and you find that you have a large block of text in the wrong place Of course this can happen just because you change your mind about how you want the text to look even if you ve made no mistakes in using DISKATE Moving text is one of the key components of editing and DISKATE has a simple move command The M command M for move will move the characters in the interval specified by the argument and insert them the same place text entered with an E command would go between the target charac ter and the preceding character Let s look at an example Suppose the current file consists of NAME JOHN DOE E OCCUPATION DEER SEX MALE ADDRESS THE WOODS and suppose we want to move the line with the address so a SAFE COMMAND TO PUT ENTRY PTR AT END OF FILE a1
52. interval is the whole source area this argument will name evaluate as exactly the beginning of our macro Of course in giving a macro a name in this way you have to be careful that the pattern macname where macname is the name of the macro does not occur anywhere in the source area ahead of the beginning of the macro Note that using this method of naming a macro using the command and executing it by D macname you can have as many macros as you want and they can all be in a single file after your text file s or they can be in several separate files It doesn t matter where they are within the source area as long as the name is properly unique The commands within a macro end in a carriage return just as if they were typed in a command line and of course there can be several on a line separated by commas When a macro is being executed when the end of the file is reached this is automatically treated as a return If the macro was invoked from the terminal by a command line control will be returned to the termin al However a macro can contain a D command so that Macros can call other macros In this case when the macro returns control is passed to the next command in the calling macro Thus the D command works like a GOSUB in BASIC We saw above that the command QF can be used to break out of a repeat loop It can also be used in a macro to return the to caller QF will quit whatever is the innermost process in which it
53. is any thing other than S or ESC the process resumes with the panic state no longer in effect Command Format n Blanks allowed anywhere EXCEPT WITHIN AN ARGUMENT A blank is mandatory to separate multiple arguments of one command Multiple commands separated by commas may be given on a single line DISKATE User s Manual 122 Number Symbols A simple number is given by a string of digits or hexadecimal digits beginning with a decimal digit preceded by a minus sign for negative numbers and suffixed immediately by no suffix decimal octal H hexadecimal A number can be given in the SPLIT FORM a b where a and b are simple numbers a b denotes 256 a b Numbers denote 16 bit values Negative numbers are represented in two s complelement form as signed 15 bit integers Addresses are 16 bit non negative integers When a negative number is treated as an address the sign bit will be treated as a digit bit Variables A variable name may be any length must contain only upper case letters and digits and must begin with a digit Variables denote 16 bit values Arithmetic Operations The arithmetic operations are performed left to right with the same priority and all operations are modulo 65536 The operation gives only the quotient discarding the remainder Arithmetic Expressions Variables and numbers may combined into an arithmetic expression using the arithmetic operations Parentheses are not
54. keep having to type in every time Let s see how to set it up as a macro so that it can be executed with a very short command line First we have to enter it into the memory We can do this with the commands N E R l QF I R I K E initial reference interval F lt S gt For the moment we ll assume that there will be only one file with the text and then a second file with the macro The F lt S gt command will restore the text file as the current file Now the question is how do we exe cute the macro Macros are executed using the DISKATE command D which stands for Do Like the F command THE INITIAL REFERENCE INTERVAL FOR THE D COMMAND IS THE i _ ENTIRE SOURCE AREA and the D command only uses the lower address of its argument To execute the macro all we have to do is figure out an argument which will spe cify where the macro is in the source area and give a D command with this argument The current file is the text file on which the macro will operate lt F gt is an argument which will specify this file and the upper address in lt F gt is the last character of the file lt F gt gt 1 would specify the forming the upper bound of the file so since the macro immediately follows the text file lt F gt gt 2 will specify the first character of the macro So every time we want to execute the macro we can type the command D lt F gt gt 2 There is something very unsatisfying about this The command D lt F
55. lt T gt S SYSIO B lt F gt S SYSIO lt F gt Note that this procedure does not necessarily call for you to transfer every file to the new disk The rest are discretionary It s very important though that whenever you reassemble IO you also save the symbol table as the file IOTBL gt and so forth DISKATE User s Manual II 2 Personalizing DISKATE Settings When DISKATE is brought up many DISKATE variables will already have been given values Because these values serve to allocate the memory you may wish to change their power up settings In working with the assembler for instance memory must be allocated for the source area the symbol table and the area of memory into which the object code is to be placed These may require different amounts of area depending on the type of job For complex programs the symbol table area will have to be quite large perhaps upwards of several K On the other hand when editing source files where there is no question of assembly the only requirement for a symbol table is to hold variables to be used in commands and macros so that a symbol table of only 512 bytes or less may suffice DISKATE provides two ways that your own choice of values at power up time can be established and both of them are both simple and systematic o The easiest way to establish your own settings it to include commands for them in the STARTUP file Suppose you want the memory at power up to be allocated
56. most process be it a repeat loop or a macro The QS command can be used in repeat loops in commands entered from the terminal just like the QF command If the QS command is given with no argument the argument used will be the previous interval computed as with any other DISKATE command This argument can be presumed not to have given a matching failure so that QS WITH NO ARGUMENT CAN BE USED AS A RETURN COMMAND This is especially useful if you want to have several macros in a Single file Let s put some of these pieces together and look at a macro which performs a function which occurs frequent ly in text editing formatting paragraphs After mak ing editing changes in prose text it will often happen that the margins of the lines will have to be readjust ed The macro we will look at will have three parameters which are given by setting the values for the variables X Y and Z Z will give the number of characters that we want as the maximum line length and X Y will serve as an interval forming the boundaries of the paragraph Once X Y and Z are set we want the command D PARAGRAPH to adjust the right hand margins The macro will work in two stages First we will delete all the carriage returns in the paragraph replacing them with blanks This will in effect turn the paragraph into one enormous line The second stage will chop it up into lines of the proper length Each of the two stages will have its own s
57. occurs whether this is a repeat loop or a macro if its argu ment produces a matching failure There is a special argument that can be used in connection with macros As a macro is executed a special command interpretation pointer is maintained l internally by DISKATE If the macro aborts in an error AS ARG the value of this pointer is saved in a special loca SYMBOL DISKATE User s Manual tion The contents of this location are denoted by the symbol which does not call for matching In other words is an argument which gives the address of the character of a macro that caused an error You can use this to help debug a macro which produces errors For instance if we had a macro in the source area which began with the command IRI and this macro was produ cing errors we could find the point where the error occurred by the command lt S gt IRI This would print the macro on the terminal up to the point that caused the error When the R command was introduced we said that it would cause the rest of the command to be repeated This is correct if the command is entered from the ter minal but if an R command occurs as part of a macro it will cause THE ENTIRE REST OF THE MACRO to be repeated until either an end of file or a quit command is encountered Analagous to the QF command is the QS command which stands for Quit on Success If its argument does not produce a matching failure it will quit the inner
58. oe matches anything in the initial reference interval Any number of consecutive dots can be used Non Matching Interval Symbols exp an arithmetic expression specifies the address given by the value of the expression Expressions are treated as non negative 16 bit numbers DISKATE User s Manual 124 lt F gt lt S gt lt T gt lt R gt exp denotes the current file See below under Memory Files denotes the source area See below under Memory Files denotes the symbol table See below under Symbol Table denotes the last record read from disk See below under Disk File Reference Symbols denotes the address next after the end of the source area denotes the address next after the end of the symbol table denotes the address given by the entry pointer The char acter whose address is the value of is called the target character The value of is given by the 16 bit number stored at CHPTR where CHPTR is a symbol from ATETBL denotes the lower address of the last interval computed This is assigned a value as soon the lower address is determined even if the upper address has not yet been determined The value of lt is given by the 16 bit number stored at Pl where Pl is a symbol from ATETBL denotes the upper address of the last interval computed Assigned a value when the complete argument is evaluated or is encountered The value of gt is given by a 16 bit number stored at P2 where P2 is a sym
59. of how to march through your file seeing successive instances of a given text pattern Another way to do it is using variables for occurrence numbers Suppose for instance that your file is a list of lines each one having a similar format to NAME JOAN DOE OCCUPATION DEER SEX FEMALE ADDRESS THE WOODS and you wanted to page through successive entries To do it using variable occurrence numbers begin with the command X 1 Then to see the next entry you can use the command X NAME 4_ X X 1 Note once again that we must use the pattern matching here rather than which specifies an absolute line number One slight drawback to this method is that in evaluating the argument X NAME 4 DISKATE must search the entire current file from the beginning each time so it will tend to be a little slow The we haven t encountered before so it s worth taking a minute to understand what it does On the surface a command like X X l appears straightforward it simply adds one to the value of X However DISKATE arguments are generally intervals and a single number is simply an interval where both addresses are the same can be used with any argument so we should clarify what it means with a general interval An argument of the form Tiled SPOR cle yest die vy ithe aearvelue is evaluated by adding value to BOTH addresses of arg In our ranger Stree s for instance the command NAME 1 will print A M E spa
60. of the form n as we have done will not guarantee that first line of the resulting interval is the entire line from which it was taken CAN USE AS ARGUMENT OF E COMMAND TO ENTER CARRIAGE RETURN OCCUR RENCING nPATTERN MATCHES nTH OCUR RENCE OF PATTERN DISKATE User s Manual 9 For instance the command JOAN 4_ would print on the terminal JOAN DOE OCCUPATION DEER SEX FEMALE ADDRESS THE WOODS there is nothing in the argument JOAN 4 which instructs DISKATE to begin the interval with the begin ning of a line We ll see how to get entire lines ina minute Remember that a construct like pattern n will give you whatever matches pattern through the NEXT n line endings A common error is to assume it will give you the interval from pattern to the n th line ending in the current file which is not in general what will happen As we just saw we will often want to get entire SUFFIX lines The easiest way to do this is with the oper ator By adding the suffix to an argument it will EXPANDS normally expand the argument to an entire line The TO command WHOLE a eee 4 LINE JOAN used with our now weary current file will print on the terminal NAME JOAN DOE Now Suppose we want to see 4 lines beginning with the first line containing JOAN Here we ve got to be careful The command JOAN 8 64 will not give us what we want but will give an error message T
61. opcode and be sure to enter the speed constant as a 16 bit address low order byte first 3 Jump to the beginning of IO There is no need to patch anything in the jump table Once you have DISKATE up for the first time the next thing you will want to do is to customize your I O routines unless you should happen to want to run from the on board serial port at 1208 baud indefinitely Weve seen that the module IO begins with a jump table and that the four jumps after the first jump to the loader are for the terminal I O routines Here are the guidelines you will need for your own terminal routines l All of the four terminal I O routines will be passed a number in the A register which will serve as a DEVICE NUMBER Initially this device number is and is changed by an IO command If you don t have multiple I O devices this feature can be ignored If you have both a video terminal of some kind and a printer for example then you could have the output routine check the A register for 9 branching to the printer driver if the A register is not Note that the IO command will cause its argument to be supplied to all the terminal I O routines in the A register Thus your character input routine should always go to the keyboard unless you have more than one keyboard You can have up to 256 different devices the maximum number of codes that can be contained in the A register 193 OTHER THAN 1200 PATCH
62. some changes resetting the entry pointer will then match the portion of the current paragraph from the entry pointer onwards You must be very careful when using this editing method since the REF command CHANGES THE MEANING of the symbol In fact virtually every matching argument may have a different meaning if the igre command feature is being used In ety the same War the pair of commands REF E oe DEF 1 ahaa aa le Sear be used to edit line by line The REF command establishes the current line i e the line contain _ ing the entry pointer as the initial reference inter 5 REF INTERVAL FOR ALL COMMANDS NORMALLY TAKING IT AS lt F gt TO BE arg REF NO ARG SHOWS LAST ARG GIVEN REF DISCON TINUES NOTE REF CHANGES THE MEANING OF AND OTHER MATCHING SYMBOLS COMMANDS TO EDIT DISKATE User s Manual ae val By giving the default command carriage return LINE BY with an empty command line the entry pointer will be LINE set to the next line and that line will be printed on the terminal Note that when giving the default command the part of the default command 1 sets the entry pointer to the address of the byte beyond the first carriage return in the initial reference interval which because of the REF command is the line containing the entry pointer rather than the whole source area DISKATE User s Manual I 2 Using DISKATE as a Monitor
63. the immediately after the K will cause the initial reference interval to be the previous interval computed That interval remember went from the N in the JOAN before the one we want to the next JOAN which is the one from which we want to delete the A So the argument up through JOAN specifies the first occurrence of JOAN in the previous interval computed which is just the JOAN we want The next sets this that is the JOAN as the initial reference interval so that the total argument JOAN A specifies l the A within the first JOAN within the previous interval computed before this command is executed This is exactly what we want so the K JOAN A does the trick Note that this is an example of using more than one in a single command You can use the repeatedly in this way to keep narrowing down the scope of matching to get what you want We saw that the symbol gt can be used to give the upper address of the last interval computed by DISKATE There is a corresponding symbol which denotes THE LOWER ADDRESS OF THE LAST INTERVAL COMPUTED lt Here are some ways this can be used Suppose you just saw an interval using a command and it didn t amount to a whole line and you want to see the line containing this interval The command neg oe will ae Notice that the command right after the sets the initial will not work The 21 MISSING ARGUMENT USES PREVIOUS INTERVAL COMPUTED
64. the storage counter is left completely untouched until Pass cases you will and from within Another pseudo op that affects both amp and is ORG The value of separating the function of amp and is that it allows you to assemble programs which will execute from a place in the memory currently occupied by some 84 GIVES STORAGE COUNTER LOCATION WHERE INSTRUC TIONS ARE STORED BY PASS 2 amp COMMAND SETS amp TO ARGUMENT COMMAND SETS TO ARGUMENT AORG PSEUDO SETS amp TO OPERAND SORG SETS TO OPERAND DISKATE User s Manual thing else in the DISKATE environment In such a case you want the object code to be stored in a different place from the place where it will go when the program is run Here you will need separate AORG and SORG Statements However in many cases you will be able to assemble a program directly into place This will have the advantage that you can begin debugging immediately after assembly while the symbol table is still in the memory In this type of situation you will probably want to use the ORG statement The ORG statement resets both amp and by the following rule 1 amp is reset from the value it had before the ORG which we ll call old amp to the value of the operand of the ORG statement which we 1ll call new 2 is reset to new amp old amp In other words first amp is reset to the operand of the ORG state
65. the maximum number of characters which can be stored in the buffer This number should not include the length of the prompt string A subroutine which determines if an argument follows the invokation of a machine language program Returns with the Z flag OFF if an argument is present A subroutine which evaluates an interval argument after the invokation of a machine language program The lower address is placed in HL and the upper address in DE In case of error it exits via WHAT rather than returning to the machine language program that called it A subroutine similar to CVALS which evaluates an inter val argument after the invokation of a machine language program This routine will return to the calling pro gram even in case of error Returning with the Z flag OFF indicates an error When called the lower and upper addreses respectively of the initial reference interval must be supplied in HL and DE The initial reference interval need not be given if it is known the argument Win concern no marching a A subroutine which prints the character in A at the terminal The I O device number is whatever has been established by an IO command and is at power up Before printing the character the panic detect routine is called All registers and flags are preserved a Tine feed is supplied automatically after a carriage return The internal print head counter is updated When the terminal width is reached a carri
66. then GO file is equivalent to L file X lt R gt The same cautions that applied to the L command apply here but in addition you will have lots of trouble if the file type has somehow become incorrect If you are processing a source file with other software which requires it to be of a non type and then forget to retype it as 9 a GO command to that file will cause DISKATE to call it as a machine language subroutine which will almost certainly result in a crash Note that just as with the L command you must be careful to make sure that is in the right place or you can clobber part of the memory with a GO command to a source file If the file is an edit macro to be used to work on the current file SAFE PROCDURE IS TO GIVE AN N ag hr W COMMAND WRITES VALUE OF 2ND ARG IN MEMORY ADDRESS PART OF DIREC TORY FOR FILE GIVEN AS IST ARG GO COMMAND LOADS FILE AND EXECUTES IT EQUIVA LENT TO L amp D COMMANDS FOR SOURCE FILES L amp X FOR NON SOURCE DISKATE User s Manual COMMAND BEFORE THE GO COMMAND Note that in this case the macro will have to restablish the file that was current before the N command as the current file 278 DISKATE User s Manual I 4 Using DISKATE as an Assembler Space will not permit this to be a general textbook on 8888 assembler programming It will be assumed that you are at least roughly familiar with what an assembler is all about and that you ve had som
67. to see but what we want is the characters from the interval gt JOAN starting with the first character after the initial N The command below will be an effective replacement for the gt JOAN eliminating the initial N problem It introduces a new matching symbol which we haven t seen before gt JOAN 2 The symbol will match ANY SINGLE CHARACTER As MATCHES an argum nt all by itself it matches the first character ANY in the initial reference interval In the command SINGLE above the argument to the left of the is exactly what CHAR we used before so let s look at the part to the right of the 2 matches the 2nd occurrence of the pattern or the 2nd occurrence of any character In other words 2 matches THE 2ND CHARACTER in the initial n reference interval 2 therefore matches everything MATCHES in the initial reference interval starting with the 2nd THE character If we re using the command above as a nTH CHAR replacement for the gt JOAN before the command above is given the last interval computed will presumably be the interval of text starting with the first character after an occurrence of JOAN or the first character in the current file if we are just Starting out So gt JOAN will match everything from the N in the cur rent JOAN to the next JOAN The sets this as the ini tial reference interval so the argument gt JOAN 2 _ becomes everything starting with the seco
68. you must be extremely careful when you do this Here is the problem In many cases you will want to enter new text at a given RELATIVE LOCATION within the source file If you save the contents of the entry pointer in a variable make a change in the text and then restore the entry pointer to the original address the entry pointer may not be where you want it to be if the change you made in the text changes its size In this case the entry pointer TO must be restored by invoking a command which sets the RESTORE entry pointer to a pattern matching argument We can ENTRY see how to do this using commands that we already know POINTER Suppose there is a simple pattern that you can be sure TO does not occur anywhere in your text such as RELATIVE Before changing the entry pointer to make the changes in LOCATION your file give the command ENTER E 1 l a UNUSED l se a a gt ie eo er a DATTEREN then after the changes have been made the command BEFORE K 11 Se CHANGE will remove the exclamation points and restore the entry THEN pointer in a single step Of course the entry pointer KILL IT will end up in the wrong place if this pattern should symbol DISKATE User s Manual actually occur in the file higher up We ll have a great deal to say below about the special case of setting the entry pointer to the end of the file Variables can also be used as occurrence numbers We saw above examples
69. 1 M 788 908 Why on earth should you go through all the work that we did when such a simple command will work The answer is that the simple command will work because we happen to know it is correct When you are working on editing in actual usage you may be just as likely to make a mistake in a command such as the one just above as you were to make the mistake which this command is supposed to correct The simple command above gives you no feedback at all to see if the operands are specified correctly before tha actual move takes place There is nothing more frustrating in text editing than to make a mistake in entering a command which is simply aimed at correcting a still previous mistake Of course the more experience you get with DISKATE the more you will be able to use such simplified commands with no trouble By now we have presented all of the most basic DISKATE text editing commands The editing commands that are presented below are on a more advanced level and allow you to perform editing tasks repetitively and create your own edit macros programs that accom plish complex editing functions You should probably attempt to familiarize yourself with the use of the commands discussed so far before going on to try the techniques that are going to be presented now Above we saw that you can replace a portion of text with something else by giving a K command followed by an E command This is fine if you want to change a single
70. 6 DISKATE User s Manual Disk Drive Designator 67 126 Disk Drivers 195 DISKBUF 128 145 DS Pseudo 86 141 DW Pseudo 86 148 E command 1 8 12 54 55 127 ECHO command 66 134 END symbol in ATETBL 144 END Pseudo 88 141 End of the file 14 31 42 Entry Pointer 1 See target character Recovery from being in wrong place 34 EOFP 116 144 _EOSAP 115 144 ERSAV 145 EQU Pseudo 85 141 ESC 4 66 121 F Command 41 138 File names disk 67 126 File Type disk 68 Files memory 1 46 43 126 FS Command 72 136 G Command 44 131 Global Search and Replace 498 GO Command 77 135 H 53 122 Hexadecimal numbers 53 I Command 72 135 IF Pseudo 87 141 Indexed Sequential Files 147 INECO 143 Initial Reference Interval 18 41 45 48 5l INTE Command Pseudo 83 998 139 141 Input Routine Character 194 Intel Format 83 149 Interval 5 122 Interval Argument 125 IO Command 64 134 7 IO Jump Table 191 196 IO Module 97 98 101 195 J Command 89 137 aaa K Command 15 59 61 leat KEEP dummy variable see JUNK 89 L assembler error code 91 142 L Command 73 134 M assembler error code 91 142 M Command 32 59 por 128 Macros 44 N Command 48 129 Matching 6 18 Matching Failure 39 40 123 147 DISKATE User s Manual MAXBS 145 Memory Address in directory entry 68 Number Symbols 122 Negative numbe
71. ATETBL before assembling your program by the same proce dure we saw in the last section The routine to be called to see if an argument is present is called VCHK It returns with the Z flag OFF if an argument is pre sent otherwise the Z flag is on This allows writing a program which takes special action in the default case of no argument If an argument is present you can call the subrou tine called CVALS to evaluate it It is important that a blank separate the argument and the address of the subroutine in the case of an X command or the file name in the case of a GO command CVALS will return with the lower address of the argument in HL and the upper address in DE unless an error occurs In this case CVALS WILL NOT RETURN AT ALL but will exit by jumping to the same error exit used by DISKATE commands which have an error This exit routine is called WHAT and its address is also given in ATETBL You could trap all DISKATE errors by patching at the location of WHAT though this is not recommended except for experienced programmers For example you might have a memory test program located at an address in memory which has been set as They Interfacing DISKATE to Machine Language Programs 114 CALL VCHK TO SEE IF ARG PRESENT Z FLAG OFF IF THERE IS AN ARG GET ADDR OF VCHK F ROM ATETBL CALL CVALS TO EVALUATE ARG RETURNS LOWER ADDR IN HL UPPER IN DE JUMPS TO WHAT IF ERROR DISKATE User
72. BAT This would print the whole current file up to the 2nd occurrence of A which would be a way to make sure that the 2nd A is really what we want But a more serious problem with this approach is that in a long file it would hardly be convenient to have to count occurrences DISKATE User s Manual just to specify the one we want At this point we are really getting to the whole crux of using DISKATE as a text editor Because matching is content oriented it allows us to do some incredibly powerful things but at the same time we have to take some care to be sure that when an interval is specified by matching it is the one we want Although this particular change is simple and we ve already seen one way that will work let s look at several different methods that will be very helpful ina great variety of situations Let s suppose for the moment that we can be confi dent that the JOAN in which we want to delete the A is the first occurrence of JOAN in the file Here of course we can be because the file is so simple The thing that we want to do is delete the A in JOAN DISKATE has a construct which reflects exactly this meaning We have said repeatedly that in determining intervals using matching arguments DISKATE works with the current file Actually this has been an oversimpli fication which normally does no harm but at this point we need to be more precise Whenever an argument involves matching before the argument is eva
73. Disk ATE User s Manual Manual Rev 1 1 By Jim Rosenberg Disk ATE C 1978 by Gary Fitts INTRODUCTION DISKATE is a powerful general purpose program development system for the 8 8 which can be used as a text editor a monitor an assembler and even an operating system The text editor and monitor allow powerful content oriented commands The entire sys tem is programmable through variables repetitive loops and com mand macros The assembler provides great flexibility by allowing its two passes to be invoked separately The user has full control over all memory allocation including placement of the symbol table Source programs object programs and symbol tables can all be stored on and retrieved from the disk The disk commands auto matically allocate and deallocate disk space handling compacting of the disk whenver necessary Although configured for the Disk Jockey disk controller board all I O calls are handled through special vectors so that if you have some other type of hardware this manual will explain how DISKATE can be customized The manual is divided into three major parts Part I called Usage explains how to use the various features of DISKATE You will want to read this carefully before beginning to use DISKATE but once you are familiar with the program you will probably con sult this section only infrequently The right hand margin of this part has left space for important keywords so that you can quickly
74. E when DISKATE is brought up from this version our user command table will be there The second enang that must be done is to include the com _ mand L PRNTSYM in the STARTUP file for otherwise the code for the user command PSYM will not be in the memory at power up and giving the command PSYM would probably cause a crash A In general there are three things to keep in mind when you create a new user defined command ic The code for the new part of the user command table must be ADDED ON TO THE END of any code that you already have for a user command table so that existing user i CAN SAVE NEW COPY OF ATE WITH USER COMMAND TABLE TO HAVE IN PLACE AT POWER UP COMMAND TO LOAD CODE FOR EXE CUTING USER COMMAND SHOULD GO IN STARTUP FILE DISKATE User s Manual defined commands are not lost Thus you should always keep on disk a copy of the file which assembles the user command table 2 After assembly of a new user command table you have to resave ATE 3 You must make sure you have provided a way at power up time for the object code for the routine which will carry out the user defined command to be brought into the memory The easiest way is to include the proper L command in the STARTUP file The third method of interfacing a machine language program to DISKATE has a completely different structure It is possible for DISKATE to serve as a subroutine to be called by a machine language program
75. I O board with a UART and one port serves as a Status port issue an input to the status port to see if input is ready If not return to the calling program with the Z flag off If the proper status bit indicates there is an input present take the input if you want to use it to determine whether or not to interrupt the process but your routine should not loop in case there is no input present The panic detect provided in SYSIO will respond to any key being pressed if it can detect it Because there are timing considerations involved in sampling the input line with the software UART on the Disk Jockey it s best to request a panic stop by hitting the break key on your terminal if it has one since this transmits 9 s continuously 6 In all of the terminal I O routines the only regi ster that need be preserved is the stack pointer If you are using DISKATE with a North Star Disk the protocol above is compatible with that for the North Star DOS I O routines though less restrictive You don t have all of their requirements to preserve the contents of registers Thus for North Star users the jumps in the IO jump table can go directly to the North Star I O vectors When you go to reassemble IO if this is necessary you can give an I command to the DISKATE diskette to find out where in the memory ATE and IO will go The memory address listed in the directory printout gives the address and the size in blocks is also given
76. IN VIRTUAL BASIC VERSION 1 8 990 REM 188 REM BASIC PROGRAM FOR BENCHMARKING CHARACTER HANDLING 200 REM 386 REM PROGRAM WILL ENTER INTO A STRING VARIABLE A 499 REM SEQUENCE OF RANDOM LENGTH INITIAL SEGMENTS OF 459 REM THE ALPHABET TERMINATED BY A CARRIAGE RETURN 460 REM 509 REM ON A SECOND PASS7 REM 888 REM PROGRAM WRITTEN IN VIRTUAL BASIC VERSION 1 6 900 REM THE SEGMENTS WILL BE 606 REM SEPARATED AND PRINTED OUT gt Help This is not the way the text was supposed to look Now the question is how to put it back together In this case we re somewhat lucky because the statement numbers provide an easy way to figure out where things should go If this were an assembler program things might be tougher After staring at this garbaged up text for a minute we can see that the last thing we entered statements 700 8800 and 999 are out of place They should have gone at the end of the file but instead have wound up in the middle of statement 500 Here is one way of correcting the damage We ll use the entry pointer to be able to see just where things are by means of the command Clearly the problem starts on what is now line 568 ae a l gt 300181 79917 500 REM ON A SECOND PASS 700 REM gt X S 990 81 gt He Stee aT 968 REM ig ae gies S X Y eave a care iat 700 REM 809 REM PROGRAN WRITTEN IN eee BASIC VERSION l 998 REM gt lt P gt gt 1 a SMX oak ee Oe ey ee a 188 REM BASIC
77. If ATE follows IO immediately in memory then you must be sure that your reassembly of IO will not result ina module to oe to fit After an assembly of I0 the command IOEND SYSIO will tell you how big IO is If the number of blocks needed to contain it is more than the number listed in your directory printout you may have to change the AORG statement in SYSIO and reassemble so that IO goes some This will not affect ATE since ATE picks up the location of IO by the jump at memory location 3 The I O system in DISKATE also allows you to customize the drivers for the disk In most cases you would not need to do this unless DISKATE were being used with some other kind of hardware but you might want to consider this for certain special applications For instance if you wanted to include on diskettes files which you would not want DISKATE to touch or even 1985 1 0 PROTOCOL COMPATI BLE WITH NORTH STAR DOS WHEN IO REASSEM BLED IF TOO BIG FOR SAME SPOT MUST BE RELO CATED BY CHANGING THE AORG DISKATE User s Manual 196 know about this could be achieved by rewriting the DISKATE disk drivers to only begin reading the disk at a particular block Or to interface DISKATE to someone else s operating system a single very large file could be created for each diskette within this operating system and then the DISKATE disk drivers could call direct access reads and writes to this file through
78. In using this method you must be very careful that the macro DOES NOT RESIDE IN THE SOURCE AREA Since the O command in the macro collapses the source area Using the COM command can be slightly tricky since you can easily get confused about whether you are talk ing to the terminal at the ground level or via a COM command If you give an ESC to the command line being typed in response to a COM command DISKATE expects the line to be retyped On receipt of the line whatever process contained the COM command continues Thus typ ing ESC to a command expected by the COM command DOES NOT return you to the ground level If you want to return to the ground level when DISKATE expects a com mand line for a COM command you can give the command RENT for Reenter This command performs a jump to the DISKATE reentry location The stack will be reinitial ized but all of your variables will keep their old values oS Note that the COM command will accept only a single command and then returns to the process that invoked it There is a similar command ATE which when execu ted CALLS DISKATE RECURSIVELY The ATE command allows you to give DISKATE any number of commands precisely as if you were talking to the ground level DISKATE An error will cause the error message to be printed but you will NOT be returned to the ground level Instead you will still be within the invokation of DISKATE which is a subprocess of whatever process contained the ATE
79. KATE to issue the line with the prompt character Suppose you want the 3rd such group The command n amp m_ 4l 19 GIVES m 1 will give the desired result Remember once again that LINES in a command of the form n m m will give WIDTH STARTING that is the number of lines being viewed and not the WITH THE number of the final line to be viewed Again note that nTH LINE even though we want to see 29 lines we have to use 19 in the second part of the command because 41 matches a whole line If we had said 41 29 this would give 21 lines not 20 Acutally the final in this command is superfluous 41 19 will give exactly the same output Of course with the JOAN DOE file we have been using as our standard example this command would return an error message since there aren t enough lines in the file to match even 41 Suppose you want to see the last line in the cur NEGATIVE rent file but you don t know how many lines in the file OCCUR there are This can be easily achieved using occurren RENCE cing with a NEGATIVE OCCURRENCE NUMBER In an instance NUMBER of occurrencing of the form Npattern if N is negative COUNTS DISKATE will look for the Nth occurrence of pattern BACKWARD COUNTING BACKWARDS FROM THE END OF THE CURRENT FILE FROM END The last line of the current file can be viewed by the OF FILE command olg i At this point we should note that DISKATE con POS Siders occurrence numbe
80. KEEP 9 8 Remount new diskette type S ATE BEGIN END ZKEEP S ATETBL lt T gt 9 Remount DISKATE diskette type L IOTBL lt T gt KEEP 9 19 Remount new diskette type ZKEEP S lt T gt 11 Remount DISKATE diskette type O L STARTUP 12 Remount new diskette type Sena 5 3 rar 4 a amp oea en ana 199 type S IO B SYSIO IOEND type T ATE ATE B T ATETBL ATETBL B type T IOTBL IOTBL B type T STARTUP STARTUP B DISKATE User s Manual 119 At this point your new diskette will have a copy of the DISKATE system in which the patches to the IO jump table and or new Disk Jockey Serial port baud rate will have been installed If you need to reassemble IO to include your own I O routines continue on Be sure to change those entries in the jump table for disk drivers so that there is a STC RET for each non existent drive 13 Mount DISKATE diskette type type O L SYSIO O L SYSIO 14 Edit SYSIO using the DISKATE editor commands to install your I O routines in place of those there Change the AORG statement to relocate IO in case your I O routines make IO large enough to conflict with the space occupied by ATE 15 Type Z A If the assembly has errors go back and correct with the editor 16 Remount new diskette type type S 10 B T 10 lt IOEND SYSIO S IO T 19 lt IOEND SYSIO S IOTBL B lt T gt S IOTBL
81. O 2ND FILE TD COMMAND COPIES ALL FILES ON 1ST DRIVE TO 2ND DRIVE LEAVING ALONE FILES UNIQUE TO 2ND DRIVE DISKATE User s Manual DISKATE will check BEFORE copying the file to see if there is room and if not the TD command will terminate When this happens you may want to give the command I newdrive to see what files exist on this drive that weren t on olddrive and delete some of them with a U command to make more room The memory address portion of the directory for a file can be set to a given address by using the W com mand which stands for Write address The W command has the form W filename argument where the argument is the same kind of DISKATE argument that would be recognized by the editing and monitor com mands The only thing that is changed by the W command is the memory address This is especially useful in connection with the assembler where during assembly the object code is placed one place in the memory but when executed the code is to be loaded at another place We ll look at this in more detail in the next section We conclude this section with a command which is extremely convenient but which can also cause serious grief if used incautiously The GO command which takes only a file name as its argument will load a file from disk and then automatically execute it If the file is a source file Eneng GO file is equivalent to L file D lt R gt whereas if the file is NOT a source file
82. PON PANIC DETECT PROCESS WILL STOP amp WAIT FOR INPUT ESC ABORT S DO ONE CHAR amp STAY IN PANIC STATE OTHER RESUME AS ARG WILL DISKATE User s Manual symbol can be used as an argument and will give the address within the file where the output was halted For instance if you abort a command and then decide you want it to go ahead and finish the command ne will work In general if you ve given no commands since aborting the printout you can use the command 2 gt gt l to resume the printout and have it continue to whatever upper bound you orignally specified This feature is implemented only for the command and will not work if a command is similarly interrupted As we mentioned at the beginning of the manual DISKATE scans its input and if the back space key 98H or Control H is pressed the pointer to the line buffer holding the line currently being typed is backed up one character and the following is echoed to the terminal a back space a space and another back space If you are uSing a video terminal this will erase the last character typed from the screen You can do this any number of times to wipe out Several characters that have been typed that contain mistakes If you type ESC then the entire line is ignored A backslash is printed on the terminal to echo this l The details for creating the actual drivers to han dle these I O calls are disc
83. PROGRAM FOR BENCHMARKING CHARACTER HANDLING 260 REM ek l 39 REM PROGRAM WILL ENTER INTO A STRING VARIABLE A 400 REM SEQUENCE OF RANDOM LENGTH INITIAL SEGMENTS OF 458 REM THE ALPHABET TERMINATED BY A CARRIAGE RETURN DISKATE User s Manual wage 460 REM 5080 REM ON A SECOND PASS THE SEGMENTS WILL BE 608 REM SEPARATED AND PRINTED OUT 700 REM 880 REM PROGRAM WRITTEN IN VIRTUAL BASIC VERSION 1 9 9080 REM Success Let s go over this carefully First a word about the general method The problem is that we have a piece of text in the wrong place and to correct that it has to be moved with the Move command To specify the Move command requires we determine three things the lower address of the interval to be moved the upper address of the interval to be moved and the target location to which we want the interval moved The method to be used here which is a fairly cautious use of DISKATE is to determine each of these separately First the lower address of the interval to be moved We decided that the thing which is out of place is statements 7998 through 900 and statement 799 now begins in the middle of line 599 The first thing we do then is set the entry pointer to the beginning of where statement 709 now is and use the command to make sure we ve got it The value of the pointer is then saved in X which will be the lower address in the interval to be moved The use of the pointer here is not absolutely necessary but settin
84. R FORM IT AT THE END SPACE WITHIN ATE FOR IT AT USRCT ROMEND DISKATE User s Manual table Suppose we want to make this into a user defined command which to abbreviate we ll call PSYM Recall that PRNTSYM was assembled to begin at 7F0 H and that to assemble PRNTSYM properly we had to set the symbol table to ATETBL to get addresses for OUT PHLSB and SYMTB Let s suppose that we ve just assembled PRNTSYM that the source code for it is the current file and that the symbol table has not been zeroed Let s also assume that there are no user defined commands at this point To create the user command table we can give the following commands N E ORG USRCT ASC PSYM DW PRNTSYM DB A Remember that PRNTSYM sets the assembler format to Intel This simple source program will create a user command table with only one entry PSYM Now any time we want to print out the symbol table we can give the command PSYM as if it were a built in DISKATE command However for this to become permanent there are two things that must be done immediately First we have to save a new copy of the module ATE so that when DISKATE is brought up the new user command table will be in place Since the symbols for ATETBL are already in the symbol table we can do this by mounting a diskette with ATE as the second directory entry and typing S ATE BEGIN END Since the user command table is located within the space for the module AT
85. RMIN ATES LOOP WITHOUT ERROR IF ARGUMENT GIVES MATCHING FAILURE DISKATE User s Manual simpler Recall that if a command takes an argument but none is given DISKATE uses the last argument computed In the command line above the argument of the QF command is identical to the argument of the K command so the command can be simplified to R 1 QF SEX 3 K This is an example then of a general form of command which will delete every occurrence of a pattern in a file In general you would use the command R 1 QFpattern K to delete from the current file every occurrence of pattern Suppose we want to REPLACE EVERY OCCURRENCE of one pattern by a piece of text This can be done exact ly the same way remembering that a K command followed by an E command replaces the argument of the K command with the entered text Thus the command R 1 QFpattern K E text will replace every occurrence of pattern with text In word processing systems this operation is often called global search and replace In determining whether a matching failure has taken place for an argument of a QF command even if there are no matching symbols in the argument a matching failure perhaps argument failure is a better term will still occur if the lower address of an interval turns out to be larger than the upper address This can acutally be used for comparison operations We ll see an example of this in I 2 l In effect a command lin
86. RS UPDATED K COMMAND WITH ARGUMENT OUTSIDE SOURCE AREA WILL DISKATE User s Manual useful for machine language programming where you may want to be sure that a program which is still being debugged is only modifying those areas of the memory that it s supposed to By zeroing a large block of memory before you begin work you can easily see with the command whether a program has dumped garbage in the wrong place This method also makes it especially easy to see how much stack is being used The DISKATE commands discussed in this section make DISKATE the most powerful monitor available for personal computers They can be used in macros just as the editing commands discussed in the previous section For instance suppose you wanted to fill a block of memory with repetitions of the pattern consisting of 8 ASCII blanks followed by a carriage return and a line feed Let s say you want to set up 190 such lines beginning at location 18980H The commands N E LINE R89 E N E FILL 1000H B16 R108 D LINE E D A will set up a macro which will do this job and can be invoked by the command D FILL Suppose we want to do the same thing but we want to fill all of the memory available up to 280 H but without overwriting anything beginning at 20090H Of course we could calculate the number of lines that would be needed figuring 82 bytes per line and substi tute that for the 19 in the macro above
87. SPECS U SPECS S SPECS e The O command will collapse the source area so that the combined file will not pick up any extraneous informa tion The I command is given so that we can verify that there will be enough room in the memory to hold the com bined file Note that the O command will put the entry pointer at the end of the empty current file which will be the only file in the source area After L SPECS the entry pointer will be at the end of the material loaded so we can go ahead and give the second L command is given to make sure the file looks correct before we go ahead and make any changes on the disk Now that we know that the file is correct in the memory we FIRST delete SPECS That way there will be no doubt that there will be room on the disk for the combined file pate AT GIVE O COMMAND BEFORE L TO WORK WITH ONLY ONE FILE O L filel L file2 WITH filel file2 SOURCE FILES WILL COMBINE THE FILES INTO ONE FILE AS THE CURRENT FILE IN MEMORY DISKATE User s Manual The S command will save the file and juggle the position of the files on the disk if necessary to make room for Lt Suppose you are working with source files in this way and give an L command but by mistake specify the wrong file so that you want to delete that part of the current file in the source area that you just loaded Of course you could backtrack give an O command and set up the current file all over
88. TE this is handled by two internal 16 bit locations the assembler PROGRAM COUNTER which is denoted by amp and the assembler STORAGE COUNTER which is denoted by The program counter keeps track of the location of each machine instruction in terms of WHERE THE INSTRUCTION WILL BE IN THE MEMORY WHEN EXECUTED The storage coun ter keeps track of WHERE EACH INSTRUCTION IS STORED IN 83 INTEL FORMAT INTE PROS COMMANDS OR PSEUDOS SET FORMAT AS INTEL PRO TECH RESPEC TIVELY amp GIVES PROGRAM COUNTER LOCATION OF INSTRUC TIONS WHEN EXECUTED DISKATE User s Manual THE MEMORY AS THE OUTPUT OF PASS 2 OF THE ASSEMBLER The storage counter is not changed during Pass 1l since no object code is generated during this pass DISKATE needs an initial value for the program counter You may want the default value for amp to be 8 but then again you May not The value of amp at power up can be set by the customization procedure described in Part II You can give it any value within the range of 16 bits You can also set its value using the amp command This command is analogous to the command it sets amp to the value of the argument Likewise the command will set the storage counter to the value of its argument When you customize DISKATE you must be sure that the power up value for is the location of a place in memory where object code can safely go in case you forget to set from withi
89. The string of commands located at the lower address of the interval given as argument is executed Execution of this string is terminated by either an end of the file a or a QS or QF command When execution terminates control is passed back to whatever command string invoked the D com Mand Analogous to subroutine call The initial refer ence interval is the source area regardless of any REF commands G lt Interval Argument gt Goto Execution of commands is transferred to the lower address of the interval given as argument There is no return to the command string invoking the G command Analagous to a GOTO command The initial reference interval is the source area regardless of any REF commands R lt Interval Argument gt Repeat The rest of the command line or file in which the command occurs is repeated the number of times given by the lower address of the argument which is usually an arithmetic DISKATE User s Manual 132 expression The value of the lower address of the argu ment is treated as a 16 bit non negative integer If the araqument is the remainder of the command string contain ing the R command is not executed Execution of an R loop May be terminated by a QS or QF command R loops may be nested If the command string containing the R command is to issue a return on completion the return is issued after the repetition has been terminated lt string gt where commen
90. Those commands which use the whole source area for their initial reference interval i e G D or F are not affected and a missing argument lt and gt also are not affected The best way to see how this works is by an example Suppose you give the command REF 3 Normally the argument by itself would match the current file but after the REF command it would match the line with the pointer Note that the argument of the REF command is kept as a character string and reevaluated each time a command would normally take the current file as its initial reference interval With the command above the initial reference interval changes every time the entry pointer is moved to a new line As with DEF REF with no argument shows the string that is being saved as the argument of the last REF command executed and REF discontinues the use of this feature until the next REF command REF and DEF can be used in combination to work quickly through blocks of text For instance suppose you are editing a letter in which paragraphs are sepa rated by two carriage returns in a row The commands REF DEF ha can be used together to edit paragraph by paragraph You would begin by giving the command lt F gt This sets the pointer at the beginning of the file Now the symbol matches the CURRENT PARAGRAPH To get to the next paragraph simply type carriage return in response to the prompt character Of course if you make
91. Using an S command for an interval smaller than the whole file gives a simple way to BREAK UP A FILE INTO SMALLER PARTS Editing a file in parts has some advan tages There is less chance that you will run out of room in the memory just at a time when you don t want such considerations to disrupt your train of thought in editing the text and DISKATE will run faster with smaller files Suppose you have a file called SPECS for instance and you want to break it into 2 parts the first part being the first 147 lines and the second part being the rest Let s call the two parts SPECS and SPECS Note that is a valid symbol to be part of a file name We ll also assume that the whole thing is in the memory as the current file To break it up into two pieces you can use the commands S SPECS 147 S SPECS 148 In this case we re assuming that the REF command is not being used Note that after these commands used as a file name will denote SPECS You can also save intervals from outside the source area We ll see some examples of this below To examine what is on a disk you can give the I command I for Identify The I command with no argu ment will list the files on the current drive If you want the files on a specific drive listed the drive designator can be given as an argument For each file the listing will tell you the length of the file whether it is a source file or not and if not will give the memory ad
92. VAL _ et the argument Lets go back to our standard example NAME JOAN DOE SEX FEMALE s ADDRESS THE WOODS 0 Suppose we want to make the following changes which DISKATE User s Manual siga both involve only deletions the name JOAN is to be changed to JON and FEMALE to MALE Let s do the second one first What we want to do is delete the FE from FEMALE It happens in this case that the only occur rence of FE in the file is in FEMALE so the command K FE will do what we want But Remember that the argu ment FE specifies the FIRST occurrence of the charac ters F E in the file How do we know for sure that this first occurrence is in the right place Of course in this case the whole file is small enough that we can look and see but even then it is possible to make a mistake The best thing to do is ALWAYS USE A COMMAND TO MAKE SURE THE INTERVAL YOU GIVE IS THE ONE YOU WANT If we were being careful we would have said FE and DISKATE would have responded with SEX FEMALE Then we would have the confidence that the K FE command would do what we intend When you give a K command the only thing that will be printed on the terminal is the prompt character when DISKATE is ready to accept another command unless the argument you give involves match ing and DISKATE cannot find anything that matches in which case an error message will be printed If we re being as cautious as possible perhaps w
93. a disk drive designator At this point we come to another important command the command for loading a file Although it might seem straightforward there are several ways that trouble can occur if not used carefully The L command L for Load will load the file given as the first argument into the memory The memory address where it is loaded can either be given as the second argument or omitted in which case DISKATE will determine the load address in the following way If the file is not a source file then it is loaded at the memory address given by the memory address portion of the directory entry for the file If the file was saved by DISKATE then this address will be the address in memory where the file began at the time it was saved If the file was created by another piece of software you should always give the load address as a second argument to the L command unless the file is of type Otherwise there may be no telling what is writtten in the space in the directory where DISKATE expects the memory address You will get quite a surprise if you give an L command without a load address and the file clobbers part of the resident software If the file to be loaded is a source file and the load address is not given as a second argument then it is INSERTED into the current file between the target character and the preceding character i e excatly where text is inserted by an E or M command The target character will then
94. age return and Tine feed are printed aii f lt A subroutine which obtains a character from the terminal and echoes it via the OUT routine above All flags and registers are preserved except A which contains the input character ESC and backspace are treated like any Other character DISKATE User s Manual 144 PHLSB PHLDC USRCT ROMEND UCTAD ASPC STCTR BOSAP SYMTB END EOSAP BOFP _ EOFP TABA _ CHPTR Pl P2 A subroutine which prints the value in HL in the current base including leading s The value is printed in split form with 3 digits per byte for all bases other than 16 If the base is 16 4 digits are printed A subroutine which prints the value in HL in base 19 suppressing leading O s An address giving the beginning of space provided inside ATE for a user command table An address giving the upper address of space provided inside ATE for a user command table The address of a 16 bit value giving the address of the beginning of the user command table If this is not changed by personalization the contents of UCTAD UCTAD 1 on power up will be the same address as USRCT The address of the assembler program counter The address of the assembler storage counter The address of a 16 bit value giving the lower address of the source area The address of a 16 bit value giving the beginning address of the symbol table The address in memory of the last byte of
95. also ADJUSTS THE SOURCE AREA file boundaries determined place the file entirely within the current source area the Source area stays unchanged If the file boundaries place the file entirely outside of the current source area the source area is redefined to be the file determined by the F command Nothing happens to the old source area in the memory the only thing that is changed is that the internal pointers DISKATE uses to define the limits of the source area are reset If one of the file boundar ies Ties inside the current source area but the other doesn t the source area is expangeg to include the file determined by the F command Let s see how the PF comniana would work with a simple example Let s suppose we re using DISKATE to edit a letter to George Morrow expressing how much we like his memory boards Somewhere in the first few _ lines will be a line containing the pattern George If the siie EMPTY FILE AT TOP OF SOURCE AREA amp MAKES NEW FILE CURRENT F COMMAND SETS CURRENT FILE TO THE FILE CONTAIN ING THE LOWER ADDRESS OF ARG INITIAL REF INTERVAL IS WHOLE SOURCE AREA TARGET CHAR IS END OF FILE SOURCE AREA ADJUSTED TO INCLUDE THE FILE DISKATE User s Manual 42 Now suppose we use the N command to set up a new empty file and in this file we put a string of DISKATE commands used to format the letter We ll see below some examples of such
96. aluate The lower address of the interval given as argument is left in the HL register pair and the upper address in the the DE register pair Nothing is printed on the terminal except for error messages For use only in machine language programs calling the entry point ATECOMS where ATECOMS is a symbol in ATETBL A jump to ATECOMS is written at memory location Assembler Formats Processor Technology Format The label field begins in the first character of each line Ifa statement is not to be given a label the first character of the DISKATE User s Manual 149 line must be blank The label is given without any symbol to ter Minate it A label must contain only upper case letters or digits and must begin with a letter A statement consisting of only a label field assigns the value of the program counter to the label The opcode field is separated from the label field by one or more blanks If the statement takes an operand the operand field is separated from the opcode field by one or more blanks The com ment field is separated from the previous field by one or more blanks The entire line can be designated as a comment by giving as the first character on the line Intel Format Same as above with the following differences a label is termi nated with a colon If a statement is not to be given a label the opcode may begin the line The comment field begins with a semicolon 3 Assembler Pseudo ops AORG
97. an EQU more than once In any assembly labels are defined once and then retain that value throughout the duration of the assembly process unlike variables that may be changed at any time However you CAN attatch several labels to the same location in a program For instance if you had a label called DATA to indicate the beginning of a data section and TABLE was to be the first label in this section both DATA and TABLE should label the same place This can be accomplished by giving an assembler statement with the label DATA as THE ONLY FIELD PRESENT In Intel format this might look like DATA TABLE Most of the other pseudo ops in the DISKATE assem bler are standard for most assemblers To define con stants you can use the pseudo ops DB for Define Byte and DW for Define Word DB creates a one byte constant with the value of the operand and DW creates a 16 bit constant with the value of the operand In addition to numerical expressions you can also enclose an ASCII character in single quote marks Thus to define a one byte constant with the label STAR which is an ASCII asterisk in Intel format you could give the statement STAR DB with the EQU pseu will also allow or DW separated consecutive bytes Operands such as can also be used do and machine instructions DISKATE you to use multiple operands for a DB only by commas Thus to define three of s you could use DB 0 0 9 To de
98. an abbrevi ated version of the symbol table for an assembly of DISKATE and a complete list of all the Symbols in it and their meaning can be found in Part III This table provides the addresses for a number of internal DISKATE routines and registers As mentioned above the table can be made into the current symbol table by the commands L ATETBL lt T gt KEEP 9 Then to zero all of the new symbols defined during the assembly of this program the command Z gt KEEP can be given The addresses for OUT and PHLSB have come from this table It should be emphatically emphasized that in the listing below these addresses refer to the corresponding locations ONLY IN THE VERSION OF DISKATE CURRENTLY BEING USED BY THE AUTHOR OF THIS MANUAL IN YOUR VERSION THEY WILL ALMOST CERTAINLY NOT BE THOSE SHOWN IN THE LISTING In no case should you take addresses for OUT and PHLSB by reading them off from the object code of this listing All of this is more fully discussed in Part II oo The program is set to be executed from location 7FOOH Note that the program sets this initial value for both amp and at once using the ORG pseudo You must always be careful when you do this since the ORG state ment preserves the relative offset of amp to so that in DISKATE User s Manual 93 this case we re assuming no previous assembly gave a different value than amp One way of avoiding this situation is to always give separate AORG and
99. are in PROM or otherwise manage to find their way into the memory without needing to be loaded within IO IO should be reassembled so that the jump table con tains the proper jumps The procedure for connecting a serial device that such as a TTY to the Disk Jockey serial port is a special case that is somewhat simpler In this case the only thing that will have to be patched is the instruction which loads the baud rate constant for the serial I O port Here is what you do 102 DURING FIRST STARTUP I O ROUTINES CAN GO AT 2900H 29FFH AFTER PATCHING JUMP TO 9100H UNLESS OTHER WISE INDICA TED ON CORREC TION SHEET TO RUN SERIAL TERMINAL FROM D J PORT AT DISKATE User s Manual l Obtain the address of the terminal initialization routine from the jump table The low order byte of this address is at I0 19 and the high order byte at IO 11 2 Unless otherwise indicated in a corrections sheet the first byte of the terminal initialization rou tine will be a NOP or The next instruction will be mnemonics octal hex LXI H 41Q G41 841 BOB 21 21 This is the instruction which loads the initial speed constant for the Disk Jockey Serial port Consult the Disk Jockey manual to find the speed constant for the baud rate you want then patch it in in place of the 41 688 octal or the 21 9 hex in this instruction Be sure to patch only the address portion of the instruction and not the
100. as follows 3200H 41FFH 7000H 7FFFH Symbol Table Source Area Object Code Area These settings can be achieved by the following com mands 3200H M lt T gt 0428H amp 7006H amp In order for these commands to be executed automatically at power up all you have to do is make sure they are included in the STARTUP file This file can also include such commands as INTE or PROS Q or J TAB WID Since STARTUP is a source file it is easy to change just type O L STARTUP and then edit it like any other source file then type S lt F gt when you are finished When STARTUP is executed at power up time it may be anywhere on the disk The advantage of concen trating your personalization in the startup file is that it s easy to change and you can initialize your system easily by typing O GO STARTUP Because STARTUP will most likely be a very small file you may find it con venient to include a copy on each diskette gt In some instances this may be inconvenient or lll CAN PERSON ALIZE BY SETTING VALUES IN STARTUP FILE O COMMAND FOR lt S gt amp FOR OBJECT CAN DISKATE User s Manual won t even work at all For instance if you want a drive other than A to be the default drive at power up this cannot be done with a CD command in the STARTUP file since at power up DISKATE will get STARTUP from the default drive which unless you change it will be A The other meth
101. be printed This is the P command which takes as argument any of the kinds of intervals discussed above The P command will format the printout in whichever of the Intel or Processor Technology formats is current If the source code contains INTE or PROS pseudos these will NOT cause the P command to change formats Thus the argument of the P command should not specify an interval that includes code with both formats The P command WILL respond to INTE or PROS commands given from the terminal so if the format is wrong one of these commands can be used to change it Note The P command can saree p used with ordinary text to 99 COMMAND WITH NO ARGUMENT WILL PRINT CURRENT ASSEM BLER TAB SETTINGS ARGUMENT IS LIST OF NUMBERS WHICH REPLACE OLD SETTINGS OMMITTED SETTING KEEPS OLD VALUE P COMMAND PRINTS INTERVAL GIVEN AS ARGUMENT IN ASSEM BLER FORMAT DISKATE User s Manual 9 obtain line numbers for referencing with arguments like n but the lines will come out looking rather peculiar Since they will be formatted as if they were assembler source statements There are several ERROR CODES that the assembler ASSEMBLY will generate if it finds something wrong in your source ERRORS code If an A Al or A2 command results in assembler TERMIN errors the message ATE A COMMAND ASSEMBLY ERRORS STRING will be printed on the terminal when execution of the command is finished and if
102. be sure we had the right JOAN just by seeing the line containing it Here s an approach where we only view a line at a time which will introduce a new DISKATE command that is very important to keep in mind in conjunction with the K command We recall that the command is a simple com Mand with no argument that will show us the line with the target character and the position of the target character in the line What we re going to do is set the entry pointer to successive JOAN s and use the command to view the line with the target character until we get the right JOAN Then we can go ahead and use the K command The first step is to set the entry pointer to the first JOAN in the file This is accomplished with the command JOAN We encountered the symbol before DISKATE prints this symbol between the target character and the preceding character in response to a command When used as a command SETS THE ENTRY POINTER TO THE LOWER ADDRESS IN THE INTERVAL GIVEN AS THE ARGUMENT That is the command sets the target character as the first character in the interval given as the argument In this case JOAN matches the first occurrence of JOAN in the file so the JOAN command sets the entry pointer so that the J in this JOAN becomes the target character Now if we give a command and the target character is the J in the wrong JOAN the command gt JOAN will set the J in the next JOAN to be
103. be using the stack of the machine language program that called it 169 bytes of stack should be sufficient for most applications 3 When control is passed back to the machine language program the carry flag will have been set if an error 118 MACHINE LANGUAGE PROGRAMS CAN CALL DISKATE VIA ENTRY POINT ATECOMS EXECUTES COMMAND STRING WHOSE ADDRESS IS IN HL JUMP TO ATECOMS WRITTEN AT MEMORY LOCATION CAN BE IN VOKED BY RST RETURNS WITH CARRY SET IF ERROR will have the addresses of the argument DISKATE User s Manual has occurred so your program can check for this case Because of the jump to ATECOMS at memory location 8 the RST will automatically cause a call to be vectored to ATECOMS One way for your machine language program to pass information to DISKATE using this method is to set the value of internal DISKATE variables For instance sup pose you want to generate a source file using a machine language program and then operate on it with DISKATE commands The contents of BOFP and EOFP can be set to point to the beginning and ending of the file with the contents of BOSAP and EOSAP being set to point to a at the beginning and a at the end of the file Then when you call DISKATE the current file will be the file generated by your machine language program Information can be passed from DISKATE to a machine language program with this method using a special com mand which
104. bol from ATETBL denotes the address of the last error in a macro or com mand line or the point at which printout from a command was aborted via panic detect The value of is the 16 bit number stored at ERSAV where ERSAV is a symbol from ATETBL denotes the assembler program counter The value of amp is _the 16 bit number stored at ASPC a symbol from ATETBL denotes the assembler storage counter The value of is the 16 bit number stored at STCTR a symbol from ATETBL where exp is a numerical expression denotes the exp th line in the current file exp is interpreted as a posi tive 16 bit number and if this is larger than the number of lines in the file exp denotes the at the end of the file without exp and are the same as l Interval Arguments DISKATE User s Manual 125 An Interval Argument may be specified by a Matching Interval Sym bol a Non Matching Interval Symbol or an expression in which interval symbols are combined by Interval Operations Concatenation Denoted by one interval symbol immediately followed by another Neither interval symbols may contain any non matching interval symbols The interval denoted is the interval matching the concatenation of the patterns specified by the two interval symbols Arithmetic Operations The symbol lt I1 gt lt op gt lt I2 gt where lt Il gt and lt I2 gt are interval symbols and lt op gt is an arithmetic opera tion has lower ad
105. buffer always keeps an extra place for the carriage return at the end of a line In addition to this limit on the maximum length a line can be DISKATE has an internal setting for the width of the terminal s line This setting can be changed by the user at any time We won t discuss this at the moment a full discussion is contained in sec tion I 3 under the WID command Now that we have a basic idea of how text is entered using DISKATE an important but still very basic function is viewing the text to make sure that it is correct There are a number of DISKATE commands which te BACK SPACE TO IGNORE CHAR JUST TYPED ESC TO IGNORE WHOLE LINE CURSOR WON T MOVE LINE TOO LONG TYPE CAR RET OR BACK SPACE CAR RET refer to memory location dddd hex the pair 2066H DISKATE User s Manual serve this purpose The simplest is the command which will QUOTE ONE LINE This command will cause the line containing the target character to be printed on the terminal The symbol this prints as a carot on some terminals and as an upward arrow on older termin als which is used in DISKATE to denote the entry pointer will appear excatly where any text entered with an E command will go That is assuming the line has no invisible control characters the will go between the target character and the preceding character The command has no arguments and is a simple way to be sure the entry po
106. ccession in this case we will SHOW THE PROMPT CHARACTER so that the example listings will look just like they would on your terminal Here is how the current file looks gt aaa 18 REM BASIC PROGRAM FOR BENCHMARKING CHARACTER HANDLING 200 REM 388 REM PROGRAM WILL ENTER INTO A STRING VARIABLE A 490 REM SEQUENCE OF RANDOM LENGTH INITIAL SEGMENTS OF 450 REM THE ALPHABET TERMINATED BY A CARRIAGE RETURN 460 REM 588 REM ON A SECOND PASS THE SEGMENTS WILL BE 608 REM SEPARATED AND PRINTED OUT gt We notice that there is an extra space following the word PASS in statement 599 and we want to get rid of it gt 500131 500 REM SK ges 106 REM 208 REM 300 REM 400 REM 456 REM 460 REM a 50 REM ON A SECOND PASS THE SEGMENTS WILL BE 60 REM SEPARATED AND PRINTED OUT gt 7 le ON A SECOND PASS THE SEGMENTS WILL BE BASIC PROGRAM FOR BENCHMARKING CHARACTER HANDLING PROGRAM WILL ENTER INTO A STRING VARIABLE A SEQUENCE OF RANDOM LENGTH INITIAL SEGMENTS OF THE ALPHABET TERMINATED BY A CARRIAGE RETURN Everything looks good Remember that in a command like K serves as an interval consisting of only one character the target character Forgetting to reset the entry pointer to the end of the file we go blithely 34 EXAMPLE OF USING M COMMAND TO RECOVER FROM POINTER IN THE WRONG PLACE DISKATE User s Manual 35 along entering additional comments gt E 780 REM 899 REM PROGRAM WRITTEN
107. ce area T is a non matching symbol which denotes the first byte beyond the symbol table This symbol can be useful in commands which serve to allocate memory We ll see an example of how this can 7 be used below It is the user s responsibility in DISKATE to see that the symbol table is not overlaid and is located within the memory so that there will be enough space for it THE MOST COMMON SOURCE OF MYSTERIOUS ERRORS USING THE ASSEMBLER COMES FROM A FAILURE TO MAINTAIN THE INTEGRITY OF THE SYMBOL TABLE For instance you can be making an assembly listing and everything is going normally when all of a sudden every symbol is flagged as not having been defined when you can see their definition right in front of your eyes What will have 79 lt ASSEM BLER PASS 1 COMPILES SYMBOL TABLE SYMBOL TABLE FORMAT CHARS OF SYMBOL LAST CHAR HAS HIGH ORDER BIT SET THEN 16 BIT VALUE OF SYMBOL INCLUDES VARIA BLES lt T gt DENOTES SYMBOL TABLE A T DENOTES IST BYTE BEYOND lt T gt UNUSUAL ASSEM BLER ERROS CHECK DISKATE User s Manual happened in this case is that lt T gt is located at the top of RAM and has overflowed off the top No matter where you have the symbol table in the memory when assembling a large program for the first time it is a good idea to have a lt T gt or perhaps several in the middle of the assembly so you can keep track of how large lt T gt is
108. ce on the terminal Arithmetic operations you can use are and with their usual meaning Division will give only the quotient 99 ARGI a ARG2 PERFORMS OPERA TION ON BOTH ADDRES SES OF ARG1 WITH LOWER ADR OF ARG2 can continue in the proper place DISKATE User s Manual discarding the remainder and multiplication will give only the low order 16 bits of the possibly 32 bit product However these operations are carried out left to right with the same priority so that for instance 1 2 3 evaluates as 9 rather than 7 as it would in a typical BASIC As with all of the arithmetic operations Operate on both values of an interval argument Let s see some ways these operations can be used Suppose you want to see the first 1 characters of the last interval computed by DISKATE One way to do it would be by the command 18 This would say print everything up to the 19th charac ter within the last interval computed Another way would be Kee lt St9 This method does not actually call for any matching It says to print the interval from the lower address of the last interval computed through 9 more characters Suppose you wanted to advance the entry pointer by one character The command will do the job Likewise 1 will back it up by one character Many other uses of the arithmetic opera tions will show up when using DISKATE as a monitor which well see in Sect
109. code in the module ATE e Ss The address of a 16 bit value giving the upper address of the source area The address of a 16 bit value giving the lower address of the current file The address of a 16 bit value giving the upper address of the current file The address of a 16 bit value giving the ending the symbol table Not consulted when values are defined for a new symbol a The address of the entry pointer The address of a 16 bit value giving the lower address the last interval computed and the value of lt The address of a 16 bit value giving the upper address DISKATE User s Manual RECAD RECND ERSAV PHD MAXBS DISKBUF RAMEND 145 the last interval computed and the value of gt The address of a 16 bit value giving the lower address the area of memory occupied by the most recently read record from disk The address of a 16 bit value giving the upper address the area of memory occupied by the most recently read record from disk The address of a 16 bit value giving the address of the most recent program execution error or the memory address at which a printout was interrupted i e the value of The address of 1 byte value giving the column in which the print head is waiting A 16 bit constant giving the maximum number of blocks allowed on a disk drive The beginning address of the ATE disk buffer After an I command a copy of the entire directory will begi
110. command can also be used to enter numerical code as well as text This is discussed in section I 2 Note that since parentheses are not used at all in DISKATE there is no problem with entering a string containing unbalanced parentheses i The text you enter with an E command may include ASCII control characters in addition to the carriage return Some care should be taken in doing this however Since such characters may not show on the terminal and could become hidden booby traps When you type a carriage return during text entry only the carriage return character is entered into the file though both a carriage return and line feed are echoed back to the terminal If you are using DISKATE to WONT RESPOND STILL IN TEXT ENTRY ERROR MESSAGE Is ngh CR ONLY BETWEEN LINES IN MEMORY DISKATE User s Manual prepare text to be used by another program you may have to make sure that this program issues a line feed when it encounters a carriage return One way of doing this is to create an edit macro which inserts a line feed into the text immediately after every carriage return We ll see below how to do this If you are entering characters for an E command or any other command for that matter and see a mistake on the current line you can type BACKSPACE to back up the cursor to the place where the error is Each time you type BACKSPACE DISKATE will ignore the last valid character on the line back up the cur
111. controlled drives 1 through 4 drives E F G are vectored to routines for North Star Disk drives l 2 3 and drive H is undefined To specify the disk drive along with the file name the file name is given followed immediately by a colon followed by the drive designator Thus a file named CONTRACT on Disk Jockey drive 2 could be specified CONTRACT B In your version of DISKATE as supplied the current drive on power up will be drive A In Part II we ll see how this can be changed Each file has associated with it the following information 1 The file name 2 The location on the disk where the file resides 3 The length of the file 4 The file type 5 The address in memory from which the file was saved File names we ve already discussed DISKATE assumes that each disk is divided into units called BLOCKS which are 256 bytes long The block is the smallest unit of information for dealing with the disk The actual details of how characters are stored on the disk are irrelevant heres the DISKATE block structure can be thought of as a logical construct which does not depend on the physical details of the acutal disk con troller organization The Disk Jockey controller for example will read and write information in units of 128 bytes The location or address of the file 2 above is the BLOCK NUMBER of the file on the disk The length is the number of blocks Thus the smallest file possi ble in DISKATE is 256 bytes long Th
112. d memory address are also copied The T command works analogously to the S command in that if newfile does not exist it is created An entire disk worth of information can be copied at once with the TD command which stands for Transfer Disk The TD command has the form TD olddrive newdrive where both olddrive and newdrive are drive deisgnators The TD command is more flexible than the usual copy disk command found in floppy disk operating systems Usually such a command will copy the old disk onto the new disk sector for sector completely overwriting whatever was on the new disk The DISKATE TD command WILL COPY TO THE NEW DRIVE ALL FILES FROM THE OLDRIVE LEAVING ALONE ALL FILES ON THE NEW DRIVE THAT DON T EXIST ON THE OLD DRIVE To be specific TD olddrive new drive _ is equivalent to giving T file olddrive file newdrive for every file on olddrive Thus if a file from Olddrive does not exist on newdrive it is created If it does exist the contents from the file on oldrive are copied into it Of course it is possible that if there is already considerable information on newdrive that DISKATE will run out of room before the TD command is finished If this happens a message informing you of this fact will be printed In this case not all of the files from olddrive will have been transferred However there will be no files which were only partially transferred 76 AS 2ND FILE T COMMAND COPIES FIRST FILE T
113. dress lt lower address of I1 gt lt op gt lt lower address of I2 gt upper address lt upper address of Il gt lt op gt lt lower address of I2 gt The symbol lt I gt where lt I gt is an interval symbol expands the upper end of the interval to an entire line Occurrencing The symbol lt exp gt lt I gt where lt exp gt is an arithmetic expression and lt I gt is an interval symbol containing no non matching symbols denotes the exp th occurrence of the pattern given by lt I gt lt exp gt is interpreted asa Signed 15 bit number If the value of lt exp gt is negative lt exp gt lt I gt denotes the ABS exp th occurrence of the pattern given by lt I gt counting backwards from the end of the initial reference interval gt Any number of consecutive dots can be used lt Il gt lt I2 gt denotes the interval from the beginning of Il to the end of I2 If lt I2 gt contains only matching symbols they match the first occurrence of the pattern given which occurs AFTER Il If the end of I2 is lower than Il an error and a matching failure will occur even if lt I1 gt and lt I2 gt contain no matching symbols J 70 I1 gt lt I2 gt causes the initial reference interval for all matching symbols in lt I2 gt to be the interval denoted by I1 gt and as soon as is encountered gt is given the value ee the upper address of Il ne cee ee et ETN l Concatenation gt a Highest 2 A
114. dress The listing will also print the number of blocks of free space on the disk as well as the number of remaining directory entries The address of the file on the disk is not included in the listing from an I command since you don t have to worry about specific disk addresses However if you need all the information that exists in the directory DISKATE keeps a copy of the entire directory in the memory and we ll see in Part II how you can access this if you need to If you are using many S commands without using the file name giving the I command periodically is a good idea to make sure there are no extraneous files on the disk resulting from an erroneous S command You may want to know only how much space is left on a disk without needing a complete listing of all of the My ae I COMMAND LISTS FILES ON DISK DRIVE GIVEN AS ARGUMENT WITH NO ARGUMENT LISTS CURRENT DRIVE FS COMMAND PRINTS AMOUNT OF FREE SPACE ON DRIVE SAME ARG CONVEN DISKATE User s Manual files This can be obtained with the FS command FS for Free Space The argument for this command works the same as for the I command either a disk drive designa tor Or no argument indicating the current drive Of course there has to be a way to delete a file and this is accomplished with the U command U for Un save The U command takes a file name as its argument and deletes the file from the disk Of course the file name can include
115. dress of the interval given as argument If this address contains this becomes the lower bounding of the file otherwise the lower bounding is the first backward in the memory from this address The upper bounding is the first in the memory forward from the lower bounding The entry pointer is set to the upper bounding If either bounding is within the source area the source are is expanded if necessary to include DISKATE User s Manual 131 the new file If the new file is entirely outside the source area the source area consists only of the new file and its bounding s The initial reference interval is lt S gt regardless of any REF commands O lt Interval Argument gt Originate source area The interval given as argument is established as the source area If the lower address of the interval and the upper address are the same an empty source area consist ing only of two consecutive zeros is created at the address Otherwise the forming the lower bound of the source area is written at the lower address of the inter val and the upper bound is written at the upper address The highest possibly empty file in the new source area is made current and the target character becomes the 9 at the end of this file and hence also at the end of the source area O no argument An empty source area is created at the beginning of the current source area D lt Interval Argument gt Do
116. e it current Of course the command Se oc ee antenna DISKATE User s Manual would work just as well At this point a digression is in order to discuss one of the side effects of the F command Above we gave a great deal of attention to the problem of having as close to a foolproof method as possible for setting the entry pointer to the end of the file after correc tions have been made on entered text We also saw an example of how unpleasant it can be to forget to do this We ve also mentioned that the F command will position the entry pointer to the end of the file that it makes current Perhaps this is a clue to how we might be able to have a shorter command which will set the entry pointer to the end of the current file One possibility which is certainly a short command is A F While this will work in many cases it is somewhat dangerous to rely on it habitually The reason is that there is no guarantee that the target character is actually within the current file In all of the exam ples we have seen so far the target character did stay within the current file but in those cases where you are uSing DISKATE as a monitor which is discussed in the next section you will frequently execute commands which place the target character outside of the source area altogether A much safer alternative is the command F lt F gt This command says take the current file and make it current A bit redundant of cour
117. e of the block than can be moved After a move command has been executed the target character will be the same character within the text as it was before but the address of that character will be changed by the inser tion In the example above after 3 the target char acter is the S in SEX and after the move command the target character remains the S in SEX even though this is now in a different place To be specific after an M command the target character is the character after the group of characters inserted by the move 2335 AFTER M COMMAND TARGET CHAR IS 1ST CHAR _ AFTER CHARS INSERTED DISKATE User s Manual Let s look at another example of using the Move command one that hopefully you won t have to use Over and over we ve mentioned that to resume text entry after having made some changes the entry pointer must somehow be restored Several methods for doing this have been shown But Suppose the worst happens and you forget to do this How do you repair the text Typi cally it will be with a Move command and what follows is a realistic example To make it seem natural and easy to understand we ll assume in the example that we re creating and editing a BASIC program even though you may not be using DISKATE to edit BASIC Were just at the stage of writing the body of comments that will go at the beginning of the program Because this example will deal with several DISKATE commands in su
118. e are using the word however too often Let s say the current base is 16 The commands B10 X 0 lt F gt R 1 QF ee however gt X X 1 X B16 will restore the current base after printing on the terminal in split decimal the number of times the word however occurs in the current file Of course not only will you need to know where various things are in the memory you will need to be able to put things where you want them When you bring up DISKATE it will have an initial value for where the source area is to go in the computer s memory Suppose you want the source area somewhere else This can be achieved with the O command which stands for Originate new source area When the O command is given with an argument the interval comprising the argument is estab lished as the source area A is written at the upper and lower address of the interval to form the boundaries of the source area unless the upper and lower addresses are equal In this case two consecutive sS are written at this address to establish an empty source area Nothing is changed in the old source area unless the argument of the O command overlaps it in which case the s written will overwrite something there What is changed are DISKATE s internal pointers to where the source area is When the O command is executed the LAST file in the new source area is made current and the target character is set to the giving the end of the file Of
119. e disk drivers themselves and the only customizing needed as far as the disk is concerned is to maintain in the jump table the proper indication of non existent drives Let s go review now the step by step procedure for getting DISKATE running for the first time l Bootstrap in the DISKATE diskette with a serial terminal at 12 baud connected to the Disk Jockey serial port if this is possible If not boot anyway and wait for the disk to become quiet Even if you intend to run at 1208 baud from the serial port you should make a backup copy of DISKATE so you will need to do some of the steps below 2 Put the patches in the jump table for jumps to your I O routines if needed If you want to run from the serial port at a slow speed patch in the initial baud rate constant The STC RET for non existent drives can also be patched in if needed 3 If you have needed to make patches force the compu ter to jump back to the beginning of IO DISKATE should now respond 4 single drive dual drive Mount a fresh diskette on type E drive B and type I GO DSKINT Mount a fresh diskette GO DSKINT DSKINT will ask which disk is to be initialized s A Remount DISKATE diskette type L IOTBL lt T gt KEEP 0 type L IOTBL lt T gt KEEP A m aa n OO oasann DISKATE User s Manual 6 Remount new diskette type S IO SYSIO IOEND 7 Remount DISKATE diskette type L ATETBL lt T gt
120. e exposure to the 8880 instruction set The DISKATE assembler operates in TWO PASSES In Pass l the assembler does not create any object code Rather it goes through the source code keeping track of where each symbol is defined In the process it compiles a SYMBOL TABLE The symbol table is a string of items containing the symbol and then a 16 bit number giving the value of that symbol At the end of the table is a The symbol itself is recorded in the following format There is no field for the length of a symbol Instead the characters of the symbol occur until the last character which has the high order bit set Thus the symbol AB would be given the hex value 41 C2 In hex normally the character A has the value 41 and B has the value 42 Setting the high order bit for the B gives the value C2 The symbol table will also contain an entry for every DISKATE variable which has been given a value There is no distinction made between a variable given a value with an command and a variable given a value by Pass 1 of the assembler There are several commands which can be used to deal with the symbol table The symbol lt T gt analogous to lt S gt is a DISKATE symbol which can be used in an argument to DENOTE THE SYMBOL TABLE without any match ing For instance if you want to know where the symbol table is in the memory you can give the command lt T gt Just as the symbol S denotes the address of the first byte beyond the sour
121. e length of a file is a one byte number so that DISKATE files are LIMITED TO A LENGTH OF 65K or 256 blocks to be exact DISKATE recognizes only two file types source files and non source files A source file has a type of 9 and a non source file may have any other type The file type is given by a one byte number When DISKATE saves a file on the disk it automatically sets the file type Any file saved FROM WITHIN THE SOURCE AREA will be assumed to be a source file and will be given type while any other file will be given type l 7 All of this information is stored in a special location on the disk called the DIRECTORY The direc tory occupies the first 4 blocks on each disk and each file on the disk has an entry in the directory which is 16 bytes long The format of the directory is 68 DESIG NATED BY LETTER A D DISK JOCKEY E G NORTH STAR BLOCK 256 BYTES Is SMALLEST UNIT OF DISK STORAGE FILE LENGTH LIMITED TO 256 BLOCKS SOURCE FILE IS FILE SAVED FROM SOURCE AREA DIREC TORY USES IST 4 BLOCKS DISKATE Users Manual byte contents 7 file name 8 9 disk address 19 file length 11 ignored 12 file type 13 14 memory address 15 unused North Star Disk users will note that this format is com pletely compatible with North Star DOS In this table bytes are numbered with low order bytes getting low numbers and addresses follow the 8888 convention of low order byte in the low order
122. e memory type the PRINTS command UPPER 7 noes RIS SEA Meas p EAs l l AND 2 lt S gt LOWER ADDRESS Likewise lt F gt will give you the location of the current OF file It is strongly recommended that while you are INTERVAL editing you give such commands from time to time to make GIVEN AS sure the source area is not getting too big There is ARGUMENT no built in protection to prevent the source area from overflowing the actual amount of RAM you have but by giving a lt S gt command from time to time you can find out how big the source area is getting and save your file on the disk before things get out of hand If you deter mine that a file is in danger of getting too big you can Save part of it on the disk kill that part and then go on editing the rest The commands for working with the disk are discussed in the next Section _ The command can also be used to find out the value of a variable Remember that a single address is interpreted by DISKATE as an interval where both addresses are the same So for instance if X is a variable the command 2K will print on the terminal the value of X twice since it interprets X as an interval where both addresses DISKATE User s Manual equal the value of X Note that this can be used to count the number of occurrences of a pattern in a text file Suppose for instance that we are writing a piece of prose and we get the sneaking suspicion that w
123. e only 69 16 BYTES PER FILE THOU SHALT MAKE BACKUPS PLEASE S COMMAND SAVES INTERVAL GIVEN AS SECOND ARGUMENT ON FILE WHOSE NAME IS FIRST ARGUMENT DISKATE User s Manual SPECS was given as the file name DISKATE assumes that SPECS is to be saved on the current drive Before the command above was given presumably there is no such file as SPECS No matter DISKATE will create the file The information saved on the file is the interval which is the second argument of the S command In this case we ve given the entire current file as the argument Because lt F gt is surely within the source area SPECS will automatically be given type 9 which makes it a source file In the memory address section of the directory the current address of the beginning of the current file will be written though for source files this informa tion is generally not consulted Now suppose you do some more work editing the file and want to save the newer version The identical command S SPECS lt F gt will save the current file as the disk file SPECS Note that the old contents of SPECS on the disk are overwritten by the S command So the S command will create a new file if there is no file corresponding to the file name given as the first argument if there is such a file its old contents are overwritten Note that this is a great convenience as long as you haven t made a mistake in typing the S command However it does
124. e ought to take a look after the K command to make sure everything looks the way we want it this is after all the goal of text editing The simplest way to get a quick look is to use the command If we give the command right after the K FE command then DISKATE will print on the terminal SEX MALE gt Let s take a careful look at this Remember that the command prints the line with the target character K putting the symbol between the target character and COMMAND the previous character i e exactly where any text SETS entered with an E command would be inserted It s been TARGET a while since we mentioned the target character and its CHAR cohort the entry pointer The command does not affect 1ST CHAR the entry pointer but the K command will SET THE ENTRY AFTER POINTER After a K command the TARGET CHARACTER WILL INTERVAL BE THE FIRST CHARACTER AFTER THE INTERVAL DELETED DELETED That s why the is in front of MALE in the line shown above This setting of the entry pointer can be both DISKATE User s Manual 17 useful and also cause severe difficulties if not kept in mind The difficulties can arise because it frequently happens that you want to enter some text then make some changes using the K command and then go on to enter more text Becuase the K command resets the entry pointer if you re not careful to put the entry pointer where you want it the text you enter thinking it is going at t
125. e serves the same purpose Sets the current format as the Intel Format Sets the current format as the Processor Technology Format Conditional assembly The operand must be in the form lt expression gt lt label gt If the value of the expression is assembly skips ahead to the label Otherwise assembly continues with the next statement ie ee ee 7 l Assembler Error Codes DISKATE User s Manual en Argument Error The operand field is invalid for the given opcode The A code usually occurs only in Pass 2 though an invalid EQU operand will cause it to occur in Pass l Missing Label An EQU statement occurs without a label Printed during both passes Doubly Defined Label A label is given where the label is identical to a symbol already defined Label Error The first character in a label field is not an upper case letter of the alphabet When this occurs 3 NOPs will be generated in place of the statement Opcode Error The opcode field is not a proper opcode or pseudo op 3 NOPs will be generated in place of the statement Symbols in ATETBL ATETBL is an abbreviation of the symbol table from an assembly of ATE the non I O portion of DISKATE It correlates with the version of DISKATE supplied but in different versions of DISKATE the symbols in ATETBL may not have the same value Any reference to such a symbol should take its value from ATETBL itself and not a listing of any previous vers
126. e terminal I O routines so that all of the termin SUPPLIED al I O from the time the IO command is executed will go AS I O to device n instead of device Also when this com DEVICE mand is executed a carriage return and line feed will NUMBER be printed on the device i e the output routine will be called with the new device number in the A register CR LF first for carriage return and then for line feed ISSUED Here is how this can be used Suppose you have TO DISKATE User s Manual both a video terminal and a printer and your drivers are set up so that when the output routine is given a call to device it prints on the video terminal but device 1 prints on the printer To get a hard copy listing of the current file you could give the command IO Lp sep tO p This would cause the command to send its output to device 1 following which printout is restored to the video terminal by the IO command The IO command applies to both subsequent input and output If your printer does not have a keyboard your input driver should be set up so that it gets the call no matter what the device number is Otherwise if you forget to give an IO command at the end of a command string containing an IO 1l for instance DISKATE will be unable to get any input The IO command can also be given an argument other than a constant The command IO DEVICE would supply the value device number If you don t have of the variable
127. e the command X3 In addition to these entry points DISKATE provides some recursive features whereby it can be invoked from within DISKATE These commands allow a macro to issue a prompt accept a DISKATE command and then resume The command COM for command takes as argument a string enclosed in brackets When this is executed the string is printed on the terminal as a prompt replacing the normal DISKATE prompt You can then type in one com mand line which may include a multi line E command When the command has been executed execution of your macro will resume Here s an example of how this can be used Suppose in loading files you want to have a way to prevent yourself from giving an L command without first having given an O command That way you will avoid clobbering the wrong part of the memory You can do this by having the following macro tucked away in an out of che way piace in the MEMORY 98 IZE IS LOAD ADDRESS OF ATE REENTRY POINT GIVEN BY SYMBOL RENT IN ATETBL IO PUTS JUMP TO LOADER AT MEMORY LOCATION 3 TO RELOAD DISKATE JUMP TO 3 IF IO INTACT COM COMMAND PRINTS STRING GIVEN AS ARGUMENT FOR A PROMPT ACCEPTS ONE COMMAND DISKATE User s Manual O I1 COM TYPE LOAD COMMAND to invoke the macro is to set a variable address of the macro then give the command D variable Note that the macro can be loaded and the variable set to its location by the STARTUP file
128. e to perform the given commands Whenever a command is given DISKATE first scans this user command table to see if there is a command with that name If so it calls the program whose address follows the symbol for the command in the user command table If not it scans the internal DISKATE command table Hence if you have any user defined commands their names will have priority over the DISKATE built in commands For instance if you had a user defined command called PRINT the command PRINT would invoke your command instead of invoking the DISKATE P command and assuming RINT is a variable name When the version supplied of DISKATE is brought up a small amount of space within the code for the module ATE is allocated for a user command table and the user command table will be in this space unless you make a new copy of DISKATE with the user command table Of course to start with there are no user defined commands so the initial contents of this Space is a indicating the end of the user command table This space consists of the interval USRCT ROMEND where USRCT and ROMEND are symbols from _ATETBL Let s see an example of how to create a user defined command In Part I we gave a functional but inelegant program called PRNTSYM to print out the symbol 116 BOFP EOF P USER COMMAND TABLE SAME FORMAT AS SYMBOL TABLE SCANNED BEFORE TABLE OF BUILT IN COMMANDS LISTS COMMAND NAME THEN ADDR OF PROGRAM TO PE
129. e using the R command constitutes a miniature program Becuase it so often happens that you will want to execute a programmed sequence of DISKATE commands DISKATE provides the facility to execute a sequence of commands that are stored in the memory as all or part of a file Our next major goal is to discuss this facility but before we do it s time to present the commands that enable you to use multiple files Often you may only work with a single source file at a time but when entering a sequence of commands which is to be executed by DISKATE more than likely they will want to go into a file separate from the material they operate on As we mentioned at the beginning of the manual a file is a sequence of bytes bounded by s The current file is included in the source area which may contain several files Suppose you want to begin editing a new file while retaining the current file in the source 4g DELETE EVERY OCCUR RENCE OF A PATTERN REPLACE EVERY OCCUR RENCE OF A PATTERN WITH TEXT N COMMAND CREATES NEW DISKATE User s Manual area for use later The command to be used is the N command N for New file which takes no argument This command creates a new empty current file at the end of the source area and expands the source area to include it Any editing you do on this new current file will leave the old file intact That s assuming of course that in editing the new file you don t make a
130. ecomes the character in the source area after the last character of the disk file S lt filename gt lt Interval Argument gt Save disk file The interval given as the second argument is saved as a disk file whose name is the first argument If a file with this name already exists it is replaced by the con tents of the interval If no file exists with this name it is created If enough room for the file exists on the disk but not in one contiguous block the disk will be com pacted and the message COMPACTING ON DRIVE drive designa tor will be printed on the terminal If the interval is within the source area the file type is set at 9 Source file otherwise the file type is set at 1 The memory address portion of the directory will contain the lower address of the interval given as the second argument If the disk or directory is full an error message to that effect will result and the contents of the disk will be unchanged l GO lt fitename gt aa aeeoa anii The disk file whose name is the argument is loaded and executed via an internal D command if it is a source file or vi an internal X command otherwise Equivalent to L lt filename gt D lt R gt for source files or L lt filename gt X lt R gt for non source files The same conventions for loading the file that apply to the L command apply to the GO command 1 lt optional Disk Drive Designator gt Identify The directory on the disk d
131. ed into edit macros if you are going to use them frequently Using a matching argument of the form codes together with the command can be a powerful way to use the DISKATE monitor Suppose for instance there is a jump somewhere between 29 9H and 219 H with the follow ing hex code C3 63 29 56 codes ARGUMENT THAT MATCHES SEQUENCE OF CHARS WITH VALUES MATCHING codes DISKATE User s Manual 57 and we want to patch this so the 93 29 is changed to 7A 38 However we don t remember just where this jump is located Let s assume that the current base is 16 The command 2000H 2100H C3 83 29 will print on the screen the address where the jump is located and then the codes The patch can then be put in and we can see the result with the command lt 1 E 7A 30 Note that the last in the command line above is a command using the default argument One of the main differences between using DISKATE aS a monitor and using it as an editor is that in using it as a monitor you will tend more to use numeric codes and explicit addresses In this regard it is important to be able to know the explicit address for constructs that we specified before by matching and such symbols as lt F gt lt etc The DISKATE command will print on the terminal in the current base the lower and upper addres COMMAND ses of the interval given For instance if you want to know where the source area is in th
132. ence this macro will allow you to do any editing work you want and will automatically handle storing the file on the disk and restoring the entry pointer If properly written such macros can decrease the likelihood of catastrophic errors significantly One final point Although this is discussed in detail in Part II we must note that DISKATE can be called by other machine language programs In addition to the jump to IO written at memory location 3 a jump is written at location to a special DISKATE entry point called ATECOMS This entry point will cause DISKATE to be called as a Subroutine with the HL register pair pointing to a string of DISKATE commands This string will be executed and then control will be returned to the calling program In this case if this string does not cause DISKATE to leave the ground level then a BYE command in the command string will cause control to be returned to the calling program Control will also be returned to the calling program if a indicating the end of the string is encountered Note however that invoking the entry point RENT either by a RENT command or by jumping to the entry point RENT will cause reentry at the ground level and the return address to DISATE s caller will be lost If you are at the ground level and there was no calling program i e you invoked DISKATE by the usual power up procedure then the BYE command will cause DISKATE TO BE RELOADED FROM THE DISK 180 BYE
133. ept in a format consisting of 1 the characters of the symbol with the last character having the high order bit set followed by 2 a 16 bit number giving DISKATE User s Manual 92 the value of the symbol 3 with a at the end of the table It would be useful to have a simple program which will print out a symbol table The example below will print on the terminal the entire current symbol table lt T gt Because the program uses DISKATE s own internal print routines all of the usual conventions regarding the panic state current base etc apply precisely as if this were a normal DISKATE command Some comments about the program First it is quite elementary doing only the minimal in the way of formatting the output and the symbol table is not sorted The symbol is printed then two spaces then the value of the symbol in the current base You could make it a bit more elegant by having the symbol printed out in a given column There are three labels not defined within the program These are OUT which prints one character on the terminal PHLSB which prints on the terminal in the current base the value in the HL register pair and SYMTB which is the address of the pointer to the beginning of the symbol table Where did the addresses come from for these three labels This question is discussed thoroughly in Part II but to jump the gun a little bit your version of DISKATE is provi ded with a file called ATETBL This file is
134. eral variants of an assembler program that share most of their code Last appropriately is the pseudo op END which is used to terminate an assembly The END pseudo is not really necessary since the at the end of the last file in an assembly will also work to terminate the assembly One use for the END pseudo is if you should want to incorporate an assembly source program and one Or more edit macros into the same file Now let s focus on how to invoke and use the assem bler A strong advantage of DISKATE over many other assemblers is that the two passes of the assembler may be invoked separately There are three pertinent assem bler commands in DISKATE The A command for Assemble will invoke both passes of the assembler while the Al command will invoke only Pass 1 and the A2 will invoke only Pass 2 There are two ways for giving arguments for the various A commands IF NO ARGUMENT IS GIVEN THEN THE ARGUMENT IS THE CURRENT FILE The command A all by itself will simply assemble the current file On the other hand you can supply as an argument A LIST OF FILE NAMES SEPARATED BY BLANKS When given such a list the assembler will assemble the program logically equi valent to the concatenation of the files given in the list though the files are not actually concatenated The files in the list can be pieces of one large pro gram or separate subroutines When a list of files is given as the argument the source area is left entirely free
135. erminal such as a TTY that won t run at 1200 baud is as follows Ts 2 won t run at 1209 baud Bootstrap in the diskette anyway The system will not respond but that doesn t matter Stop the machine and using a front panel or moni tor install your own I O routines in the computer s memory The guidelines that these routines must Satisfy are given below A safe place to put them when you first bring up DISKATE is the 256 bytes from 2909H to 29FFH Once you have DISKATE running you can use it to incorporate your I O routines within IO If your I O routines won t fit in 256 bytes use the top of the memory available Again using a front panel or monitor change the jumps in the jump table to jump to your I O rou tines Be sure to remember that the addresses of the jumps must be stored with the low order byte first Cause your system to begin executing instructions at the beginning of IO This should be 919 H 981 9090Q unless your version of DISKATE comes with a corrections sheet indicating a different starting address The module ATE will be reloaded from the disk and at this point a start up message should appear on the screen and then the prompt symbol gt Once DISKATE is running the source code for IO is contained in the disk file SYSIO This should be reassembled to contain your own I O routines so that they will automatically be loaded when DISKATE is bootstrapped If your I O routines
136. eturn with an empty command line the string given as the argument to the DEF command will be executed as a default command For instance above we had examples of using the same command repetitively to see successive intervals of text such as gt JOAN 2 By giving the command DEF gt JOAN 2 whenever you typed a carriage return in response to the prompt character DISKATE would execute the command gt JOAN 2 If you don t remember what the default command is typing DEF with no argument will show it to you If you have given a default command and then want to stop using this feature the command we omo o iF DEFT DEF with an empty argument in brackets will discontin ue using the default command feature until another DEF command is given A similar command to DEF is REF This command also takes as argument a character string up to 24 characters in length enclosed in brackets This string is treated as a DISKATE argument and is evaluated prior to any 59 PAUSE COMMAND STOPS PROCESS IN PANIC DETECT STATE DEF cmd CAUSES cmd TO BE EXECUTED FOR CAR RET ONLY AS COMMAND DEF WITHOUT ARG SHOWS cmd DEF DISCON TINUES USE OF cmd REF arg CAUSES INITIAL DISKATE User s Manual subsequent command which normally takes the current file as the initial reference interval This argument then becomes the initial reference interval for every such command
137. fine a block of storage where you don t need to give the bytes in the block an initial value you can use the pseudo DS for Define Storage It will advance amp and by the value of the operand creating the area as a block of free storage For instance to define a block called BUFFER 88 bytes long in Processor Tech nology format you could use the statement BUFFER DS 80 86 FOR MULTIPLE LABELS AT ONE LOCATION GIVE STATE MENTS WITH NAME FIELD ONLY DB PSEUDO DEFINES BYTE DW DEFINES WORD 16 BITS WITH VALUE OF OPERAND MULTIPLE OPERANDS ALLOWED DS DEFINES STORAGE BLOCK WITH LENGTH GIVEN BY OPERAND DISKATE User s Manual In addition DISKATE has a statement for defining a field of ASCII constants which is not standard The pseudo ASC will define the string of bytes given by the operand The string is NOT delimited by quotes but by blanks The string may also be terminated by the carriage return at the end of the statement line For instance to define a constant which is the ASCII string ERROR you could use the statement ASC ERROR There are two special rules to keep in mind with ASC First if mnemonic ASC is suffixed immediately no intervening blank with a non alphanumeric character then ANY PLACE THAT CHARACTER APPEARS IN THE OPERAND IT WILL BE REPLACED BY A BLANK IN THE STRING CONSTANT GENERATED Thus the statement ASC A B C defines a 5 byte constant whose
138. g as it fits into the area of memory allocated for the source area Brackets are allowed within the text but they must be balanced For instance E Computers of course are complicated is a perfectly valid DISKATE command This entire string of characters will be entered into the file including the brackets Now you ask what will happen if brackets are not balanced Trouble dear user If you type E Computers of thinking that this will enter the string Computers of then the DISKATE command processor will never have received the closing right bracket which ends text entry The result will be that you may think you are typing commands when in fact what you type is still being entered into the text This will garbage up your file with what should have gone as commands When you are typing in a command check to make sure that the first character on the line is the prompt character gt if not you are probably still in text entry The opposite problem too many right brackets will cause the text entry to be terminated prematurely DISKATE will interpret the rest of the text as a com mand which probably won t make sense When DISKATE gets a command that it can t make sense out of or an error occurs it will come back with a question Mark What if a piece of editing requires that you insert a string of characters with unbalanced brackets There is a way to do this which for the moment well post pone the E
139. g the pointer and then viewing its location with the command is so convenient as a way of seeing what we re doing that it s worth the extra bother Remember that the command sets the entry pointer to the LOWER address of the interval which is its argument The next step is to set the upper address of the interval to be moved which we ll do by setting the pointer verifying that it s in the right place and then saving it in the variable Y Right here is another important possible source of error The command as well as a command of the form variable argument will take the LOWER address of the argument What we want here is the UPPER address of the interval to be moved That s why we had to give the command 980 gt instead of simply 980 9890 gt sets the entry pointer to the upper address of the interval 9696 g Finally we set the entry pointer to the place where we want the interval moved which is the end of the file The command for doing this we ve already seen As a final precaution the command X Y will show us the interval we are about to move so that we can have one last check that it is specified correctly At last we are ready for the Move command itself which is Simply MX Y since we have set everything else up DISKATE User s Manual 37 This may seem like a lot of trouble to have gone to In fact the whole thing could have been accomplished by the single command lt F gt gt
140. getting to make sure it won t overflow the area it is safe for it to occupy Every symbol will be retained in the symbol table until a command is given to remove it The Z command given with no argument will Zero i e remove from the symbol table all symbols except those symbols that are always initially present in the symbol table on power up As mentioned in I 1 these are A B C D E H L M SP PSW when the Z command is given with no argument DISKATE rewrites a fresh copy of the initial symbol table so that if the variables had inadvertantly been given the wrong values the Z command will restore the initial symbol table to the correct values If you assemble a program see some mistakes then assemble it again without having zeroed the symbol table you will get an enormous number of assembler errors for having defined symbols twice If you have symbols of your own that you want to keep in the symbol table such as the addresses of macros for instance you can use the Z gt command This command takes a symbol from the symbol table as its argument and will ZERO ALL SYMBOLS DEFINED SUBSEQUENTLY TO THE ARGUMENT SYMBOL Suppose for instance you have defined all of the symbols that you will be wanting to keep You can give the command KEEP 9 to define the variable KEEP as a marker for those varia bles that will need to be zeroed for each assembly Then to zero all of the symbols except those you want to keep you ca
141. give a K command and then IMMEDIATELY afterward realize that it was a mistake GIVING THE M COMMAND WITH NO ARGUMENT JUST AFTER THE K COMMAND WILL RESTORE THE DELETED TEXT Suppose you delete an interval with a K command and then decide you want the deleted text back but only after you ve already given a command after the K com Mand say a The text comprising the interval will still be just outside the source area unless the com mand s you gave after the K command entered text The deleted interval can be moved back to where it should go if a way can be found to specify the right interval The first thing to do is to SAVE THE ENTRY POINTER by giving a variable command Unless you have given a command after the K command which changes the entry pointer it will point exactly to the place where the text is to be put back Suppose you know that you deleted 3 lines In this case the interval of the deleted text can be specified by the argument 1l 3_ S is a symbol which we haven t seen before and it denotes the address of the FIRST BYTE BEYOND THE SOURCE AREA What about the 1 The l occurs in a position where DISKATE expects an interval argument of some kind which means it is looking for something it can turn into an address or actually a pair of addresses to be correct Remember that addresses are treated as 16 bit non negative numbers This means that the sign bit in 1l is treated as a digit bit so l is the same
142. gt gt 2 is one that is difficult to remem ber and it would be easy to make a mistake in typing it It would be better if we could give the macro a name and then execute it by calling it by name DISKATE has just such a provision To name a macro you use the com mand The command does nothing and is ignored This means that the characters following the in a command Let s see how this would work The can be used as a label Let s say we want to give our macro the name IRI es aa Orig saat ty looked P oe Rl OF T R I l K E initial reference interval 45 D COMMAND EXECUTES STRING OF DISKATE COMMANDS STARTING WITH LOWER ADDRESS OF ARG INITIAL REF INTERVAL IS SOURCE AREA COMMAND DOES NOTHING IS l IGNORED CAN BE USED AS LABEL TO NAME MACROS DISKATE User s Manual 46 To add the name to the macro it should look like IRI R 1 QF I R I K E initial reference interval When the macro is executed the command IRI will simply be ignored Now the question is how do we execute the macro by name Remember that the initial reference interval for the D command is the whole source area The command COMMAND D name D IRI EXECUTES will take as its argument the beginning address of the MACRO first occurrence within the whole source area of the WHICH characters IRI Assuming the text file does not BEGINS contain such a pattern since the initial reference WITH
143. hange is made and then restore it after making the change when we go back to entering text we will be entering it in the wrong place How do we get the entry pointer back to where it was before we made the change As usual there are several methods Perhaps the safest method is to save the value of the entry pointer and then restore it This can be done using a DISKATE feature we havent seen before variables A DISKATE VARIABLE is named by a string of characters beginning with a letter and containing only upper case letters and digits Variable names can be of any length When you bring up DISKATE certain variables used by the assembler will already have values and should not be changed unless you are certain you will not be using the assembler These variables are A B C D E H L M SP PSW A variable can be assigned a value using the command which has the form _ variable argument In this case tere MUST NOT BE K BLANK between the sign and the variable name since if there were DISKATE would interpret the variable name as the name of a command There can however be blanks between the sign and the argument if you like The argument can be any DISKATE argument and the variable is given the value of the LOWER ADDRESS of the interval specified by the argument So if you are going to make some changes in text you have just entered with the E command and want to save the entry pointer you can use the command
144. he current source file of the characters N A M E up to the first occurrecne AFTER the N A M E of the characters A D D R E S S so that the command NAME ADDRESS would print on the terminal JOAN DOE NAME OCCUPATION DEER SEX FEMALE wt ADDRESS Remember that in using in this way if two patterns are combined as in patternl pattern2 DISKATE will look for the first occurrence of pattern2 after patternl to find the upper limit of the interval For instance if the current file is the same as the example above the command TAL Al would print on the terminal Fa eee ee In many editing applications it is useful to view the text in units of entire lines There are several ways to accomplish this The character _ which prints as an underscore on most terminals and as a left pointing arrow on older terminals is a matching argument which will MATCH A CARRIAGE RETURN For example if we are using the same current file as in the examples above the command MISEN is will PrARE on the terminal Se ag PATTERN1 PATTERN2 GIVES FROM 1ST PATTERNIL TO 1ST PATTERN2 AFTER THE PATTERNI _ UNDER SCORE OR LEFT POINTING ARROW MATCHES wu REE DISKATE User s Manual SEX FEMALE The matching in this case works just the way it did in the previous uses of to combine intervals The argument SEX will match the first occurrence of the characters S E X in the current file The argu
145. he end of the file will actually be going into the middle The reason DISKATE resets the pointer how ever is that a K command followed by an E command will REPLACE the deleted interval by the text given with the E command which is extremely useful These two consi derations are very important but in order to see some more examples of using the K command we ll postpone a detailed discussion Now let s do the second change we had planned changing JOAN to JON In this case what we need to do is delete the A from JOAN Before getting involved let s give a command and see what the file looks like In response to DISKATE will print NAME JOAN DOE OCCUPATION DEER SEX MALE ADDRESS THE WOODS Right away we can see that to simply say K A would be trouble The first occurrence of A in the file is in the word NAME so K A would delete the A from NAME which is not at all what we want One approach would be to try to count the number of A s in the file before the one we want so we could specify the right one by using occurrencing In this case the A in JOAN is the second A in the file so the command K2 A will in fact do what we want But this method has real ifficulties First of all we can hardly find out using a 2 A command whether we ve got the right occurrence of A since 2 A will only print A not telling us WHICH A it s printing One way to remedy this would be with the command
146. he symbol table is i aaa but no object code is generated A2 lt list of file nanets A2 no argument Identical to the A command but invokes only Pass 2 of the eee Q no argument Quiet o Assembler listings are suppressed execpt for assembler error messages until the next J command is encountered i J a argument le Jabber The assembler will resume printing listings DISKATE User s Manual 138 amp lt Interval Argument gt set assembler program counter The assembler program counter is assigned the value of the lower address of the interval given as argument lt Interval Argument gt set assembler storage counter The assembler storage counter is assigned the value of the lower address of the interval given as argument Z no argument Zero the Symbol Table All symbols are removed from the symbol table except the initial symbols used by the assembler see above under Symbol Table Z lt symbol gt The symbol either a variable or assembler label is removed from the symbol table and the symbol table is compacted Z gt lt symbol gt All symbols defined after the symbol given as the argument are removed from the symbol table The symbol given as the argument is left intact INTE no argument use Intel Format All subsequent P commands and Assembler commands will use the Intel format until a PROS is encountered PROS no argument use Processor Technology Format
147. hich are related to the process of assembling Normally the assembler will produce a listing on your terminal of the source and object codes during Pass 2 In many cases however you will want to assemble a program without producing the listing This is accomplished using the Q command Q for Quiet The Q command which takes no argument will cause the assembler to henceforth refrain from producing a listing during Pass 2 However during both passes if the assembler was given a file list for its argument it will print on the terminal the name of the file currently being processed Assembler error messages are always printed out regardless of any Q command The assembler will also print the value of amp in the current base on encountering a pseudo INTE or PROS To cause the assembler to resume printing listings give the J 89 SORG TT 19 PUTS OBJECT CODE JUST BEYOND lt T gt Q COMMAND CAUSES ASSEM BLER TO QUIT MAKING LISTINGS DURING PASS 1 J COMMAND CAUSES LISTINGS TO RESUME TAB DISKATE User s Manual command J for Jabber also taking no argument When listings are printed during Pass 2 the source code is printed with a tab setting for l The label field 2 The opcode field 3 The operand field 4 The comment field Line numbers will be printed at the beginning of the field for each source statement The normal tab settings for these four fields are repectively 8 15 28
148. his form cannot extend accross more than one line CAN USE You can also use concatenation with the E command CONCAT For instance the command ENATION WITH E E TABLE _ COMMAND would enter first a carriage return then the characters T A B L E and then another carriage return If is a matching argument theh how should we interpret the command PRINTS Literally it would mean print anything on the ter ENTIRE minal Remember though that anything is restricted in CURRENT scope to the current file Thus the command will FILE PRINT THE ENTIRE CURRENT FILE ON THE TERMINAL Suppose we want to see everything after the 35th line The command 35 will do what we want Acutally to be completely pre cise this is not quite correct since this will give us the 35th carriage return followed by anything Seeing this extra carriage return at the beginning will not do us any harm of course but an argument error like this can be very serious with other commands To get exactly what we want we should really use the command 36_ ait Since the interval consisting of everything after the 35th line is the same as everything starting with the 36th line It is a frequent but inescapable nuisance in many areas of programming to worry about whether something should be n or n l Similarly to get the first 5 lines of the current file though the command 4 would work a simpler command is eee 4 Yan
149. his state until the BYE command is given at which time execution of the process that invoked the ATE command continues BYE no argument 7 __ bye bye return to to DISKATE caller DISKATE returns to its caller If it was called internal ly by an ATE command the process invoking the ATE command resumes If it was called by a machine language program execution of that program resumes If DISKATE was invoked only by the power up procedure the BYE command will cause DISKATE to be reloaded from the disk RENT no argument Reenter _DISKATE is reentered at the ground level at the same DISKATE User s Manual 134 entry point given by RENT where RENT is a symbol in ATETBL The stack is reinitialized but all other varia bles and pointers are left intact All processes active at the time the RENT command is given are terminated including processes invoking ATE commands If a machine language program has called DISKATE the RENT command will cause the return address to be lost Y no argument Wipe Calls the terminal initialization routine The vector to this routine is located in the vector table which begins the module IO at IO0 9 On power up this routine is called by an internal Y command WID lt arithmetic expression gt terminal Width The width for the terminal is set to the low order byte of the value of the argument If the terminal produces a new line automatically on line overflow a width equal t
150. hould be sure that the power up version of ATE you create allocates enough memory in the symbol table to hold ATETBL You can find the size of ATETBL by giving an I command for a diskette containing it ATETBL is not particularly large so this should be no problem Don t make the mistake though of thinking you don t need any space at all for the symbol table even if you want to prepare a version only for editing prose text If you want to make a copy of DISKATE on a new diskette with new settings you can follow the same type of procedure already given to make a backup of DISKATE when you first get up and running Rather than review ing all of the details of this procedure again there are a few important points to note The minimum steps needed to create a new copy Of DISKATE are l The new diskette must be initialized 2 A copy of IO must be saved as the first entry in the directory This is not necessary if your system will be bootstrapped from a North Star Disk 3 A copy of ATE must be saved as the second entry in the directory 4 A copy of STARTUP must be saved Whenever you reassemble IO you should also update IOTBL so that IOTBL is the symbol table for an assembly of the same code as IO In this way IO can be saved by set ting the symbol table to IOTBL and typing _ S 10 SYSIO IOEND ee a Likewise ATE can be saved by pia the symbol table 0 ATETBL and Sere E ae tse 0 g BEGIN END LI o gt
151. icularly BEGIN and END The interval BEGIN END where BEGIN and END are symbols from ATETBL is the interval in memory occupied by the code for the module ATE Thus to save a new copy of ATE here is the procedure l Give the command L ATETBL lt T gt KEEP 9 This will establish ATETBL as the eee table with KEEP at the SPOS a 2 Change whatever settings you want to personalize 3 Give the command S ATE BEGIN END Remember again that if the copy of ATE is being cre ated on a new diskette ATE must be the second directory entry When a file is saved which does not exist in the directory DISKATE creates a directory entry for it in the first empty slot in the directory The I command lists the files in the order in which they occur in the 112 PERSON ALIZE BY MAKING SETTINGS THEN SAVING ATE THEN AT POWER UP ALL SETTINGS WILL BE AS THEY WERE AT THE TIME ATE WAS SAVED EXCEPT AS CHANGED STARTUP PROCE DURE FOR SAVING NEW COPY OF ATE ATE MUST BE 2ND DIREC TORY ENTRY DISKATE User s Manual directory If you are creating a new copy of your sys tem on an initialized diskette if ATE is the second file saved it will be the Second directory entry If another file is already the second directory entry you can reclaim the second directory entry by making a copy of the file under another name with the T command unsaving the file under the orginal name and then renaming the copy Also you s
152. ield The format in this example of course is Intel By way of contrast here is how the original source code looks INTE PROGRAM TO PRINT SYMBOL TABLE RG 7F0GH e jae O se e e we e RNTSYM LHLD SYMTB SET POINTER DISKATE User s Manual 95 SYMBEGIN MOV A M 3 GET SYMBOL ORA A SET FLAGS JZ SYMEND QUIT IF SYMLOOP MOV A M GET SYMBOL LETTER MOV B A SAVE COPY ANI 8 H CHECK HIGH ORDER BIT JNZ ENDSYM JUMP IF ON MOV A B RESTORE LETTER CALL OUT PRINT LETTER INX H INCREMENT POINTER JMP SYMLOOP GET NEXT LETTER r ENDSYM MOV A B RESTORE LETTER ANI 7FH STRIP OFF BIT CALL OUT PRINT OUT LETTER MVI A 3 PRINT OUT A CALL OUT SPACE CALL OUT AND ANOTHER INX H INCREMENT POINTER MOV E M LOW ORDER SYMBOL VALUE INX H INCREMENT POINTER MOV D M HIGH ORDER INX H INCREMENT POINTER PUSH H SAVE POINTER XCHG HL CONTAINS SYMBOL VALUE CALL PHLSB PRINT VALUE MVI A 13 CARRIAGE RETURN CALL OUT PRINT CR LF POP H RESTORE POINTER JMP SYMBEGIN GET NEXT SYMBOL STROND RET T RETURN TO DISKARTE Obviously the assembler listing is far more intel ligible whereas the compressed format of the original source code saves on storage The listing generated by the P command is identical to that given by the assem bler except that it includes no object code Of course it can happen that a line which is too long in the assembler
153. ight may seem confusing so we ll look at it carefully to see why it works When assembling by giving a list of disk files there is no need for a source area at all However you will need 1 room for the symbol table and 2 room for the object code How do you allocate this division One method is to let the assem bler do it for you by giving the following pseudo op in your source program SORG T 18 This statement says to put the object code 18 bytes past the end of the symbol table The 18 byte cushion is just to play it safe But as your program is assem bled during Pass 1 the size of the symbol table is expanded considerably How does this statement not result in having the symbol table clobber the beginning of the object code as more and more symbols are defined The answer comes from an understanding of the separation of function of the two passes of the assembler All symbols are defined during Pass 1 but no code is gener ated and the storage counter is not changed The SORG statement DOES NOT TAKE EFFECT UNTIL PASS 2 Thus by this time the symbol table has been completely compiled so the object code can safely be placed just beyond it If you use this method though you must save the object code on disk before going on to assemble something else A subsequent assembly definitely will cause the symbol table to begin overwriting the area of object code for the first assembly oo nee te tn ae There are a number of commands w
154. ightly odd since if there are multiple spaces surrounding the first three words on a line the multiple spaces will be reduced to only a single space so if the first line of a paragraph is indented for instance this indentation will be sup pressed However giving this command should allow you to view prose text with line numbers conveniently It may seem that we ve been spending an undue amount of time with the command DISKATE has many more commands than just the ones we ve discussed so far The reason for spending so much time with the command is to develop the concepts of interval matching and many of the ways that intervals can be specified by using for illustration a command which simply shows us the interval in question Although we are not finished with seeing new operations by which you can combine intervals perhaps now s the time to begin introducing some more commands Before doing this I should emphasize as strongly as possible that if you are ever in doubt as to just what interval an argument actually specifies it s always best to give DISKATE a command and see So far the DISKATE commands we ve discussed have K been concerned with two kinds of operations entering COMMAND text and viewing text already entered Another impor tant major operation involved in editing is DELETING KILLS text In DISKATE this is accomplished by the K com DELETES mand which will KILL or delete the interval given as INTER
155. ine feed but you should be careful that it doesn t cause an error You may find it more convenient to initialize all devices at one time 5 The panic detect routine should inspect your key board for whatever condition you want to use to indicate that a process should be interrupted This can be a special character having been pressed such as control C or any key having been pressed The panic routine should return with the Z flag on if it determines a panic stop and with the Z flag off otherwise The device number is in the A register If you have a printer and a video terminal and the printer is a TTY or Decwriter or the like that has a keyboard of its own in addition to the keyboard on your video terminal you will have to decide whether to have the panic routine consult the keyboard for the device in the A register or always consult your terminal keyboard Note that the panic detect routine could also consult a front panel sense switch if you have one but this is usually less convenient than inspecting the keyboard The panic detect routine MUST NOT WAIT for a key board input but should simply inspect to see if a panic is being requested For instance if you have a serial 194 CHAR IN B REGISTER CHAR IN PUTS CHAR IN A REGISTER PANIC DETECT RETURNS WITH Z FLAG ON IF PANIC OTHER WISE OFF SHOULD NOT WAIT FOR INPUT _ where else in memory DISKATE User s Manual terminal being run from an
156. ing any work with DISKATE but your work will be more pleasurable if you do Although DISKATE may seem complicated it is one of the most powerful pieces of soft ware in the personal computer marketplace and this complexity is a reasonable price to pay for the power you get TABLE OF CONTENTS Part I Usage LE E E E EE E E E E O O E O O O O O O E E O E E E E E O E eeee7 s e eestet amp lh igs tater soca I l Using DISKATE as a text editor eer reer ea ee eee I 2 Using DISKATE aS a monitor 0 EER ge ne eee ee 53 I3 Using DISKATE inp t ou utp t crisser oie i enra eataa 04 I 4 Using DISKATE as an assembler s scscesecesosssoossenoo 79 Dest Iny king DISKATE seseekirerts trine DEEL KANNE aiae 97 Part II Installation and Maintenance Tere eee ree eee II l Bringing up DISKATE sectoare Wad eta ew a4 Reese wR ee E etiees 181 II 2 Personalizing DISKATE Settings cies Scene Va eos ee a II 3 Interfacing DISKATE to Machine Language Programs 114 Part III System Reference Summary ccecccccccecccceeveeetal Index eececcesceccecocoececcecsecooessosesoeseesesoeooeceoseeseee l46 DISKATE User s Manual Part I USAGE I l Using DISKATE as a Text Editor A text editor is one of the most crucial pieces of microcomputer software It is through the text editor that you will prepare and maintain a good deal if not all of the material your computer will deal with DISKATE ATE stands for A Te
157. inter is where you want it to be It is highly recommended that you use this command often while you are first getting the hang of DISKATE It will show you excatly where the changes you are about to make will go to the nearest line at least To use the example from above if the line with the target character is My computer is well fed with the target character being w giving the command will print the line My computer is well fed Obviously we ll need to be able to see more than just the line with the target character The command is a more general command that will let us view practic ally anything The argument of the command is an INTERVAL in the computer s memory This concept of interval is probably the most important one in DISKATE and so we ll spend some time with it The underlying idea of an interval is extremely simple an interval is a pair of addresses in the computer s memory In order to be valid for use by DISKATE commands the first address of the pair should not be larger than the second address For instance if we use the notation ddddH to is a valid interval though it is not yet in a form that DISKATE can recognize It refers to all of the charac ters in the memory from 2900H through 2A98 H inclusive This is why the first address must be less than or equal to the second address to refer to all of the Characters from 2A88H to 2080H would not make sense To return to the command this c
158. ion I 2 Let s return to the problem of restoring the entry pointer after changes have been made so that text entry There is one problem with the method of saving the value of with a variable and then restoring it What if you have made some changes with a K command say but suddenly don t remem ber whether you saved the value of or not It would be useful to be able to put the entry pointer in the right place regardless of what you did before Fortun ately it s easy to specify where this place will usually be If you want to go on entering new text after changes have been made in what has already been entered the new text should go at the END of the current file Because the E command places text between the target character and the preceding character to add text to the end of the file you want the target character to be THE CHARACTER AFTER THE LAST CHARACTER IN THE FILE How 3G ARITH OPS EVALU ATED LEFT TO RIGHT SAME PRIORITY A A n ADVANCES ENTRY POINTER n CHARS TO ADD TEXT TO THE END OF THE FILE TARGET CHAR MUST BE CHAR AFTER LAST CHAR IN THE FILE DISKATE User s Manual can we specify this The argument specifies the current file The argument gt specifies the upper address of the current file since the causes a new value to be assigned to gt The address we want is one higher than gt so the command gt 1 will put the entry pointer so that any text
159. ion of ATETBL BEGIN RENT _ATECOMS The beginning of ATE This location is an entry point which performs initialization The stack and some internal variables are initialized The command 0 2Z Y is executed a message is printed on the terminal and the command GO STARTUP is executed The entry point to ATE Ue avoids initialization Only the stack is initialized If DISKATE has been called by _ another machine language program jumping to this loca tion will cause the return address to be lost An entry point which executes an arbitrary string of DISKATE commands and then returns HL must contain the address of the beginning of the command string A jump to ATECOMS is written at memory location 9 when DISKATE is loaded from the disk ATECOMS may be invoked by loading the address of a command string in HL and then DISKATE User s Manual S143 WHAT READER VCHK CVALS VALUS OUT INECO executing the instruction RST The entry point to the DISKATE error exit A machine language program which detects an error can return to DISKATE by jumping to this entry point A subroutine which reads one line from the terminal DE must point to a buffer which begins with a string ter minated by a byte This string will be printed on the terminal as a prompt The characters are stored in the buffer beginning immediately after the at the end of the prompt The byte in the B register must be 2 greater than
160. ist in your system This way any access to drives that DISKATE User s Manual don t exist will properly signal an error Be sure that when you do this the name fields are preserved If you are running DISKATE with only North Star disks then the loader part of SYSIO must also be changed so that all of the calls to DRIVEA go to DRIVEE The protocol for the disk drivers is as follows When the disk drivers are called the B reister contains a code giving the disk command The codes are B 3 The disk drive should be initialized This is given when a disk is first accessed after power up and also after a disk error If initializa tion is not required this should simply return Note that this command is given to initialize the disk DRIVE and or drivers rather than to initial ize a diskette For the Disk Jockey diskettes are initialized by invoking the program DISKINT Nothing should actually be written on the diskette when this command is executed Usually the routine executing command will want to be sure it knows where the head of the disk is This can be achieved by moving the head to track B 2 The size of the disk in blocks is to be returned in BC For the Disk Jockey this is 1000 The drivers supplied for the Disk Jockey leave the bootstrap record for controller routines untouched This record is not included in the 1908 blocks and as far as DISKATE is concerned is not even there For the North Star disk this n
161. ive you a significant surprise What we would like is a method of setting the current file to be the next file in the source area but which would give an error message if we happened to already be at the last file Here is a COMMAND command that will do the trick TO MAKE NEXT RED OTL E eel 2 ONS eee eee i FILE IN SOURCE Let s see how this works lt F gt gt l we have already AREA seen as a command which will set the pointer to the end CURRENT of the file This may not be necessary if you have done no editing on the current file but we ll include it to be safe Now let s look at the argument in the second command on the command line above Becuase it is the argument of an F command the initial reference interval is the entire source area so specifies everything in the source area following and including the at the end of the current file 2 specifies the second character within the interval we just mentioned If there is a file following the current file then the argument 2 will give the first character of that file 1 would give the forming the lower bound of the file However if there is no such file then after lt F gt gt 1 will point to the highest address in the source area so the argument 2 will produce a matching failure and hence an error Thus the F 2 will give an error if there is no file following the current file or else if there is such a file will mak
162. le to the next JOAN Our command gt JOAN will do exactly that with a minor bit of nuisance This com mand will print the interval from the N in the JOAN we just saw to the next occurrence of JOAN If this JOAN is not the right one either we can type exactly the same command gt JOAN and keep on typing it to see the consecutive intervals of text prior to successive JOAN s Actually as we ll see at the end of section I l there is a special command called DEF which is very useful in exactly this kind of situation where we want to keep giving the same command repeatedly The DEF command will allow just carriage return to invoke a special default command The minor bit of nuisance referred to is the fact that the N from the last JOAN we looked at will begin the printout We ll see in a minute how to get rid of that The second time we type gt JOAN the symbol gt will be the upper address from the last interval compu ted which in this case will be the N from the second occurrence of JOAN Each time we type the command a new interval will have been computed so the value of gt will change Note that to use this method you must not use any other command between the instances of gt JOAN since an intervening command will probably cause an interval to be computed so that gt will not have the value we want it to si After having typed in the command gt JOAN enough times we should come to the JOAN we want
163. listing will not be in the P command list ing One note The program assumes that the Processor Technology format might be in force before it is assem bled so the INTE pseudo begins in column 2 to indicate an empty name field If the Intel format is in force then the lack of a colon automatically indicates an DISKATE User s Manual empty name field 96 DISKATE User s Manual I 5 Invoking DISKATE In this section we will discuss the various ways of invoking DISKATE including the power up procedure It might seem like a dirty trick to have put off the power up procedure until so late into the manual but there is a good reason for it When DISKATE initializes it issues internally commands that are exactly like some you might give yourself from the terminal so it is important to have an understanding of how they work First a word about how DISKATE is organized DISKATE actually consists of two separate but linked programs ATE and IO ATE consists of everything but the I O routines and is completely independent of any particu lar I O or disk configuration IO contains the 1 0 routines including the disk drivers and also contains a loader which loads ATE To bring up DISKATE all you do is instruct your system to load and execute IO If your disk will allow any program to be bootstrapped as does the Disk Jockey you will probably want it to boot strap IO This is the way DISKATE is provided If you are uSing the Nor
164. ll call for a search through the entire initial reference interval in this case the current file starting at the beginning We have already axed what was the first occurrence of SEX so the pattern SEX will NOW match what originally was the SECOND ocurrence of SEX since we deleted the original first occurrence This means that to kill the next occurrence of the pattern we can give the identical command K SEX and keep on giving the same command any number of times to kill off the offending lines one by one Suppose for R the moment that we happen to know that the file has 47 COMMAND entries so that the command will have to be given 47 times It s obviously a ridiculous nuisance to have to actually type in the command this many times Fortun Rn ately there is a DISKATE command which will cause other commands to be REPEATED a given number of times This CAUSES is the R command In a multi command command line the REST OF sequence COMMAND a ge Aigo me te he l TINE Rn othercommands TO BE REPEATED will cause othercommands to be repeated n times We are n TIMES now able to purge our notorious file of its baser instincts by the single command R47 K SEX A word of warning In giving such commands as this you have to be extremely careful that a pattern in the part of the command to be repeated does not match more than you intend Suppose for instance that instead of using the pattern SEX we had used
165. ll go ahead and register the drive as the current drive without giving an error message but after that EVERY SINGLE DISK COMMAND FOR THE CURRENT DRIVE WILL GIVE A DISK ERROR You will know what has happened since the disk error messages print the drive designator for the drive that caused the error and this will show a drive that you don t have You can RENAME a file with the RN command in the P E EME E E RN oldname newname be sure to remember to separate the two file names by a Meisn lt R gt GIVES INTERVAL OCCUPIED BY LAST RECORD LOADED FROM DISK K lt R gt DELETES LAST TEXT LOADED CD COMMAND TAKES DISK DRIVE DESIG NATOR AS ARG amp MAKES IT CURRENT RN COMMAND _ RENAMES FIRST FILE DISKATE User s Manual blank Of course in renaming a file the file does not change its location on the disk so that a disk drive designator in newname is superfluous If one is present it is ignored even if it incorrectly specifies a drive other than the one the file oldname resides on It frequently happens that you want to copy the contents of a file from one file to another This is done with the DISKATE T command T for Transfer The T command has the form T oldfile newfile In this case drive designators are significant since the file to which you want the information copied could be on another drive from the file which is to be copied The DIRECTORY ATTRIBUTES i e size file type an
166. locate where concepts are explained You may also want to use this space for notes of your own Part I is divided into sec tions based on how DISKATE is to be used This is done for great er coherence in explaining how the commands work but there will inevitably be some overlap Some of the commands work differently in monitor usage than in text editing usage for example and so are discussed in both sections All DISKATE commands are intro duced in Part I with the single exception of the V command which can only be used in machine language programs and so is discussed in Part II Part II is called Installation and Maintenance It explains the nitty gritty of how to get this software running on your hard ware and how to make the proper changes in DISKATE to customize your system It also gives the details of how to interface DISK ATE to machine language programs Part III is the System Reference Summary Almost all the ma terial in Part III is all contained in Parts I and II but appears here in a condensed form so that you can quickly locate something you need Experienced programmers who are used to learning new languages from tersely written manuals may want to skim lightly over Part I and mainly consult Part III If you are confused about what a command does by a discussion in Part I it may help to refer back and forth to both Parts I and III This is a large manual You may be daunted by the task of reading all of it before beginn
167. luated DISKATE restricts the scope of the argument to an INITIAL REFERENCE INTERVAL For most commands this initial reference interval is the current file which is why our previous oversimplification is normally not misleading DISKATE has a special operation which allows you to in effect set the initial reference interval to something else Whenever DISKATE encoun ters a in evaluating an argument THE INITIAL REFER ENCE INTERVAL FOR THE REST OF THE ARGUMENT IS THE INTERVAL CORRESPONDING TO THAT PART OF THE ARGUMENT ALREADY EVALUATED The easiest way to understand this will be by example The argument A will to be precise match the first occurrence of A in the initial reference interval which is normally the current file The argument JOAN will match the first occurrence of J 0 A N in the ini tial reference interval which is normally the current file The argument JOAN A will match the first occurrence of A WITHIN the first occurence of J 0 A N in the initial reference interval normally the current file The command K JOAN A will achieve exactly what we want provided of course as we mentioned that we are sure that the JOAN in question is the first occurrence of JOAN in the current file In a larger file however we would still general problem of making sure that we were in X i have the the right 8 INITIAL REFER _ ENCE INTERVAL GIVES THE SCOPE OF MATCHING USUALLY CURRENT
168. ly a single source file all by itself it s always good practise to give an O command with no argument first That way you have an empty source area to work with Obviously you can use the L command to load into memory a file which was previously saved But there are also some powerful uses for the L command in addition to this We recall that a K command followed by an E com mand will replace the deleted text by the text entered with the E command A K command followed by an L com mand which gives a source file as the first argument and has no second argument will REPLACE THE DELETED TEXT BY THE LOADED FILE In this way it is possible for instance to maintain a library on the disk of standard paragraphs use an abbreviation to denote the paragraphs in editing and then write a macro which will substitute directly from disk the paragraph for the abbreviation This can be used to write form letters just to cite one application Another important way the L command can be used is to concatenate or combine files Suppose we go back to the example above where a file SPECS was split into two files SPECS and SPECS Suppose that after further editing the verbosity is reduced and the two file are now small enough that it would be more convenient to have them combined into a single file Because the two files were saved from the source area they are both source files The job we want to do can be done by the following commands O I L SPECS L
169. ly for the next time around the G loop There is one difficulty with this method invoking a macro or using G commands with an argument like name will work perfectly well but could end up being extremely slow since each time the command is executed DISKATE has to search through the whole source area looking for the argument One way around this is to use variables For instance the above example could also be done this way a eo seir mamm QFX Y _ K E G FUSE CHOP a E a QFX Y 2 KI 1 E_ G CHOP PARAGRAPH aes FUSE lt S gt FUSE D FUSE D a CHOP lt S gt CHOP x D CHOP QS Here the macro assigns the DISKATE variables FUSE and CHOP to the respective arguments This way the G com DISKATE User s Manual mands do not call for any matching and so execute much faster Note that whereas a D command could be given simply in the form D FUSE in the statement above the form lt S gt FUSE had to be given The reason is that the D command is one of the few commands for which the initial reference interval is the entire source area so the prefix lt S gt was not needed For the statement the initial reference interval is the current file so the source area must be explicitly set as the initial reference interval There is one VERY IMPORTANT CAUTION when using the G command with an argument which is a variable rather than a pattern matching argument If the macro is located inside the s
170. make it possible to get into the following tangle If you make a mistake entering the file name in the S command and specify a file which doesn t exist when you meant to specify a file which does exist the file will be stored under the incorrect name This will have two consequences there will be an extra file on the disk under an erroneous name and the file with the correct name will not contain the latest version of the file Below we ll see some ways to prevent this type of situation One other caution To specify the current file as the interval to be saved we used the argument lt F gt We could have used instead of lt F gt since like most DISKATE commands the initial reference interval for the S amp S command is the current file Remember though that iff you are using the REF command to specify a different initial reference interval this will also apply to the S command We used lt F gt to specify the current file because it gives the current file without matching regardless of the initial reference interval Here s one way of avoiding the problem of saving a file under the wrong file name Whenever a DISKATE command is to be given a file name the symbol can be given in place of the file name and refers to THE MOST RECENTLY REFERENCED FILE Thus if you create a brand new file and save it by a command S SPECS lt F gt as above and then do some more work on the file and want to save the updated version as the sa
171. mbols In all of our examples in this manual we have gone to the extra trouble to set the dummy variable just to be safe Now that we ve thoroughly discussed the symbol table we can get on to describing how the assembler works A complete understanding of lt T gt is crucial because as mentioned above not properly managing lt T gt is the greatest source of peculiar assembler errors Assembler language statements are divided into two types PSEUDO operations and MACHINE INSTRUCTIONS The machine instructions are each coded into a single machine language instruction of object code whereas the pseudo ops are directives to the assembler itself There are four fields to every assembler statement some of which may be empty l The label field 2 The opcode field 3 The operand field 4 The comment field The DISKATE assembler allows you to use TWO DIFFERENT formats for specifying these fields In the PROCESSOR TECHNOLOGY FORMAT a The label field begins in the first character of each line If an assembler statement is not to be 82 PROCES SOR TECH NOLOGY FORMAT first machine instruction is to be located DISKATE User s Manual given a label the first character of the line must be blank The label is given without any symbol to terminate it A label must contain only upper case letters or digits and must begin with a letter b The opcode field is separated from the label field by one or more blanks C
172. me file you can use the command oe te IF NO SUCH FILE EXISTS IT IS CREATED SYMBOL CAN BE USED IN PLACE OF FILE NAME FOR MOST a DISKATE User s Manual 71 RECENTLY S lt F gt l REFER ENCED PROVIDED you have not given any disk commands with other FILE file names or changed the current drive in the meantime Even immediately after power up will have a meaning since as we ll see below when DISKATE powers up it references a special file called STARTUP If you want to see what the most recent file is COMMAND you can give the command SHOWS lt p MOST RECENT This will print the file name and drive on the terminal FILE Use of the file name provision is highly recommended So far we ve said nothing about the length of the file after an S command When the S command is given the length of the file will be set at just enough blocks FILE to hold the interval saved Suppose the file already LENGTH exists and the length is not long enough to hold the SET AT interval If there is a contiguous block on the disk JUST large enough for the interval that is free DISKATE will LONG save the file in this block and release the block where ENOUGH the file existed before If there is enough free space TO HOLD on the disk to hold the interval but not in one block INTERVAL DISKATE will compact the disk and make the file just long enough to hold the interval In this case a message IF FILE of the form
173. ment and then is reset BY THE SAME RELATIVE AMOUNT AS amp WAS This means that a source program only needs to have one instance of a separate AORG and SORG at the beginning after that ORG will maintain the proper offset between amp and that was established with the AORG and SORG However if you go on to assemble a new program with a different offset between the initial value of amp and the initial value of you may have to first reset amp and from the terminal Operands of assembler statements can be labels and can include arithmetic operations and provi ded the expression can be given a value at the time the assembler encounters it You can also use DISKATE symbols such as lt S gt lt T gt amp etc but you can t use complex arguments involving matching of the kind we saw in section 1 The pseudo op EQU for EQUate can be used to assign the value of an expression to a label If an EQU statement has a label though that label had better not appear in the operand Thus if you had a table and wanted to set a label equal xe the length of the table you could have TABLE TBLEND EQU amp TBLLEN EQU TBLEND TABLE 85 ORG SETS amp TO OPERAND AND CHANGES BY THE SAME AMOUNT AS amp WAS CHANGED PSEUDO EQU ASSIGNS LABEL THE VALUE OF THE OPERAND DISKATE User s Manual but you can t have TABLE EQU TBLEND TABLE Also you cannot give a label a value with
174. ment _ by itself would match the first occurrence of a eee return in the current file so the argument SEX will match the interval starting with the first occur rence of S E X through the first carriage return after the S E X The symbol the E command _ can also be used as an argument for The command E will enter a carriage return just as would the command E In many cases it is advantageous to use the symbol in an E command to have the command fit on one line Suppose we want to see more than one line To do this we can use the same symbol in connection with another operation supported by DISKATE OCCURRENCING An argument using occurrencing gives a number n followed by a pattern and matches THE NTH OCCURRENCE OF THE PATTERN The argument 2_ by itself will match the 2nd occurrence of a carriage return in the current file The argument 2 A will match the 2nd occurrence of the character A To continue with our same example the command NAME e 24 will print on the terminal NAME JOAN DOE OCCUPATION DEER SEX FEMALE o _ ADDRESS THE WOODS Here the argument NAME 4 apecifias the interval from the first occurrence in the current file of the charac ters N A M E through the fourth occurrence after the NAME of a carriage return This will give us four lines from the current file beginning with the first occur rence of N A M E There are a couple of points to be noted here Using an argument
175. n error in an argument causing it to specify an address outside of the current file If you want to edit several files the N command can be given any number of times as long as you don t run out of memory By uSing the N command you can create several files in the memory Now the question is suppose you want to go back and look at one of the previous files To do this you use the F command F for File The F command is different from all of the commands we have described so far in that THE INITIAL REFERENCE INTERVAL FOR THE F COMMAND IS THE WHOLE SOURCE AREA The actual operation of the F command is somewhat detailed but the basic idea is that it TAKES THE LOWER ADDRESS of the argument given and MAKES THE FILE CONTAINING THAT ADDRESS CURRENT Remember that a file in the memory is bounded by s When DISKATE is given an F command it evaluates the argument and uses only the lower address First it looks to see if this location contains a if so it assumes that this is the beginning location of the file If not it begins searching backward from this location looking for a and when it finds it sets this as the beginning of the file Having found the beginning of the file DISKATE searches forward from the argument location looking for another which will be the end of the file When it has determined the file boundaries the file is made current and THE TARGET CHARACTER IS SET TO BE THE WHICH ENDS THE FILE The F command
176. n the target character will be the N in the JON We just deleted the A To insert the H we will want to use an E command and remember the E command inserts the char acters of its operand between the target character and the preceding character Since the target character is K now the N in JON anything entered with an E command FOLLOWED immediately after the K JOAN A will go between the N BY E in JON and the previous character which will be the O in JON This is exactly where we want to put the H REPLACES The command gt 7 KILLED 7 TE w INTERVAL DISKATE User s Manual K JOAN A E H will replace the A in the JOAN with an H in a single command In general Kinterval E text will REPLACE the characters in the interval with text At this point we come to what is possibly the greatest single potential source of error in ordinary DISKATE usage Regardless of whether you want to use an E command to replace with new text an interval which has been deleted with a K command the K command will set the entry pointer so that the target character is the first character after the deleted interval It frequently happens in using a text editor that you will enter some text discover an error perhaps go back and make a change and then resume entering more text Here we have a problem Any change in the file made using the K command will reset the entry pointer If we don t somehow save the value of the entry pointer before the c
177. n an assembler source program Here is one way the amp and can be used Suppose you have a machine language program which you wish to use in connection with editing but you don t want to commit yourself to where in the memory it will go Now suppose you have done some work on a source file and have decided you want this program to be brought in from the disk and to reside 1924 bytes past the end of the source area First give the command amp S 1924 S then assemble the program directly from disk We ll see how to do this below The method of not specifying initial values for amp and within a source program but relying on and amp commands from the terminal can be used to great advantage for flexibility as to where a program will go but it does have some risk Assembling the program and forgetting to give the values for amp and will deposit the code in the wrong place in memory which could lead to trouble In many cases possibly in most want to set the initial values for amp the assembler source program There are three pseudo ops for this purpose The pseudo op AORG for Assembler Origin will set amp to the value of the operand while SORG for Storage Origin will set to the value of the operand It is important to remember that a SORG state ment DOES NOT TAKE EFFECT UNTIL PASS 2 unless is used as part of an operand of an assembler statement A SORG statement will set during Pass 1l but after that
178. n at this location The first 1924 bytes of this buffer are used only during disk commands The rest is shared by other buffers The end of RAM used internally by ATE Memory above this address is free to the user DISKATE User s Manual 146 Index following a page number indicates the page number is part of the System Reference Summary Non alphabetic characters are given in order of their ASCII codes A assembler error code 91 142 A Command 88 137 Al Command 88 137 A2 Command 88 137 AORG Pseudo 84 148 Argument missing 21 Argument Passing to machine language 114 Arithmetic Expressions 122 ASC Pseudo 87 141 ASPC 144 ATE Command 99 133 ATE Module 97 191 112 ATECOMS 118 142 ATETBL 92 142 B Command 54 1293 Backspace key 4 64 121 Base current 53 54 BEGIN 142 Block disk 68 BOFP 116 144 Bootstrapping 97 BOSAP 115 144 BYE Command 169 133 C Command 61 129 CD Command 75 137 CHPTR 144 COM Command 98 133 Commands format of 121 Compacting Disk 71 Concatenation 11 55 125 Counting occurrences 58 Create disk file see S T Commands Current Drive 67 75 Current file 1 18 41 43 115 Cursor frozen 4 CVALS 114 143 D assembler error code 91 142 D Command 46 131 DB Pseudo 86 148 Decimal numbers 53 DEF Command 58 138 Default Interval Argument 126 See Argument missing Device numbers 64 193 ee Directory 68 72 76 12
179. n give the command Z gt KEEP You can also zero a specific individual symbol This is done by giving the Z command with the symbol as an argument The symbol will be removed from the symbol table and the symbol table will be compacted if the sym bol zeroed was not the last symbol in the table Note that in giving this command there must be NO BLANK BETWEEN THE Z AND THE SYMBOL If there is such a blank DISKATE will perform a straight Z command and zero the 89 lt T gt Z COMMAND NO ARG ZEROS SYMBOL TABLE EXCEPT FOR INITIAL SYMBOLS Z gt symbol COMMAND ZEROS ALL SYMBOLS DEFINED AFTER symbol Zsymbol COMMAND ZEROS ONLY symbol DISKATE User s Manual entire symbol table There are several ways of managing the symbol table using commands we have already seen Above we mentioned that the M command will update all of the relevant DISK ATE pointers This includes the pointers that define the boundaries of lt T gt If your memory is limited you may want to allocate only a small amount of space to lt T gt while editing when the number of symbols in use is small but you will need a lot of space for lt S gt But as we ll see DISKATE can assemble directly from disk so that when assembling you may not need a large source area but will want a large space for lt T gt Thus the ability to change the location of the symbol table in the memory is important Let s say you have the symbol table
180. n the memory of all the files in the source area This can be achieved by using the C command C for Copy which works similarly to an M command with the target character is outside the source area The C command copies the interval given as argument to the memory locations beginning with the target character Whatever used to be at the block beginning with the target character is simply overwritten The C command works the same whether the target character is in the source area or not and does not update any of the DISKATE internal pointers as does the M command Thus if you give the command 1 0 H and this location is outside the current source area the difference between M lt S gt and ces Se eae is that after the M lt S gt the current source area will begin at 1809H the pointers to lt S gt having been updated whereas after the C lt S gt command the source area will be peace y where it was before the command You should be gt r Since it would be easy to overwrite a part of the memory DISKATE is using for something else As you might suspect by now the K command also works differently if its argument is outside the source area In this case rather than removing the interval given as the argument the K command will ZERO this interval Zeroing a block of memory can be especially l C COMMAND COPIES INTERVAL GIVEN AS ARGUMENT TO BYTES STARTING WITH TARGET CHAR NO POINTE
181. nd character WITHIN the interval from the N in the current JOAN through the next entire JOAN This is exactly what we want the same interval we were getting before but with the first character trimmed off Here s a different kind of example that shows how you can use the operation together with the matching symbol Suppose the file you are editing is a program and you want an argument which will match the first occurrence of a variable name let s say BUFFER which occurs at the BEGINNING OF A LINE To say that it occurs at the beginning of a line means either that it will be the very first thing in the file or else it DISKATE User s Manual will be immediately preceded by a carriage return To simplify things let s suppose we can rule out the first possibility Here s the argument that will work _ BUFFER 2 The argument BUFFER matches the first occurrence of a carriage return followed by the characters B U F F E R so the whole argument matches everything from the 2nd character on within the first occurrence of a carriage return followed by BUFFER That is the argument Matches the everything in BUFFER after the carriage return which is what we want Let s return now to our example of deleting the A in JOAN and see another way it can be done Recall that the last way we did it we ended up marching through the whole current file using commands until we found the right JOAN But Suppose we could
182. nd if the symbol is undefined it is added to the end of the table and TABA updated T denotes the same value as 1 the contents of TABA On power up and after the Z command the symbol table contains values for the following symbols used by the assembler A B C D E H L M SP PSW Command Summary lt Interval Argument gt Ba Quote interval Prints on the terminal the characters contained in the interval given as argument with a line feed issued automatically at each carriage return no argument Quote one line Prints on the terminal the line containing the target gt character with the character immediately preceding the target character E lt Entry Argument gt where Enter DISKATE User s Manual 128 an Entry Argument is a seguence of Matching Symbols of the form text or Seguence of numbers or _ as defined above In this case text may extend across any number of lines If the entry pointer is within the source area the char acters denoted by the argument are inserted into the cur rent file between the target character and the preceding character expanding both the current file and the source area Otherwise they overwrite those characters in the memory beginning with the target character In either case the target character after the command is the charac ter after the material entered by the E command K lt Interval Argument gt Kill Kills the inter
183. ng the M command with an interval outside the source area which may save you some trouble When you are using the editor deleting text using the K command actually works internally by calling the routine invoked by the M command The INTERVAL TO BE DELETED IS MOVED IMMEDIATELY OUTSIDE THE SOURCE AREA and in the event that you change your mind and want back the interval deleted it can be recovered if the source area has not been enlarged However the deleted text is NOT surrounded by s so it may cause 59 M COMMAND UPDATES POINTERS E G M lt S gt DOES MOVE amp UPDATES lt S gt lt F gt M COMMAND WITH OUTSIDE SOURCE AREA SIMPLY COPIES INTERVAL TO BYTE STARTING WITH TARGET CHAR K COMMAND MOVES Z DELETED INTERVAL JUST OUTSIDE SOURCE AREA DISKATE User s Manual some trouble in general to determine just what the boundaries are for this piece of text However if you act fast enough you may be able to recover deleted text with very little trouble The M command updates all internal DISKATE pointers that it knows about which includes lt and gt After a Move command these two pointers will give the location TO WHICH THE INTERVAL HAS BEEN MOVED Since the K command issues an internal call to the M command routine immediately after a K command lt and gt give the lower and upper addresses respectively of the area in memory to which the killed text has been moved Thus if you
184. nos o m GIVES DISKATE User s Manual Literally the argument 5 means anything followed by the 5th carriage return Now is the time to confess to a little white lie So far in this manual we have seen the use of the symbol both as an operation for combining intervals and as a matching argument In fact instead of you can use any number of dots from one on up so for example the argument patternl pattern2 has the same meaning as patternl pattern2 or patternl pattern2 Let s see why this ambiguity causes DISKATE no problems The actual symbol which DISKATE recognizes internally is the Single dot Since can be interpreted as a matching argument which matches anything can be thought of as a matching argument which means anything followed by anything Clearly this has the same meaning as just plain anything which is why any number greater than zero of dots amounts to the same thing The use of two dots here is simply personal preference A single dot by itself in an argument is easy to miss which is why two are being used in this manual but in your own work you should use either one two or three dots together as your preference suits We ve seen so far that specifying an interval which matches a given line number or a give group of lines identified by line number can be slightly tricky DISKATE provides a way to deal with numbered lines directly which eliminates some of this difficulty but which is n
185. o see why we have to keep at the business of going through exactly how DISKATE will interpret the argument The argument JOAN is the interval consist ing of the first line in the source file containing the PATTERNS Characters J O A N This INCLUDES THE CARRIAGE RETURN INCLUDES at the end of the line The argument JOAN 4 is the FINAL interval beginning with JOAN through the 4th occur CAR RET rence of a carriage return AFTER the JOAN Since JOAN already includes a carriage return in order for this argument to work there must be at least 5 carriage returns in the file but altogether there are only 4 DISKATE User s Manual 19 This explains the error message The argument gave a pattern which DISKATE couldn t match Note that in this case there is nothing syntactically wrong with the command JOAN 4 the command caused an error because the pattern failed to match DISKATE does not distinguish in it s error message between syntax errors and logic errors i e for instance errors that arise because a pattern has no match in a particular file but which might give a perfectly legitimate match in a different file Occurrencing together with the operator gives us a simple way to specify the n th line in a file For instance suppose your terminal is a video terminal with 24 lines and you want to view your text in groups of 29 lines this way there will be a little overlap and room for DIS
186. o the hardware terminal width will produce an extra blank line for lines whose length is exactly the terminal width WID no argument Prints the current terminal width on the terminal ECHO lt one line character string gt The character string given as argument is printed on the terminal There is no matching PAUSE no argument The message PAUSE is printed on the terminal and the panic state is entered IO lt arithmetic expression gt set I O device number The low order byte of the value of the arithmetic expres sion is subsequently supplied in the A register as an I O device number to all calls to the terminal I O routines A carriage return and line feed are printed on the device L lt filename gt lt Interval Argument gt Load disk file DISKATE User s Manual 135 The file is loaded at the lower address of the interval given as the second argument The Default Interval Argu ment may not be used as the Interval Argument L lt filename gt no Interval Argument If the file is not a source file it is loaded at the address given by the memory address portion of the direct ory If the file is a source file it is loaded at the address given by the entry pointer If the entry pointer is within the source area the file is inserted into the memory file containing the target character between the target character and the preceding character with any bounding s removed In this case the target character b
187. od for personalizing values is to set them however you would like and then save ATE on the disk with these values having been established Whenever you do this YOU MUST MAKE CERTAIN ATE WILL BE THE SECOND ENTRY IN THE DIRECTORY Otherwise the loader in IO will load the wrong file If you save a new copy of ATE any commands you have given which change internal settings such as CD INTE and the like will result in DISKATE being brought up at power up in just the state in which it was saved unless settings are changed by the STARTUP file or the power up initial command string 0 2 Y Thus for instance if you move the source area and the symbol table then save a new copy of ATE afterwards at power up the source area and symbol table will begin at the same places they did at the time ATE was saved but the symbol table will contain only the initial assembler variables and the source area will be empty To aid in this process you can use the symbols contained in ATETBL ATETBL is a file on the DISKATE diskette which gives an abbreviated version of the actu al symbol table from an assembly of your version of ATE For the module I0 of course the entire source code is given ATETBL includes a large number of addresses which are important for use in machine language pro grams which will be discussed in the next section A listing of the meaning of the symbols in ATETBL is given at the end of Part III In this section two symbols concern us part
188. ommand JOAN in this case gt will be the address of the N in the first JOAN in the file The argument gt then will match everything in the current file from the N in the first JOAN onward There is an important but subtle point here that we should note before going on In all the cases we have seen DISKATE arguments are intervals i e a pair of addresses But here we ve just said that gt denotes the upper address in the last interval computed by DISKATE This implies that gt denotes not an interval but a single address In fact any symbol or expression in DISKATE that determines a single address is also interpreted as AN INTERVAL WHOSE UPPER AND LOWER ADDRES SES ARE EQUAL We said that an interval was a pair of addresses in which the first address was not greater than the second but they certainly may be equal For example in evaluating sa tebe ha DISKATE takes the SAPEESe LON 9 gt GIVES UPPER ADDRESS IN LAST INTERVAL COMPUTED SINGLE ADDRESS EQUIVA LENT TO INTERVAL WITH a UPPER AND LOWER ADDRESS THE SAME E DISKATE User s Manual 29 2000H to be completely equivalent to the expression 29000H 2000H Note that this last makes perfect intuitive sense Now let s return to the example We started with the command JOAN and could see from it that the first occurrence of JOAN was not the right one What we said we wanted to do then was view the rest of the fi
189. ommand takes an interval as its argument and prints on the terminal all the _ characters in the interval issuing a line feed also whenever it encounters a carriage return To see some examples we have to consider how intervals are specified The Simplest way to specify an interval is to give 2A9 0H r COMMAND PRINTS LINE WITH TARGET CHAR INTERVAL IS A i PAIR OF ADDRESSES COMMAND PRINTS INTERVAL DISKATE User s Manual 6 explicitly the pair of addresses For instance the command 2000H 2A90H will cause the characters in the interval from 2999H through 2A9 H inclusive to be printed on the terminal To give a complete explanation of the meaning of the symbol at this point would be premature so for the moment let s leave it that an interval can be specified in the form explicit adress explicit address Normally you will give the addresses explicitly when you are using DISKATE more as a monitor than as an editor so there will be more about this in section I 2 A more sophisticated method of specifying an interval is to use an argument that calls for MATCHING In using matching you present to DISKATE a pattern of MATCHING characters and it searches the memory looking for the pattern If it finds it then the area that matches the pattern becomes the interval A simple form of matching argument is similar to one we have already seen with the E command The argument TEXT l GIVES text
190. onse to the command DISKATE would print on the terminal something like 200D Cs 63 29 The C3 is the opcode for the JMP instruction and the jump is to location 2993H Remember that 8488 instruc tions put the low order byte of an address at the low order location The C3 we want to leave alone but the two bytes 93 29 we want to change to 7A 39 Just as we did when entering text to use the E command to enter numbers we first have to Set the entry PORRES The command to do ene Whore Job is 290EH E 7A 308 ee ee Note that we say 200EH instead of 28 DH because at _20 DH is the opcode C3 which we don t want to change _ Obviously it should go without saying that if you use DISKATE as a monitor as in this example and then go back to editing a source file having failed to restore the entry pointer to the right place back inside the source area a total disaster could result Entering 55 NUMERIC CODES CODES MUST BE IN CURRENT BASE WITHOUT Q OR H SUFFIX SEPA RATED BY BLANKS A IF OUTSIDE SOURCE AREA E COMMAND WILL OVER WRITE NOT INSERT DISKATE User s Manual text following a monitor usage such as we just saw without having restored the entry pointer could very well dump garbage on top of part of the resident software which could certainly result not in merely garbaging up your source file but in a total crash which could wipe out the contents of the entire memo
191. ot as general in building complex expressions The symbol is similar to in that it refers to an entire line but rather than being a general suffix which can be appended to any argument the symbol is only placed after an occurrence number to specify the n th line in the current file Specifically the argument a n specifies the n th Tine in the current file This is NOT A MATCHING ARGUMENT When evaluating n DISKATE searches through the current file for the nth line without regard to anything else Let s see noy this compares to our previous examples Above we saw an example of a command using expressions of the form n_ for showing 29 lines beginning with line 41 To do this using the owe could use _the command symbol 41 60 Notice here that 60r is not a matching argument so it definitely does not match the 6 th line after the 4lst sias 1ST n LINES kd o ETC ALL HAVE THE SAME MEANING n SPECI FIES THE N TH LINE DIRECTLY NO MATCHING DISKATE User s Manual line Instead 680 simply specifies the 69th line period no matter where the expression 68 occurs The is very useful when you want to see specific numbered lines in a file but not as useful when you want to specify lines relative to a content oriented argument We saw for instance that the command JOAN 3_ will print on the terminal four lines beginning with the first occu
192. ource area and follows the text to be edited editing commands within the macro can cause the size of the text to change and thus cause the entire macro to be moved about in the memory as the source area behind the change is shifted up or down In this case the value of the variables used as arguments of G commands will NOT be updated as the source area is shifted It just happens that in the macro above no net change in the length of the text is caused so this problem does not occur One way of preventing this difficulty is to locate edit macros outside of the source area completely This can be done for instance by editing the macros in the source area storing them on disk for future use then loading them into a vacant part of the memory which will not be overwritten by the source area Or as we ll see below it s possible to have more than one source area though of course only one at a time can be current Another way of getting around this problem is to put the macros at the BEGIN NING OF THE SOURCE AREA That way as changes are made within the text addresses within the macros will not change Of course when using only pattern matching arguments the problem does not occur at all An alternative method to using G commands would be to use the R command but this would require that FUSE CHOP and PARAGRAPH be in separate files since the R command will repeat to the end of the entire file until a OF or QS causes a quit To ente
193. outputs of numeric codes from DISKATE will always be in split form As mentioned briefly in the previous section num bers can be used to specify an interval explicitly For instance the first 8K bytes of memory can be specified by the intervals 0 1FFFH 37 377Q 31 255 To view an interval in memory in numeric rather than ASCII form the command may be given It works like the command except that interval given as the argument is output byte by byte in numeric form in the current base For each line of output there will also be a field at the left of the printout showing the address of the first byte being printed on that line The command will thus perform a CORE DUMP to your terminal of the interval given as the argument Arguments given with the command may call for matching just like so many of the ones we saw in the previous section The command is especially useful for viewing non printing control characters if these Suppose for instance you Suspect that there is a non printing character in the line in the current file containing the target character The command es will dump this line on the terminal and you can inspect the codes to try to find out where there may be a trouble spot You can also use the E command to enter numeric 54 OCTAL HEX NUM BER MUST BEGIN WITH DECIMAL DIGIT B COMMAND Bn SETS n AS CURRENT BASE COMMAND DUMPS INTERVAL TO TERMINAL BYTE BY
194. r each file for which there is room and then an error message will occur indicating that the disk or directory is full RN lt current filename gt lt new filename gt Rename The file whose name is given by the first argument is renamed to the name given as the second argument If the second argument includes a Disk Drive Designator it is ignored W lt filename gt lt Interval Argument gt Write memory address i DISKATE User s Manual 137 The lower address of the interval given as the second argument is written in the memory address portion of the directory for the file name given as the first argument CD lt Disk Drive Designator gt Current Drive The disk drive designated by the argument is established as the current drive A lt list of file names separated by blanks gt Assemble Invokes both passes of the assembler to assemble the source program logically equivalent to the concatenation of the files given in the argument list may be included as a file name to cause a PAUSE command to be executed while diskettes are changed The name of each file in the list is printed on the terminal as it is assembled by each pass The source area is not affected A no argument Invokes both passes of the assembler to assemble the cur rent Eiee Ea lt F gt Al lt list of file names gt Assembler Pass 1 Al no argument B Identical to the A command but invokes only Pass 1 of the assembler T
195. r the macros using this method you could use the port OMrng sequence of commands a oN ale E FUSE R 1 QFX TN E CHOP R 1 OFX N E PARAGRAPH K EL 49 EDIT MACROS WHICH CHANGE THE LENGTH OF TEXT MUST NOT FOLLOW TEXT IN SOURCE AREA CAN BE AHEAD OF TEXT OR OUTSIDE SOURCE AREA DISKATE User s Manual D FUSE xX D CHOP A special command which can be used in macros or command lines is the PAUSE command The PAUSE command which has no argument will cause the process in which it occurs to stop in a panic detect state You can stop to examine the printout on the terminal and then resume execution of the process by typing any character other than S or ESC The panic detect is discussed more fully in Section I 3 The message PAUSE is printed on the terminal to notify you that you re in the pause state This command is especially useful for such things as changing diskettes We conclude the discussion of using DISKATE as a text editor with two special commands which can be used aS a convenience to save extra typing The command DEF takes as its argument a string of characters up to 24 in length enclosed in brackets as you would for an E command DEF stands for Default When DISKATE is given a DEF command the string forming its argument is saved in a special internal location Thereafter whenever DISKATE expects a command if you simply type a carriage r
196. re 8l M lt T gt COMMAND WILL RELOCATE SYMBOL TABLE TO START AT L g filename lt T gt WILL OVER WRITE lt T gt TO SET UPPER BOUND SET UNUSED VARIABLE DISKATE User s Manual finding it But in defining a new symbol the size of the symbol table increases so a new upper bound for lt T gt must be recorded This is why an command to a non used variable will set the upper bound This technique will be the cornerstone of the process of customizing DISKATE discussed in Part II In many cases you may find that commands will work properly if you give only an L filename lt T gt command without setting a dummy variable For instance you could give a command L TBLSAVE lt T gt and then a command symbol where symbol is one of the symbols in TBLSAVE and even though DISKATE will not have recorded internally the proper upper bound for the symbol table TBLSAVE the command will work properly anyway This is because the command will scan through lt T gt taking the at the end as its indication of the end of the table rather than consulting the pointer giving the upper bound However if you have not set the value of a dummy variable and then give the command lt T gt an incorrect value will be printed for the upper bound of lt T gt Under these circumstances if you move lt T gt with the M command you may very well end up moving only a small part of it and losing most of your sy
197. re will almost always be many different ways of accomplishing a given thing in DISKATE and after using it for a while you will surely develop your own style The examples presented here reflect the style of the author of this manual a style which you may not care for once DISKATE becomes familiar One of the things that has been stressed above and will be stressed below is that you should always look before you leap give DISKATE a command or command which will preview any change you are about to make Constantly giving such viewing com mands may seem tedious but they will save an enormous amount of grief that can result from giving an erroneous command We ve seen many examples of various simple dele tions deleting an A from an occurrence of JOAN to change it to JON The deletion itself was usually easy but the work came in specifying where it was to take place Now let s suppose we wanted to change the JOAN instead to JOHN deleting an A from JOAN and inserting an H Without repeating all of the previous examples let s suppose we have set things up by various commands perhaps so that in the last interval computed the first JOAN is the one we want to change We saw that the command K JOAN A will delete the A from this JOAN Now we want to add in the H As mentioned above the K command sets the tar get character to be the first character after the inter val deleted After the command K JOAN A the
198. return and line feed for an overflow line 2697 DEVICE Y COMMAND EXECUTES TERMINAL INITIAL IZATION ROUTINE NO ARG WID COMMAND SETS TERMINAL WIDTH TO ARG WITH NO ARGUMENT PRINTS TERMINAL WIDTH DISKATE User s Manual There is a simple command for printing a message on the terminal The command ECHO one line character string will simply cause the string within the brackets to be printed on the terminal This differs from the com mand in that to print a message on the terminal using a command the message must be part of the current file The echo command will print its argument without any matching so that the contents of the current file don t matter The Panic Detect State is important so we ll spend a minute discussing it in detail The panic detect routine is called periodically by DISKATE including PRIOR TO THE OUTPUT OF EVERY BYTE TO THE TERMINAL In the version supplied with DISKATE which is written for the serial I O port on the disk controller the panic detect routine will respond to ANY KEY ON THE KEYBOARD BEING PRESSED Because there are timing considerations involved the best way to register the panic state to DISKATE if you re using the panic detect routine provided is to press the Break key if your terminal has one These details are discussed more fully in Part II If you are writing your own panic detect you may wish it to respond to only a certain character such as
199. rithmetic occurrencing ous te a we Bi eee DISKATE User s Manual 126 4 Lowest Default Interval Argument For any command except disk commands which take an Interval Argu ment where no argument is given the argument used will be the last interval computed Disk File Reference Symbols Disk Drive Designator a letter A through H A through D denote Disk Jockey Drives 1 through 4 E through G denote North Star Drives 1 through 3 H is currently unused Disk File Names Any sequence of up to 8 ASCII characters whose value is greater than hex 20 a blank except for l 4 2 z 3 as the first character of the file name 4 as the first character of the file name optionally followed immediately no intervening blank by then a Disk Drive Designator If no Disk Drive Desig nator is present it is assumed the file resides on the Current Drive In place of a file name the symbol denotes the most recently referenced file If a file name is prefixed immediately no intervening blank by the character a disk command referencing this file will first execute a PAUSE command and then execute the disk command lt R gt denotes the interval in the memory occupied by the most recently read record from the disk The lower address of lt R gt is given by the 16 bit value stored at RECAD and the upper address by the 16 bit value stored at RECND where RECAD and RECND are symbols from ATETBL
200. rive given as argument is list eds If no argument is given the directory on the current drive is listed A copy of the directory is left at the DISKATE User s Manual 136 1824 bytes beginning at DISKBUF where DISKBUF is a symbol in ATETBL FS lt optional Disk Drive Designator gt Free Space The amount of free space on the disk and on the directory for the drive given as argument is listed If no argument is given the listing is for the current drive U lt filename gt Unsave The disk file whose name is the argument is deleted from the directory T lt source filename gt lt destination filename gt Transfer The contents and directory attributes of the disk file whose name is the first argument are transferred to the file whose name is the second argument If there is no file whose name is the second argument it is created with compacting occurring as with the S command TD lt source Disk Drive Designator gt lt destination Disk Drive Desig gt Transfer Disk Each file on the disk designated by the first argument is transferred to the disk designated by the second argument with the same conventions as for the T command Files on the destination disk drive whose names are not in the directory on the source disk drive are unaffected If the destination disk drive or directory does not have enough room to complete the transfer for every file on the source disk drive the transfer will be carried out fo
201. rrence in the current file of JOAN WHEREVER that happens to be The command JOAN 3 would give an error message if the first occurrence of JOAN happened to be beyond the 3rd line in the current file Even though it may seem like more work at first to learn to specify intervals using expressions of the form n_ because these expressions can be used more widely than expressions like n it s a good idea to go to the extra effort There is another important way in which expressions of the form n differ from expressions like n_ We saw that the argument 1 refers to the last line in the current file the expression 2_ the next to the last line and so on In evaluating expressions of the form n DISKATE will TREAT ALL OCCURRENCE NUMBERS AS POSI TIVE If you specify n and there are not n lines in the file n will evaluate to the address of the 6 which forms the upper bound of the file in the computer s Memory So 2 will not give the next to the last line in the current file but will give the address of the at the end of the file as will 3 and so forth since 2 is the same thing as 65534 which is surely larger than the number of lines in the file This means that 1 is a convenient expression for the at the end of the current file We ll have occasion to need this Since DISKATE provides the facilities for referring to locations in the current file by line number in more than one way in fact you might
202. rs 18 39 69 O assembler error code 91 142 O Command 58 131 Obscure Jude the 38 Occurrencing 8 29 125 Octal numbers 53 ORG Pseudo 85 148 OUT 143 Output Routine Character 194 P Command 14 998 138 Pl 115 144 P2 115 144 Panic Detect Routine 194 Panic State 59 66 121 Pass 1 assembler 79 Pass 2 79 PAUSE Command 598 126 134 PHD 145 PHLDC 144 PHLSB 144 Power up Procdure 97 Priority of operations 125 Processor Technology Format 82 139 Prompt character 1 PRNTSYM example 93 PROS Command Pseudo 3 Q base suffix 53 122 Q Command 89 137 QF Command 39 132 QS Command 47 132 R Command 38 47 131 RAMEND 145 READER 143 RECAD 145 RECND 145 7 a ae REF Command 58 138 RENT Symbol in ATETBL 98 142 RENT Command 99 133 Reentry Point see RENT Return from macro 47 RN Command 75 136 ROMEND 116 144 E S in panic state 66 121 S Command 69 135 g Speed Constant 1863 SORG Pseudo 84 1490 Source Area 1l 41 45 48 58 115 Source File disk 68 STARTUP file 97 111 96 139 141 77 148 DISKATE User s Manual STCTR 144 Symbol Table 79 82 113 127 example of program to print 93 SYMTB 144 S 60 124 T Command 76 136 TAB Command 89 139 TABA 127 144 Target character 2 16 24 33 TD Command 76 136 Terminal Initialization Routine 65 183 T 79 124 in SORG operand 89 U Command 73 136 UCTAD 144 User Command Table 116 USRCT
203. rs to be SIGNED 15 BIT INTEGERS OCCUR That means that the high order bit is treated as a sign ENCE bit and negative numbers are represented in two s NUMBER complement notation On the other hand an address is MUST BE interpreted aS a non negative 16 bit number When you lt DISKATE User s Manual ll give DISKATE a number the question of whether it will 32767 be treated as a signed 15 bit number or a non negative 16 bit number depends on the context If you give DISKATE a negative number when it expects an address the sign bit will be treated as a digit bit and if you give a large number when it expects a signed number the highest order bit will automatically be treated as giv ing the sign bit This means that an occurrence number greater than 32767 will count as a negative number This shouldn t pose a problem since it s very unlikely you would ever need an occurrence number this large What if you want to see the last 29 lines in the file One way to do it would be by the command 26 19 This would literally give you 28 lines beginning with the 28th line counting back from the end of the file However there is an easier way The command below may throw you for a loop at first because so far we have presented the symbol as an operator which connects intervals into a larger interval In fact the symbol can itself be used as a pattern matching argument The command we just showed will give the same result as the
204. ry Once again always be sure the entry pointer is in oe right place before beginning to edit You can use the E command with numeric codes rather than text in editing source files also Once upon a time we promised to show a method for entering into a text file unbalanced brackets This can be done by using an E command with the numeric code for the bracket rather than using the E command with ASCII text The command E 133 if the current base is 8 E 5BF if the current base is 16 for instance will enter a left bracket You can also use a sequence of numeric codes enclosed in s with each number separated by a blank as part of arguments calling for matching The pattern numeric code sequence where numeric code sequence is a sequence of numeric codes in the current base separated by blanks will match a sequence of characters having the same value in the current base as given codes There are many ways this can be used Suppose for instance you have a piece of text which was created using another piece of software and each line concludes with a carriage return followed by a line feed To delete the line feeds you can use the command _ B16 R 1 QF D A gt K Here the pattern D A matches the characters carriage return line feed since we made sure the current base was 16 Thus D A gt matches a line feed at the end of a line and this command will get rid of all of them Of course such commands can be easily turn
205. se but it will position the entry pointer reliably to the end of the current file and is shorter and simpler than the com mand lt F gt gt l It s more straightforward than you can choose whichever is easier to remember Now we can get to the real goal of the present discussion how to write sequences of DISKATE commands which can be executed as edit macros The flexibil ity which this feature of DISKATE allows provides for a virtually unlimited range of applications We ll begin with examples which are very Simple One simple appli cation for edit macros which could arise very often is substituting for an abbreviation the text the abbrevia tion stands for If there is a long phrase that occurs over and over in your text for instance it would be nice to have an abbreviation for it and then use the DISKATE edit macro facility to replace the abbreviations with the expanded form To use an example that occurs frequently in this manual suppose we use the symbol I R I to stand for initial reference interval 1 but Without 44 F lt F gt ANOTHER SIMPLE COMMAND TO PUT ENTRY POINTER AT END OF CURRENT FILE DISKATE User s Manual using any edit macros we could instruct DISKATE to replace every occurrence of the I R I by the full text using the command R 1 QF I R I K E initial reference interval If we found ourselves wanting to issue this command frequently it s a long one to
206. sor and remove from your screen the offending character You ll then have to retype the rest of the line Or you can type ESCAPE ESC to tell DISKATE to ignore the entire line DISKATE will echo the ESC then a backslash then a carriage return and line feed If you ESC a line which begins a command DISKATE will expect a new com mand and will issue the prompt character If you ESC a line of an E command which is not the first line the line will be ignored and you will still be in text entry As you type characters to DISKATE major changes in the memory do not occur until DISKATE has received an entire line In the meantime what you have typed is placed in a special internal location called a line buffer The length of this buffer is 130 characters which means that DISKATE cannot process a longer line This should normally cause no problems since most terminals have a maximum line length of 88 or less It could show up however if you forgot to end a line with a carriage return and kept on going to the next line down on the screen When the DISKATE line buffer is full the CURSOR WON T MOVE If this happens while you re in the middle of the characters for an E command you should type BACKSPACE and then a right bracket and a Carriage return to terminate text entry to use other editing commands to place part of the line which is too long on the next line Or you can simply type a carriage return and keep on entering text The line
207. ss so we ll see in a minute how to avoid this One note Repetition numbers are treated as POSITIVE That means that 1 is the same thing as 65535 So if you want to use the large number method for the R command the simplest number to use for the number of repetitions is l since this is the largest repetition number DISKATE will allow 65535 is obviously large enough for any thing you would want to do since files must fit in the memory and 65K is the limit to the size of the memory The problem we just encountered can be phrased as follows How can we break a repetition loop when we ve run out of matches without terminating the loop with an error There is a DISKATE command especially for this purpose The command QF which stands for Quit on Failure will terminate a repeat loop without an error if in evaluating the argument of the QF command a match fails So we can do what we did above without having to end in an error oy the command Rl OF SEX Ja K SEX 18 D N N The R 1 will cause the rest of the command to be repeat ed 65535 times and the QF SEX command will terminate the loop without an error if there is nothing which matches SEX In fact we can make the command much But we don t really need to go to The simplest approach in giving a repeat command which you want to apply to every occurrence of a 30 IN Rn COMMAND n Is TREATED AS POSITIVE R 1 IS THE SAME AS R65535 QF COMMAND TE
208. subroutine at the lower address of the interval given as argument is called If it maintains DISKATE User s Manual 133 the integrity of the stack it may end with a RET statement and execution of the process which invoked the X command will continue lt variable gt lt Interval Argument gt The lower address of the argument is assigned to be the value of the variable If the variable is not present in the symbol table it is entered there as described above under Symbol Table There must not be a blank between the variable name and the sign COM lt prompt string gt Accept Command The string within brackets given as argument is printed on tne terminal as a prompt One command will be accepted and executed and then the process invoking the COM command will continue The command may be a multi line E command Or a multiple command command line The command line typed in response to the COM command may be ESC ed and retyped An error in the command will cause the process invoking the COM command to terminate and issue an error condition ATE lt prompt string gt invoke DISKATE recursively The string within brackets given as argument is printed on the terminal as a prompt DISKATE calls itself recursively and then will accept and execute any number of commands An error condition terminates only those processes invoked by commands given after the ATE command is executed Exe cution of DISKATE continues in t
209. t the string contains no commas blanks carriage returns or end of file characters s All characters from the up to the next comma blank carriage return or end of file are ignored allowing the string to serve as a label QF lt Interval Argument gt Quit on Failure Execution of a command string invoked by a D command or repetition invoked by an R command is terminated and a return if pending is executed provided the argument results in an Argument Failure where an Argument Failure is caused by either the failure to obtain a match for a matching argument or by an interval which results in the lower address being greater than the upper address An argument containing a syntax error or undefined variable will cause an actual error rather than an Argument Failure If an Argument Failure does not occur execution continues Only the innermost process is terminated An Argument Failure does not result in an error condition QS lt Interval Argument gt Quit on Success Execution of a command string invoked by a D command or repetition invoked by an R command is terminated and a return if pending is executed unless the argument results in an Argument Failure where an Argument Failure is defined as above If an argument failure occurs execution continues Only the innermost process is terminated An Argument Failure does not result in an error condition X lt Interval Argument gt Execute The machine language
210. th Star disk then DISKATE can be brought up with the DOS command GO IO There is one very important point here if you go to create another copy of DISKATE on another disk IO ASSUMES THAT ATE IS THE SECOND ENTRY IN THE DIRECTORY If you use DISKATE to create another copy of itself you must make sure that this holds When IO has loaded ATE it will transfer control to it Initialization then takes place This includes execursnd the command orn es 0 2 Y The last thing DISKATE does at power up is to execute the command GO STARTUP STARTUP as provided is an ordinary source file To see what it contains in your version type O L STARTUP The STARTUP file can be used to initialize your own variables which you want to have a value every time you use DISKATE However sinc the GO command will either L and Da source file or L and X a non source file the STARTUP file can also be a machine language program If you create a new copy of DISKATE you might happen to forget the STARTUP file This is not serious In this case DISKATE will issue an error message that it can t find STARTUP and then take commands from the terminal as usual You can edit STARTUP just as you 97 TO POWER UP LOAD amp EXECUTE IO VERSION PROVIDED BOOT STRAPS IO ATE MUST BE 2ND ENTRY IN DIREC TORY ON POWER UP COMMAND GO STARTUP IS EXECUTED STARTUP PERFORMS USER S INITIAL IZATION ENTRY POINT TO INITIAL
211. the operating system So the protocol for the disk drivers will be given even though in many cases you won t need to use it There is one important way in which YOU MUST CUS TOMIZE THE DISK DRIVERS The version supplied assumes that there actually exists a disk drive for all of the possible disk drive designators You should alter the jump table to indicate that a drive is not present in your system so that if you try to access a file ona non existent drive an error will be returned The format for that part of the jump table for the disk drivers is Location Contents IO 15 MVI C lt code for drive A gt MUST JMP lt driver for drive A gt SUBSTI TUTE I0 26 MVI C lt code for drive B gt JMP lt driver for drive B gt STC RET LO 25 MVI C lt code for drive C gt JMP lt driver for drive C gt FOR THE MVI IO0 39 MVI C lt code for drive D gt JMP JMP lt driver for drive D gt FOR EACH I0 35 MVI C lt code for drive E gt DRIVE JMP lt driver for drive E gt THAT DOES NOT I0 4G MVI C lt code for drive F gt EXIST f a JMP lt driver for drive F gt T0 45 MVI C lt code for drive G gt i JMP lt driver for drive G gt I0 59 STC piee RAGE On ee The pair of instructions STC RET at I0 5 is in lieu of a driver for drive H All of the disk drivers return with the carry flag set in case of error When you reassemble SYSIO then you should substitute a STC RET for the MVI JMP for each drive that does not ex
212. the command was part of a larger command string such as a macro or multiple command command line the command string will abort in the same way as for any other DISKATE error The ERROR meaning of the error codes is as follows CODES A Argument Error The operand field is invalid for the given opcode This will also occur at a statement lbl after an IF pseudo of the form IF expression 1lbl if the expression refers to an undefined variable The A code usually occurs only in Pass 2 though an invalid EQU operand for instance will cause it to occur in Pass l M Missing Label An EQU statement occurs without a label Printed during both passes D Doubly Defined Label A label is given where the label is identical to a symbol already defined Failing to give a Z or Z gt command after a previous assembly will usually generate quite a few of these L Label Error The first character in a label field is not an upper case letter of the alphabet When this occurs 3 NOPs will be generated in place of the statement This error can occur especially in Processor Technology format if there are extraneous characters in the source code O Opcode Error The opcode field is not a proper op code or pseudo op 3 NOPs will be generated in place of the statement To conclude this section let s look at a simple example of an assembler program which performs a useful function We ve already seen that DISKATE symbol tables are k
213. the target charac ter Again we can give the command to see the line sois _ PATTERN 2 MATCHES PATTERN AT BEGIN NING OF LINE A COMMAND SETS TARGET CHAR TO FIRST CHAR OF INTERVAL GIVEN AS ARGUMENT DISKATE User s Manual 25 with the target character to check to see whether we ve got the right spot Note that because the command does not take an argument the symbol gt has the same value after the command as before it What command shall we use once we have the right JOAN There are several different commands we could use that would all have the desired effect As before K JOAN A would do Or since we know that the target character is the J in the JOAN we want we could use the command A IN i ARGUMENT In this command we see still another use of the symbol The command is a Kill command so is MATCHES not used here as a command Rather is used here as TARGET part of an argument The symbol used as an argument CHAR gives AN INTERVAL BOTH OF WHOSE ADDRESSES ARE THE VALUE OF THE ENTRY POINTER Or you can think of as a SYMBOL WHICH MATCHES THE TARGET CHARACTER So the argument in the command above translated literally means the first A within everything in the current file starting with the target character Since we already arranged for the target character to be the J in the JOAN from which we want to delete an A this command will work In fact there is a
214. ti tuted for in this example The symbol lt F gt is an argument which calls for no matching and denotes the current file The command we just Saw can be replaced by the command lt F gt gt 41 There is an important difference e does not really specify the current file but rather the initial saaTs ARG1 ARG2 WITH ARG2 NOT INSIDE ARG WILL WORK IF ARG2 HAS NO MATCHING lt F gt DENOTES CURRENT FILE NO MATCHING WHATEVER INIT REF INTERVAL lt F gt gt 1 DISKATE User s Manual reference interval as we have discussed There is a special DISKATE command which we will discuss below which can be used to manipulate the initial reference interval thus changing the meaning of but lt F gt always refers to the current file no matter what the initial reference interval is The command that we ve just seen lt F gt gt l is probably the command that comes the closest to specify ing the end of the file within the DISKATE language ina The argument lt F gt gt 1 may look con straightforward way fusing but it says the end of the file quite direct ly However there is a simpler command which will work by a Side effect We recall from a previous discussion that 1 will also specify the end of the current file The simplest command for setting the entry pointer to the end of the file which is reliable is probably the command l If you use this command
215. to IO be positioned in the memory completely independently The beginning of IO is a JUMP TABLE to all of the I O drivers DISKATE will need The initial power up procedure if your terminal system is not connected to the Disk Jocky serial port is to bootstrap as if you were and then to patch into this jump table We ll discuss the first part of the jump table first because the first part deals with the terminal I O and the i second part deals with the disk drivers Here is the format for the ee elie of IO the 10 12 Location wn Contents 1043 JMP Character Output Routine I0 6 JMP Character Input Routine I0 9 JMP Terminal Initialization Routine JMP Panic Detect Routine In the standard version Sarcently provided 0100H IO begins at Mi A jump to the oadet e the jump table Thus DISKATE can be reloaded at any time by jumping to IO This Thus ATE and IO can 181 BAUD RATE FOR SERIAL TERMINAL ON D J IS 1200 DISKATE READY TO RUN WITH SERIAL TERMINAL CONNEC TED TO D J ATE MUST BE 2ND FILE IN DIREC TORY IO BEGINS WITH JUMP TABLE 1ST JUMP TO LOADER THEN TERMINAL I O DRIVERS DISKATE User s Manual Since a jump to IO is written at memory location 3 the easiest way to jump to IO to reload DISKATE is simply to jump to memory location 3 The general procedure for getting started with a terminal either not connected to the Disk Jockey or a serial t
216. to the source area But what if you want the whole source area as it already exists to be moved somewhere else The DISKATE M command which we ve already introduced is actually an extremely flex ible and powerful command and can very easily accomplish moving whole blocks of memory not only moving the actual bytes but also changing the internal DISKATE pointers so the change will be properly kept track of Suppose for instance that your source area begins at 2AG H and you have done some work editing a file Now you decide you want to consult a BASIC program which needs this area of memory Of course your file could be saved on the disk and then reloaded but perhaps the BASIC program is able to examine the file and you want the file still to be in memory for speed reasons Let s say that a block beginning at 7090H is free and big enough to hold the file The single command 7000H M lt S gt will move the entire source area to 7 9 H and update the pointers delimiting both lt S gt and lt F gt The M command is another which behaves differently if the entry pointer lies outside the source area which here it certainly does In such a case the interval to be moved is NOT deleted from anything but is simply COPIED to the area of memory beginning with the target character There is no insertion whatever used to be at the interval to which the argument is copter is simply overwritten 00000 Ue There is a special case of usi
217. u could get unintended results You can have multiple arguments as long as they are separated by blanks just call CVALS or VALUS repeatedly Suppose you wanted a machine language program to follow the same protocol as DISKATE commands and use the last interval computed in case the argument is missing How do you obtain this interval Such information as this can be obtained by reading the contents of the relevant internal DISKATE registers Again this is done by using symbols from ATETBL In this particular exam ple the lower address of the last interval computed is stored at the address denoted by the symbol Pl and the upper address at P2 As it happens this example is not really too useful since the CVALS subroutine will automatically give you the addresses for the previous _ interval computed if it detects that there is no argument If you take a minute to skim through the listing of the meaning of the symbols in ATETBL at the end of Part III you will see that you have complete access to all of the information available to DISKATE commands The To evaluate further arguments 115 VALUS WILL EVALUATE ARG amp RETURN Z FLAG OFF IF ERROR MUST GIVE INIT REF INTERVAL IN HL DE ARGS MUST BE PRECEDED BY BLANK GET ADDRS OF DISKATE INTERNAL REGI STERS FROM ATETBL E G lt AT Pl gt AT P2 lt S gt AT BOSAP SEOBE lt F gt AT somewhere else DISKATE User s Manual
218. uage subroutine is properly written and exits with one of the 8889 return instructions a call to it by an X command can be included in a command line or macro and the rest of the command line or macro will be executed upon return the address given is a program which goes into its own loop and does not return such as a BASIC or DOS then the X command will serve as a jump command since such a program will certainly intialize its own stack The details of how to write machine language programs for interfacing to DISKATE are discussed in Part II Here s an example of using the X command If you re using DISKATE with software that locates a DOS entry point at 2028H you can get back to the DOS from DISKATE with the command l X29284 _ If you want to do this frequently you can set a varia ble say DOS equal to 2928H and then use the command xDOS This method is highly recommended since as with any jump command if you make a mistake typing an explicit address for the X command a system crash could result ie apja X COMMAND ISSUES CALL TO MACHINE LANGUAGE SUBROU TINE AT LOWER ADDRESS OF INTERVAL GIVEN AS ARGUMENT DISKATE User s Manual 64 1 3 Using DISKATE input output In this section we ll deal first with terminal I O handling and then talk about how to use the disk com mands All of the I O in DISKATE is handled through calls to locations in a jump table so that any type of terminal
219. ubmacro which ME ag a GIVES CHAR OF MACRO THAT CAUSED AN ERROR R COMMAND IN MACRO REPEATS TO END OF FILE OR QS QF CAUSING QUIT QS COMMAND QUITS INNER MOST PROCESS IF NO MATCHING FAILURE IN ARGUMENT QS WITH NO ARG ACTS AS RETURN G COMMAND TRANS DISKATE User s Manual 48 PARAGRAPH will call The macro will contain a new FERS command the G command which stands for Goto The G EXECU command causes execution of a DISKATE macro to be TION OF transferred to the lower address of the argument and MACRO works just like a GOTO in BASIC Like the D command TO the initial reference interval for the G command is the LOWER whole source area And now for the macro ADDRESS OF ARG FUSE QFX Y _ K E G FUSE INITIAL CHOP REF QFX Y Z K 1 E_ G CHOP INTERVAL PARAGRAPH IS D FUSE SOURCE x AREA D CHOP Qs A few notes The final QS can be done away with if this macro ends the file If as shown the entire macro is going to go in one file the lines beginning with PARAGRAPH must NOT begin the file The reason is that if they do the pattern FUSE will match not the name of the submacro as we want it to but rather the characters FUSE within the D FUSE This could cause all kinds of problems Note that in the macro CHOP the K followed by the E command will leave the target character as the character following the carriage return just entered setting it proper
220. umber is 350 The maximum number of blocks DISKATE can handle is given by the constant MAXBS which is a symbol in ATETBL The use of ATETBL is discussed in the next section If this command returns with a value larger than MAXBS a disk error will result Each block is 256 bytes long regardless of how the controller Phys leary organizes disk _ records BEls A read E A contains the number of blocks DE contains the beginning memory address to which the information is to be transferred and HL contains the beginning disk address in blocks The directory occupies blocks 6 3 The drivers supplied for the North Star Disk invoke only the North Star DOS routine DCOM and not DLOOK B A write operation The registers A DE and HL follow the Same convention as for the read opera _ tion T 187 DISKATE User s Manual 198 For all disk commands a return with the carry flag set is the indication of a disk error If you should want to rewrite the disk drivers for your own applications all DISKATE disk addressing is relative to your drivers Thus if your drivers examine only a certain portion of the disk leaving the rest beyond the reach of DISKATE DISKATE s block n refers to the n th 256 byte block within the area of disk you ve set aside The same would apply for drivers that work through an operating system s direct access file routines In most cases however you will probably not need to touch th
221. ussed in Part II At this point we are ready to begin discussing the disk file handling commands Internally DISKATE is divided into two parts a program called IO which has all of the disk handling routines and other input output and a program called ATE which is independent of any particu lar I O requirements DISKATE comes with drivers for handling both the Disk Jockey controller and the North Star Disk However if you have another controller it is possible to interface that controller to DISKATE by rewriting part of the module IO A disk file is named by a combination of the file name and a designator which indicates which drive the file is on DISKATE maintains a CURRENT DRIVE and the disk drive designator can be omitted when specifying a file in which case the file will be assumed to be on the current drive File names may be up to 8 characters long and can contain ANY PRINTING CHARACTERS EXCEPT a comma or colon In addition the file name must not begin with or The disk drive designator is a single letter A through H In the current version drives A through D are vectored to routines for the Disk 26 7 GIVE ADDRESS IN THE FILE WHERE PRINTOUT FROM COMMAND WAS ABORTED BACK SPACE WIPES OUT LAST CHAR TYPED ESC WIPES OUT CURRENT LINE DISK FILE SPECI FIED BY FILENAME COLON DRIVE DRIVE OMITTED ASSUMES CURRENT DRIVE DRIVES DISKATE User s Manual Jockey
222. val given as argument If the interval is within the source area the interval is deleted from the file in which it occurs and moved to the area of memory immediately following the Source area by an internal M command The deleted text is not surrounded by J s Immediately after a K command the Default Interval Argu ment will be the interval occupied by the deleted text outside the source area The target character will be the first character in the source file after the material that was deleted If the interval is not within the source area the inter val is zeroed and the target character becomes the first character in the interval M lt Interval Argument gt Move Moves the interval given as argument to the address given by the entry pointer If the entry pointer is within the source area then the interval is inserted into the file containing the entry pointer between the target character and the preceding character Otherwise the characters of the interval are copied to overwrite the area of memory beginning with the target character The target character becomes the character after the new location of the last Character moved Internal PointTerg to the following are updated by the M command S gt lt F gt lt R gt lt 2 Ss T lt gt and hence the default interval argument User Command Table Macro Command Interpretation Pointer Return Adresses Repeat Addresses A pointer is updated if the address it gi
223. value is the string A BC Second the character hex 5E WILL NOT BE GENERATED into the string constant but WILL CAUSE THE HIGH ORDER BIT OF THE PREVIOUS BYTE TO BE SET This can be used to define a symbol table in the same format DISKATE uses for the symbol table it generates itself For instance to define a short symbol table called SHORTBL with entries for program labels ALHPA BETA and GAMMA in Intel format you could use the statements SHORTBL ASC ALPHA DW ALPHA ASC BETA DW BETA ASC GAMMA DW GAMMA DB There is a special pseudo IF which allows for conditional assembly The form of the IF statement not counting the label field is OIF expression label The operand thus consists of expression and label with a comma between BUT NO BLANKS A blank would prematurely terminate the operand field DISKATE evaluates the expression and if it is not the assembly continues as it normally would However IF THE EXPRESSION HAS 87 PSEUDO ASC DEFINES STRING GIVEN BY OPERAND OPERAND DELIM ITED BY BLANKS ASCX WITH x ANY NON ALPHANU MERIC REPLACES x WITH BLANK IN STRING GENER ATED IN OPERAND NOT GENER ATED SETS HIGH ORDER BIT IN PREVIOUS CHAR IF PSEUDO EXPRES SION COMMA LABEL SKIPS TO LABEL IF EXPRES DISKATE User s Manual THE VALUE THEN THE ASSEMBLY SKIPS AHEAD TO THE STATEMENT WITH LABEL label You can use this to have sev
224. ves is within the interval given by the argument before the move takes DISKATE User s Manual 129 place There is a pointer for both the lower and upper address for each of lt S gt lt F gt lt R gt and lt T gt and if only one such pointer lies within the argument because the argument overlaps only the one pointer will be updated This may cause an erroneous value lt Interval Argument gt set Entry Pointer The entry pointer is set to the lower address of the argument C lt Interval Argument gt Copy The characters of the interval are copied to the area of memory beginning with the target character No pointers are updated and the action of the command is not affectd by whether the entry pointer is within the source area B lt Arithemetic Expression gt Base The low order byte of the value of the expression becomes the current base It may be any number gt 4 For bases gt 19 ASCII characters in order beginning with A are used for digits Numbers are printed in split form for all bases other than 16 lt Interval Argument gt Quote numbers core dump The numefical value in the current base for the bytes of the interval are printed on the terminal with the address of the first byte of a line printed at the left of the MH 1 ee eae ge a lt Interval Argument gt Where 3 Prints on the terminal in the current base the upper and lower address of the interval given as argument
225. we haven t seen before The V command V for eValuate will put the lower address of the argument in HL and the upper address in DE This command will not achieve anything if given from within DISKATE itself Since the information in HL and DE will be overwritten by subsequent internal routines However if the V com mand is the last command in a string of commands execu ted by ATECOMS on returning to the caller HL and DE For instance Suppose you wanted the value of the DISKATE variable X The machine language source statements LXI H COMMAND RST LABEL e e o e COMMAND ASC VX DB 8 as part of a machine fenguage program would cause the value of X to be in both HL and DE after the return to the statement LABEL The commands in a command String executed by ATECOMS can certainly include disk commands You can even use this method to link to the value of symbols in ATETBL during the execution of a machine language pro gram rather than during assembly by having a command string include L ATETBL lt T gt KEEP or the like and then a V command for a symbol in ATETBL With this flexibility virtually any type of interface needed between DISKATE and a machine language program can be constructed 119 V COMMAND PUTS LOWER ADDR OF ARG IN HL UPPER IN DE ONLY OF USE IN COMMAND STRING TO BE EXECUTED BY ATECOMS DISKATE User s Manual If a machine language program needs access to the information in the
226. wonder if there is a way for DISKATE to provide you with the numbers of the lines in the current file There is indeed a way but well _ postpone a detailed discussion until section I 4 because it is keyed to the assembler To jump the gun a bit the P command is similar to the command and will print _the interval given as argument with line numbers How ever if you give this command with a file consisting of ordinary prose text THE TEXT WILL COME OUT IN A PECUL aie n TREATS n POSITIVE IF TOO BIG GIVES ADDRESS OF AT END OF FILE 1 GIVES AT END OF FILE P COMMAND PRINTS INTERVAL WITH LINE NUMBERS IN DISKATE User s Manual 15 IAR FORMAT with all kinds of spaces where they shouldn t ASSEMBLER be The P command is designed to print assembler source FORMAT code and the spacing is so that the listing will con form to a convenient assembler format If you re able MANY to recognize lines properly in spite of this format you EXTRA can use the P command to obtain line numbers for use in SPACES expressions like n and n The use of line numbers obtained in this way is a very good way of making sure that an occurrence of a pattern is in the right place There is a way to make the P command easier to use for editing prose by using a command which well discuss in more detail later The command TAB 6 8 8 will suppress most of the format pecularities It can still make your text look sl
227. xt Editor has a variety of editing commands which can work with any type of text The text is entered into and maintained in an area of the computer s memory called the SOURCE AREA The Source Area is divided into one or more units called FILES A file is simply a segqeunce of bytes in the memory bounded by s Most of the editor s commands work with a file which is established as the CURRENT FILE If this file is expanded or contracted the source area is expanded or contracted along with it to preserve all of the files in the source area There are special commands which serve to make a different file the current file Actually you can have several different source areas in the computer s memory but more about that later To work with files that reside on the disk you will need to move them in and out of memory using the DISKATE I O Soumandsy which are discussed in I 3 When you bring up DISKATE it will issue gt as a prompt character and wait for a command There may be blanks between the name of the command and an argument if the command takes one and blanks MUST be present between arguments if the command takes more than one Usually in this manual we will omit unnecessary blanks though in your own work you can include them if it will make the DISKATE commands more intelligible A blank should never occur within an argument unless that blank is a significant character as part of the argument Eo There are two ke
228. y concepts which you will need to be familiar with to understand how the DISKATE commands work The entry pointer and intervals The ENTRY POINTER isa special 16 bit location pointing to a mem ory location which serves as a kind of target for where an action performed by many DISKATE commands is to take lt place For instance let s consider one of the simplest DISKATE commands which serves to enter text into a file This command has the forme E text The E stands for Enter Note that if this is a command you are typing in all by itself the prompt SOURCE AREA FILES CURRENT FILE gt IS PROMPT CHAR ENTRY POINTER OE COMMAND DISKATE User s Manual mo character gt will be the first character on the line For instance to enter the string I have a weak spot for computers the line on your terminal should look like gt E I have a weak spot for computers To avoid confusion the prompt character will generally not be shown since it is not one that you type As you type in the characters of the text one by one here s what DISKATE will do Everything in the current file and in fact the whole source area start ing with the location pointed to by the entry pointer will in effect be pushed up in memory by one byte and TARGET the character typed will be entered into the computers CHAR memory at the location pointed to by the entry pointer The entry pointer is then incremented

Disk/ATE™ User`s Manual Manual Rev. 1.1 By Jim

Contents

Download Pdf Manuals

Related Search

Related Contents

Disk/ATE&trade; User`s Manual Manual Rev. 1.1 By Jim

Contents

Download Pdf Manuals

Related Search

Related Contents

Disk/ATE™ User`s Manual Manual Rev. 1.1 By Jim