Home

AMOS AlphaBasic (old)

1. MAP statements should normally come at the beginning of the program before any references to the variables being mapped If a reference is made to the variable before it is mapped such as LET A 5 8 the variable will be assigned by the normal variable allocation routines and the MAP statement will then give an error since the variable is already defined As a convenience all MAP statements force allocation to the next even byte boundary so that binary word data can be assigned properly if desired Due to the complexity of the syntax checking used for the MAP statement no syntax analysis is performed on them until the program is compiled unless the MAP statement is used in the immediate mode If you wish to check the syntax of your progress during the entering of a group of MAP statements you may force compilation and syntax checking of your partial program with the COMPILE command This will indicate your progress so far and also define all mapped variables up to this point so that you may interrogate them with the command described in the following section LOCATING VARIABLES DURING DEBUGGING Since the mapping scheme is new and fairly complex to understand fully a command has been implemented which will assist you in locating the mapped variables and in understanding the allocation techniques used by the AlphaBasic memory mapping system It is valid only as a system command and has no meaning if used within a program text The c
2. multiplication division raise to power kk raise to power string literal NOT logical not AND logical and OR logical or XOR logical xor EQV logical equivalence MIN minimum value MAX maximum value equal lt less than gt greater than lt gt unequal gt lt unequal unequal less than or equal less than or equal greater than or equal greater than or equal OPERATOR PRECEDENCE The precedence of operators determines the sequence in which mathematical operations are performed when evaluating an expression that does not have overriding parenthesis to dictate hierarchies AlphaBasic uses the following operator precedence exponentiation unary plus and minus multiplication and division addition and subtraction relational operations comparisons logical AND OR XOR EQV MIN MAX logical NOT MODE INDEPENDENCE Expressions may contain any mixture of variable types and constants in any arrangement AlphaBasic performs automatic string and numeric conversions as 13 necessary to insure the result is in the proper format For example if two strings are multiplied together they will first be automatically conver ted to E numeric format before the multiplication takes place If the result is then to a y become a string it will be reconverted back to string format before the assignment is performed In other words the statement A B 345 is perfectly legal and will work correctly Th
3. the the the the the the constant e 2 71828 raised to the power X natural base e logarithm of the argument X decimal base 10 logarithm of the argument X square root of the argument x largest integer less than or equal to the argument X integer part of X fractional part truncated factorial of Xx absolute value of the argument x a value of 1 0 or 1 depending on the sign of the argument X Gives l1 negative 0 if X is 0 and 1 if x is positive 37 RND X Returns a random number generated by a pseudo random number generator based on a previous value known as the seed The argument X controls the number to be returned If X is negative it will be used as the seed to start a new sequence of numbers If X is zero or positive the next number in the sequence will be returned depending on the current value of the seed this is the normal mode The RANDOMIZE statement may be used to create a seed which is truly random and not based on a fixed beginning value set by the system MEM X Returns a positive integer value which specifies the number of bytes currently in use for various memory areas used by the compiler system The most common use of this is with an argument of 0 which returns the number of free bytes left in the user memory partition This MEM 0 call duplicates the action performed by the FRE X function in other versions of BASIC Other values of the argument X return memory a
4. 1 unformatted area called ARRAY which is subdivided into two level 2 items Statement 110 defines the first of these which is an area called UNITX The optional dimension indicates that 5 of these identical areas exist which must be referenced in the program by the subscripted variable names UNITX 1 thru UNITX 5 Each one of these areas is then further subdivided into five level 3 items statements 120 160 Since the level 2 is subscripted because it occurs 5 times so must each of the level 3 items be subscripted There are 5 variables named SIZA 1 thru SIZA 5 occuring once in each of the respective variables UNITX 1 thru UNITX 5 The same holds true for the variables SIZB NTOT and CNAME Statement 150 however creates a special case since it contains a dimension also Normally this would create an area of 10 sequential bytes referenced as FLAG 1 thru FLAG 10 In our example however this 10 byte area will occur once in each of the higher level areas of UNITX 1 thru UNITX 5 This implicitly then defines a double subscripted variable ranging from FLAG 1 1 thru FLAG 5 10 Statement 170 causes the allocation to return to level 2 where one floating point variable is allocated The total storage requirement for the level 1 variable ARRAY1 comes out to 206 bytes as follows 40 bytes for each of the five areas UNITX 1 thru UNITX 5 plus 6 bytes for the one variable TOTAL Notice that since TOTAL starts a new level 2 it does not occur 5 time
5. 39 STR X or STRS X Returns a string which is the character representation of the numeric expression ae X No leading space is returned for positive numbers i Ga VAL AS Returns the numeric value of the string expression AS converted under normal BASIC format rules SPACE X or SPACES X Returns a string of X spaces in length TRIG FUNCTIONS The following trig functions are implemented in full li digit accuracy SIN X Sine of X COS X Cosine of X TAN X Tangent of X ATN X Arctangent of X ASN X Arcsine of X ACS X Arccosine of X DAIN X Y Double arctangent of X Y i 40 FORMATTED OUTPUT VIA PRINT USING STATEMENTS The PRINT USING statement is an extension of the standard PRINT statement which allows the output to be formatted into specific character positions suitable for business reports formal text applications and the like The format of the statement appears as follows PRINT USING lt string gt lt list gt Output to terminal PRINT lt file gt USING lt string gt lt list gt output to file The string expression is used to control the formatting of the variables as they are enrountered in the print list and must match the format of the variables to be printed The string may be either a string constant a string variable or a string expression which is interpreted as an exact image of the line to be printed The list is the sequence of variables or expressions to be printed using a
6. AlphaBasic If a previous version of the program same name already exists on the disk in the current user s file area that program will first be deleted before the new program is saved No backup file will be automatically created In actual practice the program name may be a full system file specification if desired The SAVE command may also be used to save the compiled object program on disk for later running without recanpilation To save the object program the user enters the program name followed by the explicit extension RUN which causes the program to be campiled if necessary and then the object program to be saved on disk The following two examples will show how the SAVE command is used to save first the source program and then the object program SAVE PAYROL saves source as PAYROL BAS SAVE PAYROLL RUN saves object as PAYROL RUN LOAD The specified program whose name must follow the LOAD command is located on disk in the current user s file area and then loaded into memory for editing or execution This is the reverse of the save function above The program is expected to be in ASCII format with the default extension BAS unless explicitly entered as otherwise If the program cannot be located an error message will result In actual practice the program name may be a full file specifier if desired The load command does not clear the text buffer before it loads the requested file and therefore may be used to concatena
7. OPEN OR OUTPUT PRINT RANDOM RANDOMIZE READ REM RESTORE RESUME RETURN RIGHT RIGHTS RND RUN SAVE SCALE SGN SIGNIFICANCE SIN SPACE SPACES SOR STEP STOP STR STRS STRSIZ July 1979 input output ISAM control kill file lower case string left string left string length string variable assignment line list text load program natural logarithm base 10 logarithm lookup file map variable maximum value memory size mid string mid string minimum value new program loop termination expand mode off logical complement on goto gosub error open file logical OR output print on terminal file random randomize RND function read data remark line restore data resume after error subroutine exit right string right string random number run program save program set scale factor sign set significance sine spaces spaces square root step stop program numeric to string conversion numeric to string conversion set string size ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 13 SUB sub gosub TAB tab TAN tangent THEN optional statement verb TIME system time TO to UCS upper case string USING using VAL string to numeric conversion WORD memory word WRITE write file XCALL external subroutine call XOR logical XOR July 1979
8. String variables are manipulated in the same Manner as in other versions of BASIC and may be concatenated by use of the plus sign between two strings String variables may be assigned values by enclosing string literals in quotes String functions such as LEFTS RIGHTS MIDS etc are implemented to assist in manipulating portions of strings or substrings In addition a powerful substring modifier may be used to operate on portions of strings within expressions A seperate section is devoted to this unique option of AlphaBasic Unformatted variables are also considered string variables when they are used in expressions or printed Be careful with this one If the unformatted variable is mapped to contain subfields which are not in string format it will cause some very strange results when printed or used in expressions This is because no conversions are performed on the subfields of an unformatted variable it is Fan used in its entirety exactly as it appears in memory LF VARIABLE ARRAYS Arrays may be numeric or string variables and are allocated dynamically during execution when the DIM statement is encountered in the program During execution if no DIM statement has been encountered when the first reference to the array is made a default array size of 10 elements for each subscript level is used This means that all DIM statements must be executed in the program prior to any actual references to the array Arrays may be any number of leve
9. This area is not preserved by the runtime system and the subroutine must not count on its security between XCALL statements Note that the word at R4 may be used as the first word but the word at R5 is the base of the arithmetic stack and must not be destroyed The last word of free memory is actually at 2 R5 for all practical purposes The runtime system has its own internal Memory management system and does not conform to the AMOS operating system memory Management Therefore the external Subroutine must not use the GETMEM monitor calls to generate a block of work space in memory Also if any file calls are to be done they must be done with internal buffers since the INIT call sets up a buffer by using the GETMEM Monitor call SUBROUTINE LOADING The version 3 1 release requires that the subroutines being called by the BASIC program already exist either in system memory or user memory before beginning execution of the program These subroutines must be loaded into system memory by the SYSTEM call in the SYSTEM INI file or they must be loaded individually into the user memory partition with the LOAD command while in monitor command mode Note that this is the monitor LOAD command to load the SBR module which is vastly different from the LOAD command used to load a basic source file once you are running AlphaBasic Future releases will provide capabilities for loading subroutines and overlays from within the BASIC program itself 53 E
10. X Y arguments will be treated as row column coordinates for positioning the cursor on the terminal screen The following characters will then be printed beginning in that position As in other functions the X and Y arguments may be expressions Terminals are assumed to begin with row 0 top of screen and column 0 left end of each row If the value of X is negative the function is interpreted as a special terminal command and the command code must be specified as the Y argument The codes will be transmitted to the terminal driver TDV module in 1 6 which will do the actual interpretation and perform the special function The following list will give the standard codes in use for the ADM3 SOROC and Hazeltime CRT terminal drivers Q Function Clear screen Cursor home upper left corner Cursor return colum 0 without line feed Cursor up one row Cursor down one row Cursor left one column Cursor right one column Lock keyboard Unlock keyboard Erase to end of line Erase to end of screen Protect field reduced intensity Unprotect field normal intensity Enable protected fields Disable protected fields Delete line Insert line h SCeOOWIHU amp WNHHO H H H hH p O U e Ww H The actual routines that perfom the screen controls are in the specific terminal drivers and not in AlphaBasic itself Therefore if you have a different terminal and you write a driver to perform the above functions for the terminal in use it will o
11. data to or from a file Once a 47 file has been closed no further references are allowed to that file number until another OPEN statement is executed Any files that are still open when the program exits will be closed automatically The format of the CLOSE Statement is i ZA 3 CLOSE lt file gt KILL The KILL statement erases one file from the disk It does not need a file number and no open or close need be performed to KILL a file The format for the KILL statement is KILL lt filename gt As in the OPEN statement the filename is any string expression which evaluates to a legal file description LOOKUP The LOOKUP statement looks for a file on the disk and returns a flag which tells you if the file was found and if SO how many records it contains The format for the statement is LOOKUP lt filename gt lt result variable gt As in the OPEN statement the filename is any string expression which evaluates to a legal file description The result variable is any legal numeric variable which will receive the result of the search If the file was not found a zero Will be returned If the file was found and is a sequential file a positive number will be returned which is the number of records in the file If the file was found and is a random contiguous file a negative number will be returned which is the number of records in the file In either case the number of records represents physical 512 byte disk recor
12. digit in the range of 30 to 30 and may not be a variable since scaling is done at compile time for constant values as well as at runtime for input and output conversions Negative scaling moves 43 the 1l digit window to the left and for most cases will not be of use to the average programer A few words of caution are in order here Once the SCALE statement has been detected during compilation all constant values that follow are scaled by the Scaling factor so that they are stored Properly In addition a runtime command is generated in the executable program which causes the actual scaling to be performed on input and print yalues when the program is running If two or more different SCALE statements are executed in the same program some very strange results may come out unless the user is totally familiar with what is happening in regards to compile time and run time conversions We suggest that you play with this one a bit before delving into it full steam One other word of caution Floating point numbers that are stored in files by the sequential output PRINT statement will be unscaled and output in ASCII with no problems Floating point numbers that are written to random access files by using the WRITE statement will not be unscaled first and any program that reads this file as input had better be Operating in the same scaling mode or else apply the scale factor explicitly to all values from the file Binary and string values of cour
13. distinct areas in use for variables and thus the offsets of the variables will be to one of these two areas All FIXED and MAP1 through MAP16 variables are allocated in the fixed storage area while all DYNAMIC arrays are allocated in the dynamic array storage area As dynamic arrays are dimensioned and redimensioned their position may shift around relative to one another and relative to the dynamic storage area base Variables in the fixed storage area will never change position relative to each other or to the storage area base Array location information is given only pertinent to the base of the array itself which is the location of the first element within the array The actual range of locations used by the array may or may not be contiguous in memory depending on whether overlapped dimensioning techniques are being used in the MAP statements Simple non array variables are defined as a location range which tells exactly where the entire variable lies within the storage area If you want to find out where a particular element of an array is located you may follow the variable name by the particular subscript values decimal numbers only of the element you wish to locate given the two commands A A 4 12 25 The first command will give information about the array A while the second command will define the exact location of element A 4 12 within the array A o Keep in mind that this command is to assist you in following the all
14. either a CONT command or a single step command The breakpoint remains set after it has been reached until explicitly cleared by another BREAK or COMPILE command DELETE The DELETE command is used to delete groups of source lines from the program text If the command is followed by a single line number only that line will be deleted If the command is followed by two line numbers separated by a comma space or other non numeric character all lines of text which fall between the two line numbers inclusive will be deleted from the text BYE This says goodbye to basic and returns the user terminal to monitor command mode Any program left in memory is lost forever so you may want to save it first using the SAVE command 30 PROGRAM STATEMENTS The source program contains statements which are executed in sequence one at a time as they are encountered Each of these statements normally starts with a verb followed by optional variables or statements modifiers This section will list the program statements that are supported by AlphaBasic and give some examples where necessary for clarity LET Assigns a calculated value to a specific variable during execution of the program It is unique in that the actual word LET may be omitted if desired LET AS 12 4 LET SUM 4 5 Al SOR B1 LET C JANUARY AS 12 4 SUM 4 5 A1 SOR B1 C JANUARY GOTO The GOTO statement transfers execution of the program to a new st
15. is named in the file spec The program may also optionally be resident in user or system memory if desired Some examples of legal statements are CHAIN PAYROLL CHAIN PAYROL RUN CHAIN DSK1 PAYROL 101 13 ees Due to the fact that programs are compiled and not interpreted there is no means for executing a program at any entry point other than its physical beginning There is also no internal method for passing parameters between programs Sometimes referred to as common area but this can be easily accomplished in a number of ways by making use of the XCALL statement and creating a common area within an external subroutine A Parameter could also be passed to the next CHAINING TO SYSTEM FUNCTIONS It is sometimes desirable to transfer execution to a system function or a command file from a BASIC program If the name of the file in the CHAIN statement does not have the RUN extension it is assumed to be a system command function In this case the AlphaBasic runtime package will create a dummy command file at the top of the current user partition and then transfer control to the monitor command processor The monitor will then interpret this dummy command file as a direct command and will continue execution at that point Note that the dummy command file created by the runtime package is merely the one line name specified in the CHAIN Statement and not the command file itself which may be the target function desired Some valid examples ar
16. large to fit in memory in the interactive mode In compiler mode the user compiles the program and stores the compiled object code on the disk During the actual running of the campiled program only the object code and a minimal run time execution package need be in memory thereby conserving space and increasing run speed The compiler and the runtime package are both written in reentrant code so that they may optionally be shared by all users running or debugging programs The object programs created by the compiler are also totally reentrant and sharable thereby further reducing memory requirements if it is desired to allow several users to run the same program Data formats supported include floating point string binary and unformatted variables All data formats may be simple variables or array structures In addition the unique memory mapping system allows the user to specify the ordering of variables in prearranged groupings for more efficient processing This system is similar to the data formatting capabilities of the COBOL language and lends itself nicely to business applications where the manipulation of formatted data structures is of prime concern Advanced compiler techniques have been used in all areas to give a truly commercial grade processing system which may be easily integrated into a series of programs not limited to the BASIC language If the user is running a large number of BASIC tasks in a timesharing environment the
17. or with Single step commands END Causes the program to terminate execution and return to the READY mode The END statement does not terminate campilation of the program nor is it required at the end of the program Rs OTHER VERBS AlphaBasic supports other verbs which are described in separate sections to more fully go into the details of their Operation The following verbs are described elsewhere SCALE scaled arithmetic modifier CHAIN executes a new program or command file ON ERROR controls error trapping and processing OPEN opens an I O file for processing CLOSE closes an I O file to further processing WRITE write a record to a random access file KILL deletes a file from disk ALLOCATE allocates a random access file on disk LOOKUP searches for a file and returns its size XCALL executes an external assembly language subroutine ISAM facilitates processing of indexed sequential files 36 BASIC FUNCTIONS The following is a list of the currently implemented functions which are available for use in expressions Note that the mode independence feature of the expression processor will perform automatic conversions if a numeric argument is used where a string argument is expected and vice versa EXP X Returns LOG X Returns LOG10 Returns SOR X Returns INT X Returns FIX X Returns FACT X Returns ABS X Returns SGN X Returns if X is the
18. primary index and data files named MASTER IDX and MASTER IDA respectively The structure also has Secondary index files named ADRESS IDX and PAYROL IDX which access the MASTER IDA file in different Sequences If it is desired to process the file structure via the sequence used by the ADRESS IDX index file the following two statements would be required OPEN 1 MASTER INDEXED RECSIZ RELKEY OPEN 2 ADRESS INDEXED RECSIZ RELKEY Note that the record size expression RECSIZ and the relative key variable RELKEY are identical in both Statements This is important since they both refer to the same data file MASTER IDA ISAM statements may then be made referring to either index file 1 or 2 but all READ and WRITE statements must be made to the data file which is associated with the primary index file 1 In other words READ and WRITE statements must not be made to file 2 ISAM STATEMENTS There are six functions which are performed by the ISAM statement with the codes 1 6 as listed previously They will be explained in more detail in this Section In the following descriptions there are codes which either require a relative key as input or return a relative key to be used when accessing the data record In all cases this relative key will be retuned in the variable specified in the OPEN statement for the index file being accessed by the ISAM statement This then leaves the System properly set up for an immediate access to the
19. program prior to the next access to the file Current e ae 33 34 35 36 37 38 soft error codes in effect are illegal ISAM statement code record not found in index file search duplicate key found in index file during attempted key addition link structure is smashed and must be recreated index file is full data file is full free list is empty end of file during sequential key read 67 CHAINING TO OTHER PROGRAMS AND SYSTEMS AlphaBasic supports the CHAIN statement which terminates execution of the current program and initiates the execution of a new program or system function The new program to be executed must be named in the CHAIN statement itself and may be a full file specification if desired The program named in the statement may be another AlphaBasic program compiled only or it May be a system command or command file name This allows a program to execute a command file and invoke system commands if required CHAINING TO ANOTHER ALPHABASIC PROGRAM The default extension of the file specification in the CHAIN statement is RUN which names a new AlphaBasic program to be executed If the extension of the evaluated file specification is indeed RUN either explicitly or by default the new program is loaded into memory and executed All variables in the new Program are first cleared to zero prior to execution The program must be compiled and must be in the current user area on disk unless an explicit area
20. record You should also check each ISAM statement for any possible error that might otherwise go unnoticed ERROR PROCESSING Every ISAM statement executed may potentially result in some form of error Errors will fall into one of two categories hard or soft Hard errors are defined as errors which are returned to the ISAM processor from the monitor file service system indicating some invalid disk operation Soft errors are those which occur within the ISAM Processor indicating an error or condition peculiar to ISAM files only AlphaBasic handles these two types of errors in different ways Hard errors will cause the standard BASIC error processor to be invoked resulting in either a message and program abort or an error trap if ON ERROR GOTO is in effect These errors may be detected with the normal error processing defined in the section dealing with the ON ERROR GOTO statement Soft errors will never result in an error message or error trap and it is therefore up to the programmer to test for these error conditions after every ISAM statement This is done by using the ERF X file error function where X is the file number used in the ISAM Statement The ERF function operates in a similar fashion to the EOF X function If the ERF function returns a zero the Preceding ISAM operation was successful If the value returned is not zero then 66 an error or abnormal condition was detected and proper corrective action should be taken in the
21. record will not be Modified An attempt to transfer more data than is in the logical record size will result in an error message The most efficient use of the randam files comes when the variable or variables used are Mapped by the MAP statement to the exact picture of the record format in use 50 CALLING EXTERNAL ASSEMBLY LANGUAGE SUBROUTINES External subroutines written in assembly language code may be called from any AlphaBasic program using the XCALL Statement The syntax for this statement is as follows XCALL routine argument l argument 2 argument n The routine to be called is an assembly language program which has been assembled using the MACRO assembler The resulting PRG program file must then be renamed to give it the assumed extension SBR indicating this is a subroutine and not a runnable program When the XCALL statement is executed by the AlphaBasic runtime system the named Subroutine will be located in memory and then called as a subroutine AlphaBasic first saves all registers and then sets certain parameters into these registers for use by the external subroutine The addresses of the arguments are calculated and entered into an argument list in memory along with their sizes and type codes The base address of this list is then passed to the user routine in register R3 The arguments may be one of two basic forms A variable name may be used in which case the argument entry in the list will reference the selected variabl
22. run time package is fully reentrant and may be included in the resident monitor for a further increase in memory efficiency The compiled object programs are also totally reentrant and sharable for users who are running the same application programs Variable names are not limited to the conventional single character and single digit format but may be any number of alphanumeric characters in length as long as the first character is alphabetic This is accomplished by using a dynamic tree structure for the storage of reserved words and variable names within the compiler and is another feature which makes AlphaBasic ideally suited for business applications Since the source code is compiled and need not be in memory when the program is eventually run the length of the variable name is not a significant concern Label names may also be used to identify points in the program for GOTO and GOSUB branches Label names are alphanumeric and help to clarify the program structure when used with meaningful definitions The following sections will describe AlphaBasic features and operations It is assumed that the reader is familiar with conventional BASIC language concepts This initial description is not meant as a tutorial to the novice but rather as an informational packet which will list the supported functions of conventional BASIC and give more detailed study only to those areas that differ from the norm In all cases AlphaBasic has been designed to sup
23. runtime package is located and loaded from disk if necessary and then started The runtime package then initializes user memory and locates the user program specified in the RUN command This program has the extension RUN and may be either in memory already or be loaded from the user disk area Loading of the program is autamatic by the runtime package if the program is not in memory The program is then executed and continues until the end or until aborted by the user typing a control c on his terminal Note that the RUN command serves two different functions depending on whether the user is in monitor mode or in interactive BASIC mode In monitor mode the RUN command is used to execute a compiled BASIC program which has been previously stored on disk or loaded into memory explicitly The command RUN PAYROL will run the PAYROLL RUN compiled program and then exit back to monitor mode without ever entering the campiler itself In interactive BASIC mode the RUN command will compile and run the current source program which the user is editing and testing The user should note these differences While we are on the subject of running source programs versus compiled object Programs there is one more restriction that the user should note The CHAIN Statement causes one program to terminate execution and turn control over to the next program in sequence This chained program must be a compiled object Program or it will not be found There is no mechanis
24. sign itself The above two special functions may be combined by starting the numeric format field with the symbols S combines the asterisk fill and floating dollar sign functions described above This combination also defines three digit positions one of which is taken up by the dollar sign itself TRAILING MINUS SIGN If a numeric field is terminated by a minus sign in the format string the sign of the output number is printed following the number instead of before it If the number is negative a minus sign will be printed If the number is zero or positive a blank will be printed Note that if the trailing minus sign is not used space must be provided in the numeric field for the sign to precede the number if it is negative COMMA If the numeric format field contains one or more commas in front of the decimal point a comma will be inserted every three digits to the left of the decimal point when the number is printed Each comma also defines one digit position in the format field Commas appearing after the decimal point will be treated as printing characters EXPONENTIAL FORMAT Exponential format may be specified by following the numeric field designation with four up arrows which defines the space taken up by the E NN exponent value As with other numerical formats any decimal point arrangement is A allowed and the significant digits are left justified with the exponent being adjusted as necessary 42 SCA
25. 4 5 6 left relative position values 6 5 4 3 2 1 right relative position values Allowing negative values for right relative positions provides the ability to pick out digits within a numeric string without having to calculate the total size of the string first and then work fram the left The substring length argument used by the second format may also take on negative values for a more flexible format Normally the length is a positive value which represents the number of characters counting the beginning position and incrementing the index to the left A negative length causes the index to move to the right and returns a substring whose last character is the one marked by the beginning position argument Confusing Perhaps a few examples may clarify the use of substring modifiers Assume the main string is AS and it contains the above example of ABCDEF The following substrings will be returned AS 2 4 equals BCD AS 2 4 equals BCDE 16 AS 3 3 equals C A 3 3 equals CDE AS 3 2 equals DE AS 3 2 equals CDE AS 33 2 equals BC AS 3 2 equals CD AS 4 1 equals D AS 4 1 equals D Any position values or length values which would cause the substring to overflow out of either end of the main string will be truncated at the string end AS 3 10 equals CDEF AS 14 34 equals ABCDEF The main string to which the substring modifier is applied is actually any expression and need not be a defined single string var
26. AlphaBasic User s Manual 1977 ALPHA MICROSYSTEMS ALL RIGHTS RESERVED I would like to express my thanks to the following people for their assistance in the development of AlphaBasic Mike Roach for debugging and suggestions on implementation of new features Paul Allen Edelstein for the trig package and advanced mathematical assistance Bob Hitchcock for operator and applications oriented suggestions and scaled arithmetic the myriad of users who pointed out problems in the earlier versions sometimes not so tactfully and most of all to Carolyn without whom much of this may never have been accomplished Dick Wilcox AMOS AlphaBasic and AM 100 are tra emarks of products and software of ALPHA MICROSYSTEMS Irvine CA 92714 1977 ALPHA MICROSYSTEMS ALPHA MICRO 17881 F Sky Park North Irvine CA 92714 INDEX INTRODUCTION TO ALPHABASIC ALPHABASIC GENERAL INFORMATION INTERACTIVE VS COMPILER MODES RUNNING BASIC PROGRAMS ALPHABASIC VARIABLES DATA FORMATS ALPHABASIC EXPRESSIONS LOWER CASE CHARACTERS SUBSTRING MODIFIERS MEMORY MAPPING SYSTEM INTERACTIVE COMMAND SUMMARY PROGRAM STATEMENTS BASIC FUNCTIONS FORMATTED OUTPUT VIA PRINT USING STATEMENTS SCALED ARITHMETIC ALPHABASIC FILE I O SYSTEM FILE I O STATEMENTS CALLING EXTERNAL ASSEMBLY LANGUAGE SUBROUTINES ERROR TRAPPING SYSTEMS FUNCTIONS EXPANDED TAB FUNCTIONS SCREEN CONTROL FORMATTED NUMERIC DATA VIA THE USING MODIFIER PROCES
27. L FDIV and FOMP Of the 48 bits in use for each 6 byte variable the high order bit is the sign for the mantissa The next 8 bits represent the signed exponent in excess 128 notation giving a range of approximately 2 9 10 39 thru 1 7 10738 The remaining 39 bits contain the mantissa which is normalized with an implied high order bit of one This gives an effective 40 bit mantissa which results in an accuracy of 11 significant digits STRING used for the storage of alphanumeric text data String variables may be assigned fixed lengths for efficiency and speed or may be left to dynamically vary in size as the data changes Fixed length string variables require one byte of storage for each character and may be fixed in position using the memor y mapping system Dynamic length string variables may not be mapped into memory since their position may shift during execution An indexing scheme is defined to the user if it is absolutely necessary for him to locate these strings with an external routine 11 BINARY binary variables are similar to integer variables in other implementations of BASIC A binary variable may be from 1 to 5 bytes in length and may be signed when all 5 bytes are specified When less than 5 bytes are specified for the size in a MAP statement the binary value may be loaded as a negative number but will always be returned as a positive number of full Magnitude with the upper bit preloaded as the sign taking on its specific
28. LED ARITHMETIC AlphaBasic uses a floating point format which gives an accuracy of 11 significant digits Unfortunately this accuracy is absolute only when dealing with those numbers which are total integers in the first 11 positions to the left of the decimal point This fact stems from the conversions that are required from decimal input to the binary floating point format used in the hardware For most business users the actual range of numbers contains two digits to the right of the decimal point and nine digits to the left of the decimal point When the fractional part of the number is converted between decimal and binary formats a small but significant error is sometimes introduced which may propogate into large inaccuracies when dealing with absolute dollars and cents values AlphaBasic incorporates a scaling feature which helps to alleviate these problems by storing all floating point numbers with a scale offset This offset may be used to effectively designate where the 11 absolute accuracy digits are located in relation to the decimal point This is done by multiplying every input number by the scaling factor and then dividing it out again before printing This is a simplified explanation and many other checks and conversions are done internally to scaled numbers but that is the general idea The scaling factor represents the number of decimal places that the 1l digit window will be effectively shifted to the right in any floating
29. N lt line number gt IF lt relative expression gt THEN lt line number gt ELSE lt line number gt IF lt relative expression gt lt statement gt IF lt relative expression gt lt statement gt ELSE lt statement gt IF lt relative expression gt THEN lt statement gt IF lt relative expression gt THEN lt statement gt ELSE lt statement gt The above formats may be nested to any depth and rather than go into detail we suggest that you play around with them to determine the actual restrictions that exist Some examples IF A 5 THEN 110 IF A gt 14 THEN 110 ELSE 220 IF BS END PRINT END OF TEST IF TOTAL gt 14 5 GOTO 335 IF AA 5 AND BB 6 IF CC 7 PRINT 567 ELSE PRINT 56 ELSE PRINT NONE IF A 1 PRINT 1 ELSE IF B 2 THEN 335 ELSE 345 DIM The dimension statement defines an array which will be allocated dynamically at execution time There is no limit to the number of subscripts that may be used to define the individual levels within the array The statement DIM A 20 defines an array with 20 elements referenced as A 1 through A 20 Multiple arrays may be dimensioned by a single DIM statement by separating them with comas Subscripts are evaluated at execution time and not at campile time thereby 34 allowing variables to be used as subscripts instead of fixed values The statement DIM A B C will allocate an array whose size will depend on the actual vaules of B and C at the time the DIM statement is executed Strin
30. RROR TRAPPING em AlphaBasic allows the user program to trap errors that would normally cause the system to print an error message and abort the program run During interactive processing this would return you to AlphaBasic command mode and during compiled run processing it would return you to to monitor command mode Use of the ON ERROR GOTO and RESUME statements allows these errors to be detected within the user program and immediate action to be taken when appropriate ON ERROR GOTO STATEMENT Error trapping is enabled and disabled by using the ON ERROR GOTO statement in one of two forms The first form specifies a line number or label within the Program When the program encounters this ON ERROR statement it stores the line number and sets a flag enabling error trapping If an error occurs at any time after this control will be transferred to the routine specified by the line number No error message will be printed Examples of this form of the statement are ON ERROR GOTO 500 ON ERROR GOTO TRAP ROUTINE The error routine must then take appropriate action based on the type of error which caused the trap The ERR function will return the following data based on conditions at the time of the error ERR 0 numeric code specifying the type of error detected ERR 1 last line number encountered prior to the error ERR 2 last file number accessed pertinent only for file errors The second form of the statement disables further err
31. Runtime was x x seconds BASIC is telling you how long it took to run your program Runtime package RUN PRG not found BASIC or COMPIL was unable to locate the runtime package or did not have sufficient memory in which to load it July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 10 Stack overflow BASIC s internal stack has overflowed This is most often caused by such operations as nesting GOSUBs too deep or branching out of FOR NEXT loops Subroutine not found The specified subroutine could not be found Subscript out of range The specified subscript is outside the range specified in the DIM or MAP statement for the subscripted variable Syntax error The syntax of the specified line is invalid System commands are illegal within the source program BASIC system commands LOAD DELETE LIST etc are not valid within a BASIC source program System error A system error has occurred during the execution of the specified line System error is used as a catch all error message for a variety of unlikely occurrences Temporarily all arrays must be less than 32K The array size you specified is larger than 32K bytes Undefined line number or label The line number or label specified in a GOTO or GOSUB statement is not defined within the program Write protected Your program tried to write on a write protected device Wrong number of subscripts The number of subscripts specified is not the same as the num
32. SING INDEXED SEQUENTIAL FILES CHAINING TO OTHER PROGRAMS AND SYSTEMS Note This manual reflects AlphaBasic version 3 3 release PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE PAGE IO OW 13 15 16 18 27 31 37 4l 43 45 47 51 54 57 59 60 6l 68 INTRODUCTION TO ALPHABASIC AlphaBasic is an extension of the popular BASIC language with several features not found in other implementations These features not only enhance the performance of traditional uses of the language but also make business applications easier to program COBOL users will find the 1 0 processing convenient for data manipulation while the memory mapping system will entice the assembly language programmers who wish to link up their own external routines Floating point hardware in the processor is fully supported making AlphaBasic faster for mathematical computations than any other BASIC currently implemented in a microprocessor system AlphaBasic runs in one of two modes Interactive mode operates in the traditional manner whereby the user creates alters and tests his program which resides totally in memory This mode is convenient for the creation and debugging of new programs or the dynamic alteration of exisitng programs Compiler mode is more useful for programs which are to be put into productive use or for testing programs which are too
33. Y S 14 21 230 MAP2 STATE S 4 300 MAP DOUBLE ARRAY ARRAY m 310 MAP2 UNIT 6 320 MAP3 CODE B 2 330 MAP3 RESULT F Statements 100 100 define an array with 8 floating point elements for a total of 48 bytes in memory Statements 200 230 define an area with three string variables in it for a total of 42 bytes Normally this area would follow the 48 byte ARRAY area in memory but the origin parameter in statement 200 causes it to overlay the first 42 bytes of the ARRAY area instead Statements 300 330 define another array area of a different format with 6 elements each element being composed of one 2 byte binary variable CODE and one floating point variable RESULT The origin parameter in statement 300 also causes this area to overlay the ARRAY area exactly Caution The above scheme allow variables to be referenced in a different format than when they were entered into memory If you load the 8 elements INDEX 1 through INDEX 8 with floating point values and then reference the variable STREET as a string you will get the first four floating point variables INDEX 1 through INDEX 4 which will look very strange in string format EXAMPLES The following two statements produce identical arrays 100 DIM Al 10 110 MAP1 A1 10 F Both statements produce arrays containing ten floating point variables referenced as Al 1 thru Al 10 Statement 110 however will define its placement in memory in relation to other mapped variabl
34. a continuation line by making an ampersand amp the last character on that line For example 10 IF X lt 12 2 OR B gt 0 THEN amp X 147 2 amp ELSE amp J B 20 Q 1252 The maximum size of any line including any continuation lines is 500 characters If a program with continuation lines is loaded into the interactive compiler the lines are concatenated into one line Therefore loading and saving a program with continuation lines under the interactive compiler BASIC results in the elimination of the continuation lines 10 0 MESSAGES OUTPUT BY THE ALPHABASIC SYSTEM Below is a complete list of all messages output by the AlphaBASIC system i e BASIC RUN and COMPIL along with a brief explanantion of each message Bitmap kaput Your program attempted a file operation TOREN ALLOCATE etc on a device with a bad bitmap Break at line n The program reached the breakpoint that was set at line n COMPILE BASIC is telling you that it is compiling your program July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 6 can t continue You have attempted to continue a program which is not stopped at a breakpoint or which has reached a point where it can go no further e g it has reached an END statement Cannot find xxxxxx The program xxxxxx was not found Compile time was x x seconds BASIC is telling you how long in elapsed time not compute time it took to compile your program DELETE wha
35. acted from the free list and the relative key is returned in preparation for a code 2 index key addition statement If no more data records are free in the data file a data file full error will be returned All free records in the data file are kept in a linked list called the free list This list is built initially by ISMBLD and contains all the records in the data file As code 5 ISAM statements are executed the free list delivers these records and as code 6 ISAM statements are executed the records are again returned to the free list for reuse The index file is not modified and the symbolic key in the statement is ignored This call must be made only to the Primary index file number Code 6 the data record specified by the relative key is returned to the free list for reuse by a code 5 call The index file is not modified and the symbolic key in the Statement is ignored This call must be made only to the primary index file number READ AND WRITE STATEMENTS The ISAM calls do not access the data records themselves but merely deliver back the relative key of the associated data record to be used Normal READ and WRITE statements are then used to actually retrieve or write into the data record itself These READ and WRITE statements follow the same format used when accessing a normal random access data file in AlphaBasic The relative key associated with the Primary file as specified in the OPEN statement must contain a valid relative key for th
36. al reasons why this condition is not checked for and reported as an error I will not bore you with the gory details on this except to insure you that it will be implemented in a future production relase of AlphaBasic In the meantime you can kinda keep tabs on memory usage with the MEM X function which will deliver back various usages of memory during program development See the section on functions for more detail on this EXPANDED SOURCE TEXT MODE AlphaBasic normally scans the source text in expanded mode version 3 2 which dictates that reserved words verbs functions commands etc be terminated by a space or a character that is illegal in variable names This allows labels and variables to begin with reserved words Tn other words the variable name PRINTMASTER will not be interpreted as PRINT MASTER in expanded mode In this mode the statement FOR A l TO 10 cannot be written as FORA 1T010 There are two commands which the user may apply to switch back and forth between normal and expanded modes EXPAND sets syntax scanner to expanded mode NOEXPAND sets syntax scanner to normal mode AlphaBasic initializes itself in the expanded mode Note that the mode in which a program is compiled has nothing to do whatsoever with the resultant object code which is generated either in size or execution speed INTERACTIVE VS COMPILER MODES i AlphaBasic may be run in one of two modes of execution Interactive mode deals directly with the user o
37. ame is followed by a set of subscripts within parenthesis then the variable will be assigned as an array with the dimensions specified by the subscripts just as if a DIMENSION statement had been used to assign the array For example the statement MAP A F assigns a single floating point variable called A but the statement MAP A 5 10 F assigns a floating point array with 50 elements in it 5 X 10 just as if the statement DIM A 5 10 had been executed Note that since these mapped arrays are assigned memory at campile time and not at run time the subscripts must be decimal numbers instead of variables TYPE CODE The type code is a single character code which specifies the type of variable to be mapped into memory The following variable types are implemented in AlphaBasic unformatted absolute data variable string variable floating point variable binary unsigned numeric variable Wyn gt lt If no explicit type code is entered unformatted data type X is assumed Unformatted data is absolute in memory and is usually only used to reference a group of other variables as one mit Until more specific details on data handling are determined unformatted data variables should only be moved to other unformatted data variables For all practical purposes unformatted data variables are treated like string variables except that they are not terminated by a null byte only by the explicit size of the variable 19 String variabl
38. anatic for calculations Since all binary variables are converted to floating point format before performing any arithmetic calculations binary arithmetic is actually slower than normal floating point arithmetic and is used mainly for compacting data into files and arrays where the floating point size of 6 bytes is inefficient When conversions from floating point to binary are done any data that will not fit within the defined size of the target variable will merely be lost with no error message given Range checks where required are the responsibility of the programmer prior to moving a floating point number to a binary variable area The best way to understand this is to play with a few examples in immediate mode Please take note that the use of binary numeric variables is not allowed in some instances FOR NEXT loops may not use a binary variable as the target variable although they may be used in the FROM TO and STEP value expressions The record number key in a random mode OPEN statement must be floating point also The result variable of a LOOKUP statement must be floating point SIZE The size parameter in the MAP statement is optional but if used it must be a decimal number specifying the number of bytes in the variable If it is amitted it will default to 0 for unformatted and string types 6 for floating point types and 2 for binary types The size parameter of floating point variables must be 6 or mitted VALUE An initial valu
39. and error trapping to be enabled again CONTROL C TRAPPING When the operator types a control c on his keyboard during the execution of an AlphaBasic program the program is suspended at the next statement Action taken then depends upon the status of the error trapping flag If no error trapping is enabled the program is aborted with the appropriate message being printed on the terminal If error trapping is enabled the error trapping routine is entered with the code in ERR 0 being set to 1 This feature allows the user to Prevent inadvertant aborting of Programs during critical times such as file updates Control c action is Suspended during error recovery processing to prevent accidental aborting of the program due to an error condition occurring within the error routine The control c will be detected immediately upon execution of the RESUME statement 55 Q W RNONNUNNNNERB E e BE CAAUBRUNESODIATRONEOWOUAU RUN N O UJ ERROR CODES RETURNED BY ERR 0 Meaning Control c interrupt System error Out of memory Out of data NEXT without FOR RETURN without GOSUB RESUME without ERROR Subscript out of range Floating point overflow Divide by zero Illegal function value XCALL subroutine not found File already open IO to unopened file Record size overflow File specification error File not found Device not ready Device full Device error Device in use Illegal user code Protection violation Write protect
40. arallels the ON GOTO statement in format and operation except that pointl through pointN represent entries to subroutines that will be executed based on the result of the expression evaluation As with the GOSUB Statement the verb CALL may be used in Place of the verb GOSUB giving an ON CALL statement READ DATA RESTORE These calls allow data to be an integral part of the source program with a method for getting this data into specific variables in an orderly fashion DATA statements are followed by one or more literal values separated by commas String literals need not be enclosed in quotes unless the literal data contains a comma All data statements are placed into a dedicated area in memory no matter where they appear in the source program READ statements are followed by one or more variables separated by commas Each time a read statement is executed the next item of data is retrived fram the DATA statement pool and loaded into the variable named in the read statement If there is no more data left in the data pool an error message results and the program is aborted The RESTORE statement is used to initialize the reading of the data pool fram the beginning again l gis DATA 1 2 3 4 5 DATA 2 3 0 555 ONE STRING 4 4 READ A B C READ AS READ C 2 3 B 4 RESTORE The READ statement is also used for reading data from random access files For details on this refer to the section describing the file I O system INPUT Al
41. are referenced by the compiled program code through an indexing scheme Each variable in the working storage area contains an item in the index area which contains all information needed to define and locate that variable The working storage area therefore contains only the pure variables themselves without any associated or intervening descriptive information The index area is a separate entity physically located before the working storage area in memory The allocation of the variable storage area for any program is predictable and normally done as each variable is encountered during compilation Since this scheme is not easily followed by the user a different method must be derived which can override normal allocation processes for those users who wish to have the variables allocated in a predetermined manner The disk I O system also requires that the variables used be in a specific relationship to each other when used in the more advanced modes The MAP statement has been included in AlphaBasic for the purpose of allocating variables in a specific manner MAP statements are non executable at run time but merely direct the compiler in the definition and allocation of the referenced variables Each MAP statement contains a unique variable name to which the statement applies When the compiler encounters this statement it allocates the next contiguous space in working storage as required and assigns it to that variable name The type of the varia
42. are supported via some elementary ISAM statements which allow the direct control of index file and data file items The ISAM package as implemented requires more direct user control over the files than other implementations for two main reasons The major reason is memory limitations which restrict the amount of smarts that could be put into the AlphaBasic runtime package for handling ISAM files The second reason is the current structure of the existing ISAM package which precluded a more extensive implementation within the timespan allocated to the project We feel that the version we are releasing will provide sufficient control for all users even though the initial training and programming may at first appear more complex than other versions of ISAM handlers For more detailed information on ISAM files and the ISAM assembly language package the user is referred to the Separate manual titled ISAM System User s Guide FILE STRUCTURE An indexed sequential file consists of one data file and one or more index files which link to the data file The data file is structured identical to a normal random access file with the additional restriction that all records which are not currently active are linked to each other in a chain called the free data record list All data records reside in the data file and the data records may be any size up to the maximum of 512 bytes As in the normal random access file data records will not be Split across physica
43. atement location This statement location must be either a line number or a label defined somewhere in the program The line number or label must follow the GOTO statement in the program The GOTO statement may be broken apart as GO TO if desired GOSUB CALL RETURN Calls a subroutine which starts with the line number or label following the GOSUB or CALL verb The subroutine exits via the RETURN statement which returns control to the statement following the GOSUB or CALL statement Executing a RETURN statement without first executing a GOSUB statement will result in an error message Both GOSUB and RETURN are currently illegal in immediate mode Note that the CALL verb is merely another way of specifiying GOSUB for those programmers used to this verb from other languages ON GOTO The ON GOTO statement allows multi path GOTO branching to one of several points within the program based on the result of evaluating an expression The actual format is ON expression GOTO pointl point2 point3 pointNn The expression can be any valid expression which will be evaluated down to a positive integer result The result will then be tested to branch to pointl if 1 point2 if 2 point3 if 3 etc If the result is zero negative or greater 31 than pointN the program will fall through to the next statement The points pointl through pointN may be line numbers labels or any misture of the two fr ON GOSUB The ON GOSUB statement p
44. ber defined in the DIM or MAP statement for the subscripted variable July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL 11 0 RESERVED WORDS Page 11 Below is a list of the reserved words used by the BASIC compilers BASIC and COMPIL You MUST not use any of these reserved words as variable names or labels Reserved Word ABS ACS ALLOCATE AND ASC ASN ATN BREAK BYE BYTE CALL CHAIN CHR CHRS CLOSE COMPILE CONT cos DATA DATE DATN DEF DELETE DIM ELSE END EOF EQV ERF ERR ERROR EXP EXPAND FACT FILEBASE FIX FOR GO GOSUB GOTO IF INDEXED INPUT INSTR INT July 1979 Meaning absolute value arccosine allocate file logical AND ASCII value arcsine arctangent set breakpoint exit to monitor memory byte call subroutine chain next program character value character value close file compile program continue execution cosine data statement system date double arctangent define function delete lines dimension else end of program end of file logical equivalence file error error status error exponentiation expand mode on factorial file base offset fix loop initiation program jump call subroutine program jump conditional test indexed input data search string integer ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 12 Io ISAM KILL LCS LEFT LEFTS LEN LET LINE LIST LOAD LOG LoG10 LOOKUP MAP MAX MEM MID MINS MIN NEW NEXT NOEXPAND NOT ON
45. ble arrays are those arrays which were allocated by a DIM statement or by a default reference to a subscripted variable MAP1 through MAP16 variables and variable arrays are those which were defined in a MAP statement Variable type is the type of the variable and may be UNFORMATTED STRING FLOATING POINT or BINARY If the variable is an array the dimensions will be listed after the variable type code in the format ARRAY n n n where n n n are the values of the subscripts in use by the array If the array is dynamic and has not been allocated yet the subscript values will be replaced by the letter X to indicate that they are not known at this point Remember that any variable defined in a MAP statement which is in a lower level to another variable will inherit all subscripts from that higher level variable The size of the variable will be given in bytes In the case of arrays the Size will represent the size of each single element within the array The location of the variable is a little tricky to explain since it is actually an offset to the base of a storage area set aside for the allocation of user variables As each new variable or array is allocated it will be assigned a location which is relative to the base of this storage area The location information is given here to help you understand the relative placement of the variables in the mapping system and does not represent the actual memory locations which they occupy There are two
46. ble is also specified in this statement and may be used to override the standard naming conventions of BASIC All variables not defined in a MAP statement will then automatically be assigned storage in sequence for total compatibility with existing standards As has been noted previously this memory mapping system is primarily required for advanced disk I O techniques and to assist in linking with assembly language routines Special functions are provided to deliver the absolute addresses of these areas as parameters during assembly routine calls By knowing the layout of the variables in memory the user need only pass the base address of the area to the routine and the routine can then reference all needed variables by indexing techniques The mapping system does have another distinct advantage to the sophisticated programmer in the allocation of arrays With the MAP statement the user has the ability to override the standard array allocation scheme and force the allocation to proceed in a more flexible manner Conventional BASIC arrays are allocated in contiguous memory for each subscripted variable encountered AlphaBasic allows several variables to be cambined in a single contiguous array which can provide efficiency in the manipulation of associated data structures MAP STATEMENT FORMAT Briefly the MAP statement has the following syntax MAPn var iable name dimensions type size value origin 18 Q MAPn represents the hierarchy of t
47. ble name matches The following statements are all considered to be identical FOR A 1 TO 100 STEP 2 For A 1 To 100 Step 2 For A 1 to 100 step 2 for A 1 to 100 step 2 Lower case letters used within string literals inside quotes will be retained and printed as lower case Lower case letters which are entered into string variables by means of the INPUT statement will also be retained as lower case letters The entire string processing system now supports lower case characters Note that all lower case characters are considered greater than any upper case character due to their position in the ASCII colating sequence To assist in processing and comparing input which contains lower case letters the UCS X function has been implemented This function will return a string which is identical to the argument string X with all lower case characters being translated to upper case The inverse function LCS X will return a string with all upper case characters being translated to lower case 15 SUBSTRING MODIFIERS AlphaBasic supports a unique method of manipulating substrings A substring is defined as a portion of an existing string which may be as small as a single character or as large as the entire string Substring modifiers allow the substring to be defined in terms of character positions within the string relative to either the left or right end of the string The length of the l substring is defined either in terms of its begin
48. ce This is actually the address of the first word following the argument list in memory 51 and may be used to store a terminator word to stop scanning of the argument list if desired R5 this is the arithmetic stack index used by the runtime system The arithmetic stack is built at the top of the user partition and grows downward as items are added to it When the external subroutine is called R5 points to the current stack base Since the arithmetic stack may contain valid data the external subroutine must not use the word indexed by R5 or any words above that address ARGUMENT LIST FORMAT The list of arguments Specified in the XCALL statement may range from no arguments at all to a number limited only by the space on the command line To pass these arguments to the external subroutine an argument list is built in memory which describes each variable named in the list and tells where it can be located in the user impure area The variables themselves are not actually passed to the subroutine but rather their absolute locations in memory are In this way the subroutine may inspect them and modify them directly in their respective locations This does not apply to expressions which are built on the stack as described previously R3 points to the first word of the argument list which is a binary count of how many arguments were contained in the XCALL statement Following this count word comes one 3 word descriptor block for each argum
49. corresponding data record via a READ or WRITE statement Code 1 the specified index file is searched for the key which matches the symbolic key in the statement If a match is found the associated relative key will be delivered back for access to the data file If the key is not found an error code 33 will be returned see section on error processing Code 2 the specified index file is accessed and the next sequential key is located The corresponding relative key is returned in preparation for a READ Or WRITE to the data file If this is the first access to the file following the OPEN statement the first sequential key will be located If this statement 63 follows a Previous code 1 Statement the next Sequential key following the code l key will be located If there are no more keys in the index file and a relative key Code 3 the specified symbolic key is added to the index file along with the relative key which must be in the corresponding variable Specified in the OPEN Statement This relative key will normally be already set up by a prior code 5 Code 4 the specified symbolic key is located in the index file and deleted from it The Corresponding data record relative key is returned so that the data record may then be deleted and returned to the free list by using a code 6 ISAM statement If the symbolic key is not located in the index file a record not found error will be returned 5 the next available data record is extr
50. d The source text is built up in memory as it is entered and is automatically kept in numerical sequence Multiple lines terminated by line feed instead of carriage return are not supported at this time but will be in the future No syntax checking is performed when the line is entered other than validation of the line number When the user enteres the RUN command the source program is compiled in its entirety and syntax errors reported The resulting object code is also stored in memory which results in a much greater initial memory requirement If no errors were detected the object code is then executed with no direct reference to the source code anymore Any further editing of the source code will set a switch to automatically force recompilation when the next RUN command is entered The user may store the compiled object program on disk by using the SAVE command with the explicit extension RUN appended to the program name This saved program may then later be executed without recompilation by use of the monitor RUN command which calls upon the runtime package only MULTIPLE STATEMENT LINES The system supports multiple statement lines by using colons to separate the statements The normal rules apply such as REM and DATA statements cannot contain other statements on the same line Immediate mode commands may also be Multiple statement lines IMMEDIATE MODE COMMANDS All lines typed into AlphaBasic by the user will be considered for imm
51. d multiplied by the 5 The plus sign then performs numeric addition instead of concatenation The result is in numeric format and will be converted if its destination is a string If this approach seems confusing you should try a few examples of your own on the system to see what the results are Remember any potentially ambiguous expression may always be forced to one or the other type by use of the STR and VAL functions 14 LOWER CASE CHARACTERS Beginning with release 3 1 AlphaBasic supports lower case letters a z in both the input source program and in the runtime execution of programs The line editor built into the interactive system will now accept and store source input text in lower case characters if desired Lower case letters when used within variable names and labels will be unique and separate from the corresponding upper case letters In other words the variable a is separate from the variable A and the variable Tom is separate from the variables TOM and tom Lower case letters may be used as the first character of a variable name or program label just as upper case letters may be Reserved words are treated somewhat differently from the above system When a reserved word is expected the syntax parser temporarily translates all lower case letters to upper case and then checks for a reserved word match If the word is not a reserved word the translation is not retained and the lower case letters are used for varia
52. desired This is followed by a WRITE statement to rewrite the data record The WRITE statement must be made to the primary file The index files are not altered in this operation Note that the above method should only be used to alter data that is not a direct part of any symbolic key To change a symbolic key you must delete the key in the correct index file and then add the new key with another ISAM statement The data record need not be deleted and recreated during this operation unless necessary for complete new data DELETING A DATA RECORD The deletion of a data record in an indexed file structure involves not only the deleting of the data record itself but also the deleting of all symbolic keys associated with that data record All index files must be open for this operation The data record is first located by one of the symbolic keys via ISAM code 1 and then the data record is read into the buffer with a READ statement to the primary file Each symbolic key must then be extracted from the data record and used to delete each key from its associated index file with Successive ISAM code 4 statements The data record itself is then deleted and returned to the free list with an ISAM code 6 statement NOTE A good check on the structure would be to store the relative key in another variable and then compare the relative keys returned by each ISAM code 4 statement to insure that the symbolic keys all did indeed link to the correct data
53. ds and must be divided by the blocking factor for random files if the number of logical records is desired ALLOCATE The ALLOCATE statement is used to Preallocate a contiguous file on disk which may then be opened for random Processing An attempt to allocate a file which already exists will result in an error message A randam file need only be allocated once and may then be opened for random read write Operations as many times as desired The statement format is ALLOCATE lt filename gt lt number of records gt As in the OPEN statement the filename is any string expression which evaluates to a legal file description The number of records is a numeric expression which represents the number of physical 512 byte disk records to be allocated to the file 48 INPUT Once a file has been opened for input the data is read from the file by a Special form of the INPUT statement using a file number which corresponds to the number assigned in the OPEN statement The variables in the list may be either numeric or string variables but must follow the format of the data in the file being read Weird results will occur if you attempt to read string data into a numeric variable or vice versa The general format of the INPUT statement is INPUT lt file gt lt variable l gt lt variable 2 gt lt variable n gt During the reading of the input data into the variable list all leading spaces will be bypassed unless enclosed within quot
54. dy exists Your program tried to create a file which already exists File already open You have attempted to open a file that is already open on the same file number File not found BASIC was unable to locate the specified file File spec error The file specification you gave in a file operation OPEN LOOKUP etc is in error AlI file specifications must conform to the system standard i e devn file ext p pn File type mismatch Your program tried to perform a sequential operation on a random file or vice versa Floating point overflow A floating point overflow occurred during a calculation IO to unopened file Your program tried to perform input or output to a file that is not open Illegal GOTO or GOSUB The format of the GOTO or GOSUB statement is invalid Illegal NEXT variable The variable used in the NEXT statement is not valid e g not floating point Illegal PRINT USING format The edit format used in a PRINT USING statement is invalid Illegal SCALE argument The argument given in a SCALE statement is invalid the argument must range between 30 and 30 Illegal STRSIZ argument The argument given in a STRSIZ statement is invalid Illegal TAB format Your program has incorrectly specified a TAB function Illegal expression The specified expression is not valid July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 8 Illegal function value The specified function value is not val
55. e CHAIN SYSTAT PRG 1 4 CHAIN TEST1 CMD CHAIN DSKO MUMBLE CMD 2 2 68 Note that if the account number is not specified the action taken will be the Same as if the command was entered directly from the keyboard In other words programs and command files will normally be searched for in the user area only due to the fact that the extension had to entered explicitly Note also that the system function will be executed as a mainline function and not as a Subroutine to the runtime system This means that if you wish to automatically return to some AlphaBasic program you will have to execute a command file whose final command is a RUN command to begin the execution of said AlphaBasic program Confused Me too Good luck 69 July 1979 DwM 00100 44 ADDENDUM TO THE ALPHABASIC USER S MANUAL 1 0 INTRODUCTION The purpose of this document is to provide additional information for the BASIC programmer until such time as we can issue a new AlphaBASIC user s Manual part number DWM 00100 01 For more information on using the AlphaBASIC system turn to the AlphaBASIC User s Manual 1 1 Contents The next few sections discuss new AlphaBASIC features that are not discussed in the current AlphaBASIC manual Section 10 0 lists all messages displayed by the AlphaBASIC system and Section 11 0 lists all reserved keywords used by AlphaBASIC 2 9 EDITING MASKS The section in the current BASIC manual that discusses formatted outpu
56. e within the user impure area This variable is available to the called Subroutine for both inspection and modification The argument may also be an expression numeric or string in which case the expression is evaluated and the result is placed on the arithmetic stack referenced by R5 This result is then referenced in the argument list entry instead of a single variable It is available for inspection only since the stack is cleared when the subroutine exits The user routine is free to use and modify all six general work registers RO R5 and may use the stack for work space as required When the subroutine has completed its execution a return must be made to the runtime system by executing the RIN subroutine return instruction REGISTER PARAMETERS The following registers are set up by the runtime system to be used as required by the external subroutine They may be modified if desired as they have been saved before the subroutine was called RO indexes the user impure variable area RO is used throughout the runtime system to reference all user variables Details on the format of this area are not available at this time and the user need not be concerned with them RO may be used as a work register R3 points to the base of the argument list R3 may be used to scan the argument list for retrieval of the argument parameters R4 points to the base of the free memory area that may be used by the external subroutine as work spa
57. e may be given to any mapped variable by including any valid expression in the value parameter This value may be a numeric constant a string constant or a complete expression including variables Remember 20 however that the expression is resolved when the MAP statement is executed at runtime and the current value of any variable within the value expression is the one used to calculate the assignment result MAP statements may be executed more than once if it is desired to reload the initial values Note that if the size parameter is omitted such as for floating point variables but the value parameter is used there must be an extra comma to indicate the miSsing size parameter MAP1 PI F 3 14159 MAP HOLIDAY S 10 CHRISTMAS The first example preloads the value 3 14159 into the floating point variable called PI The second example preloais the letters CHRISTMAS into the string variable called HOLIDAY ORIGIN In some instances it may be desirable to redefine records or array areas in different formats so that they occupy the same memory area For instance a file may contain several different record formats with the first byte of the record containing a type code for that record format The origin parameter will allow you to redefine the record area in the different formats to be expected When the record is read into the area the type code in the first byte can be used to execute the proper routine for the record type Each differ
58. e operation or an error will result READ CLOSING AN INDEXED FILE In order to insure that all data records have been rewritten to the data file and that all links in the index file have been Properly updated and rewritten to the disk it is imperative that all index files Primary and secondary be closed using the normal CLOSE statement and referencing the correct file number Failure to do so may result in the link structure being sacrificed to the god TRON god of electrons and integrated circuits CREATING AN INDEXED FILE The steps used in the initial creation of an indexed file will be traced here Initially the structure must be created using the ISMBLD program Refer to the ISAM System User s Guide for details on this procedure Any secondary index files must also be created by the ISMBLD program The program to add the data items will then open all index files associated with the structure For each new data record to be added the following steps are performed The next free data record is retrieved ISAM code 5 and the data record is written into it with a WRITE statement One symbolic key is then added to each of the index files using ISAM code 3 statements All keys will therefore link to the same data record After all data records have been written to the file all files are closed READING AN INDEXED FILE SEQUENTIALLY To read an indexed file in the sequence of the symbolic keys the file is first opened by the program I
59. e record size Each READ or WRITE call will then Manipulate the 100 bytes of data by reading or writing to or from variables whose size totals 100 bytes Lets Say you now want to expand the file to 120 bytes and that most of the programs will not have to make use of the extra 20 bytes until some time in the future You write a program which copies the 100 byte file into a new 120 byte file and update the main parameter file to indicate that the new record size for the vendor file is 120 bytes instead of 100 Each program will now open the file using the new 120 byte record size since it is read in from the Parameter file at runtime but will only READ or WRITE the first 100 bytes of each record due to the variables used by the READ and WRITE calls Got the message 46 FILE I O STATEMENTS OPEN In order to manipulate data to or from a file it must be opened first The open statement assigns a unique file number to a file and also specifies the name that is to be given to an output file or to be used in the locating of an input file The general format is OPEN lt file gt lt filename gt lt mode gt record size record number var iable file any numeric expression which evaluates to an integer from 0 65535 0 is defined as the user terminal and treated as such filename any string expression which evaluates to a legal file description mode specifies the mode for opening the file INPUT opens an existing file for input op
60. e return will be suppressed so that the next PRINT statement executed at some later time will resume printing on the same line For compatibility with other popular BASIC implementations AlphaBasic also recognizes the single question mark as an alternate form of the PRINT verb PRINT USING is supported for formatted output and is described in another section reserved for that purpose alone The PRINT statement is also used for writing data to sequential files For 33 details on this refer to the section describing the file I O system FOR NEXT STEP These statements allow the execution of loops within the program and follow the same format and restrictions as other forms of BASIC The variable used may be subscripted if desired If no STEP modifier is used the step value is assumed to be a positive l The variable name may be amitted in the NEXT statement if desired in which case the previous FOR statement will be the one that is incremented All normal rules for nested loops and entering or exiting from within the loops apply here as in other BASIC versions FOR and NEXT statements are illegal in immediate mode FOR A 1 TO 10 FOR B Al Bl TO Cl STEP 2 FOR A 10 TO 1 STEP l IF THEN ELSE The conditional processing features in AlphaBasic give a wide variety of formats which duplicate just about all functions performed by other versions of BASIC The formats that are acceptable are IF lt relative expression gt THE
61. ed File type mismatch Device does not exist Bitmap kaput Disk not mounted File already exists Redimensioned array Illegal record number 56 SYSTEM FUNCTIONS AlphaBasic supports a unique group of operators called system functions which Provide the programmer with the ability to get to the I O ports physical memory enclosed within parenthesis The major difference is that a System function may appear on the left side of an assignment statement Whereby it represents an output or write condition to the System function System functions used within expressions on the right side of an assignment statement will perform an input or read operation and deliver back a result to be used in the expression evaluation IO X The IO system function allows the 256 I O ports to be Selectively read from or written to In both cases only one byte will be considered and an output expression greater than 255 will merely ignore the unused bits TO X A lt expr gt twrites the low byte of expr to port X IO X treads port X and places the result into A BYTE X and WORD X The BYTE and WORD system functions allow the programmer to inspect and alter any memory locations within the 65K memory addressing range of the machine These operations have often been called PEEK and FOKE statements in other implementations of BASIC The BYTE functions will deal with 8 bits of data in the range of 0 255 and the WORD functions will deal with 16 bits of da
62. ediate execution if no line number preceeds the command System commands result in immediate interpretation and execution and are considered those commands which result in a system fucntion but are not ever included in the text of a program itself Normal program statements may be entered without a line number in which case they will be compiled as a Single line program and then immediately executed Certain commands are considered illegal in immediate mode and the user is advised that no error messages are currently implemented to prevent the user from inadvertantly attempting to execute these commands In some instances the execution of these commands will result in complete system destruction which means you must reboot the entire operating system This malady will be corrected in a future release VARIABLE NAMES Variable names are not limited to a single letter or a letter and a digit as in conventional BASIC implementations A variable name may contain any number of alphanumeric characters as long as the first one is alpha A Z Apostrophes may also be used in variable names to improve clarity Mapped variables are defined by an explicit type code and therefore do not follow the standard convention of using a dollar sign for string variables Normal non mapped variables are considered floating point variables unless they are terminated by a dollar sign in which case they are considered string variables Subscripting follows the standard con
63. ence that file should also use a FILEBASE command 4 0 EXTENDED TAB FUNCTIONS Contrary to the statement in the AlphaBASIC manual the home position of the cursor the upper left hand corner is 1 1 NOT 0 0 In addition to the standard TAB 1 n functions listed in the manual the following are also available Code Function 17 Delete Character 18 Insert Character 19 Read Cursor Address 20 Read Character at Current Cursor Address 21 Start Blinking Field 22 End Blinking Field 23 Start Line Drawing Mode enable alternate character set 24 End Line Drawing Mode disable alternate character set 25 Set Horizontal Position 26 Set Vertical Position 27 Set Terminal Attributes Not all terminal drivers have all of the functions above simply because all terminals are not able to perform all of these functions If your terminal has additional features Alpha Micro recommends starting at 64 decimal when you assign function codes in your terminal driver 5 0 MAP STATEMENTS As of BASIC version 4 0 all MAP statements must appear at the front of a program before any executable code July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL l Page 3 6 0 LIBRARY SEARCHING Whenever a program called via RUN or CHAIN or a subroutine called via XCALL is requested BASIC follows a specific pattern in looking for the requested module If you specify a PPN then BASIC uses the currer default device and the specified PPN If you specify
64. ent routine can access the record in a different format by the different variable names in that format All record formats actually occupy the same area in memory This feature directly parallels the redefine verb in the COBOL language data division Normally a MAP statement causes allocation of memory to begin at the point where the last variable with the same level number left off The origin parameter allows this to be modified so that allocation will begin back at the base of some previously defined variable and therefore overlay the same memory area If the new variable is smaller than the previous one or the exact same size it will be totally contained within the previous one f it is larger than the previous one it will spill over into newly allocated memory or possibly into another variable area of the same level depending on whether there were more variables following it Play with this one a while to get the hang of it The origin parameter must be the last parameter on the line and takes the format TAG where TAG is a previously defined mapped variable on the same map level If size and value parameters are not included in this statement they may be omitted with no dummy commas if desired The following statements define three areas which all occupy the same 48 byte memory area but which may be referenced in three different ways 100 MAP ARRAY 110 MAP2 INDEX 8 F 200 MAP ADDRESS ARRAY 210 MAP2 STREET S 24 220 MAP2 CIT
65. ent specified If there are no arguments in the XCALL statement the argument list will consist only of the single count word containing the value of zero The format of each 3 word block describing one argument is as follows Word 1 variable type code Bits 0 3 contain the hex type code for the specific variable O unformated 2 string 4 floating point 6 binary 8 through E are currently unassigned Bit 4 is set to indicate the variable is subscripted or clear to indicate the variable is not subscripted Other bits in the type code word are meaningless Word 2 absolute address of variable in user impure area This address is the first byte of variable no matter what is type or size might be Word 3 size of the variable in bytes Note that the above descriptions also apply to the expression arguments with the exception that the results will be located above the address Specified by R5 instead of below it The argument list is built in free memory directly above the currently allocated user impure area R4 points to the word immediately following the last word in the argument list The user may scan the argument list and determine its end either by decrementing the count word at the base of the list or by scanning until the scan index reaches the address in R4 52 FREE MEMORY USAGE When the subroutine is called indexes R4 and R5 mark the beginning and end of the free memory that is Currently available for use as workspace
66. erations OUTPUT creates a file for output operations RANDOM opens an existing file for random read write The remaining two options must be used for RANDOM mode only record size an expression which specifies dynamically at runtime the logical record size for read write operations on the file record number variable a non subscripted numeric variable which must contain the record number of the desired random access for READ or WRITE statements when they are executed Any attempts to read or write to a file which has not been opened will result in an error message and the program will be aborted The filename string may be as brief as the name of the file in which case it is assumed to have an extension of DAT and reside in the current user s disk file area The filename string may expand to become a complete file specification if desired giving the explicit location of the file in another user area and even on another disk drive Some examples are OPEN 1 DATFIL INPUT OPEN 15 PAYROL TMP OUTPUT OPEN A C OUTPUT OPEN 3 DSK1 O0FILE ASC 200 20 OUTPUT OPEN 1 VENDOR DAT RANDOM 100 RECNUM The OPEN statement is the only statement which references the file by its actual ASCII filename in the standard Operating system format All further references in the program are made by the file number which is assigned in the OPEN Statement lt file gt expression CLOSE The CLOSE statement terminates the processing of
67. erfacing the software outputs to be straightforward This initial description is not meant to be the final result of the internal documentation effort but merely an introduction to the basic design theory behind the campiler All variables in use by the application program are stored in a dynamically alterable area within the user program area The areas that change dur ing program execution are those which are set aSide for arrays and variable size strings Simple variables do not change position once they are assigned storage positions Assignment of storage is normally done as each variable is encountered in the source program The user has the option of overiding this assignment by the memory mapping system which will be described later Each variable is assigned as one of the following data format types and once assigned may not be changed within the same program FLOATING POINT all numeric variables are assigned as this type unless otherwise specified in the program The standard precision in use by this system would probably be called single and one half since it lies midway between what has been accepted as single and double precision formats The reason for this is that the hardware floating point instructions all work in this format and so we may as well make use of the extra precision wherever Possible Floating point numbers occupy six bytes of storage and are in the same format as dictated by the hardware instructions FADD FSUB FMU
68. ertain commands are meaningless in immediate mode and are therefore not allowed FOR NEXT etc Multi statement lines are lines which contain more than one command Separated by colons and are allowed in immediate mode as well as in the source program A unique feature which is very useful for debugging is the single step command Each time the line feed key is entered with no data on the line the next instruction in the program is listed and then executed Control is then returned to the user terminal for inspection of variables or alteration of the source program before continuing execution Note that any alteration of the source program results in recanpilation before the single step command is actually executed Compiler mode differs fram interactive mode in that the program is compiled first and then the object code is stored on disk with a special save comand The object code may then be run at any time in the future requiring that only the run time package be in memory along with the object code itself Neither the compiler proper nor the program source code is required for execution in campiler mode resulting in a large cutback in memory requirement This mode is also ideal for sequential automatic processing of multiple programs within a system structure IN this mode BASIC programs may be freely mixed with assembly language programs for maximum efficiency and flexibility RUNNING BASIC PROGRAMS AlphaBasic supports the ability to capi
69. es Similarly the two statements at 300 and 310 produce the same two dimensional array as the statement at 200 200 DIM B1 5 20 300 MAP BX 5 310 MAP2 B1 20 F Inspect the following statements 400 DIM C1 10 410 DIM D1 10 500 MAP CX 10 510 MAP2 C1 F 520 MAP2 D1 F The statements at 400 and 410 produce two arrays each with ten variables The statements at 500 510 and 520 produce one array with twenty variables in it The variables will still be referenced as Cl 1 thru C1 10 and D1 1 thru D1 10 but their placement in memory is quite different The Cl variables will be interlaced with the Dl variables giving Cl 1 Dl 1 C1 2 Dl 2 C1 3 22 C1 10 D1 10 There are also ten unformatted variables CX 1 thru CX 10 which each contain the repective pairs of C1 Dl variables in tandem Referencing one of these CX variables will reference a 12 byte unformatted item composed of the C1 Dl pair of the same subscript This type of formatting would be useful in sophisticated techniques only The following define a more complex area 100 MAP ARRAY 110 MAP2 UNITX 5 120 MAP3 SIZA B 2 130 MAP3 SIZB B 2 140 MAP3 NTOT F 150 MAP3 FLAG 10 B 1 160 MAP3 CNAME S 20 170 MAP2 TOTAL F 180 MAP THING F 190 MAP WORK1 X 40 The area that is allocated by the above statements requires a total of 252 bytes of contiguous memory storage A total of 3 levels are represented in various formats Statement 100 defines a level
70. es are terminated either by the explicit size of the variable or by a null byte 0 if the string is shorter than the allocated size Mov ing a long string to a short one truncates all characters which will not fit into the new string variable Moving a short string to a long one causes the remainder of the long variable to be filled with null 0 bytes so that the actual data size of the string will be preserved for concatenation and printing purposes Floating point variables are the normal numeric variables used for mathematical calculations AlphaBasic supports only one and one half precision floating point variables which require 6 bytes of memory each and give 11 12 digits of precision Floating point variables must begin on an even address and the MAP statement processor will leave a blank byte between any floating point variable and the variable it follows in memory if that variable ended on an odd address This probably would only cause concern during the mapping of record structures for I O transfers Binary variables may range in size from 1 5 bytes giving from 8 39 bits of binary unsigned numeric data or 40 bits of binary signed data This is handy for the storage of small integer data such as flags in a single byte or for the storage of memory references as word values with a range of up to 65535 in two bytes Binary variables are manipulated just like floating point variables with conversions to and from the floating point format being aut
71. es just as in the normal form of the INPUT statement Also all carriage returns and line feeds will be bypassed allowing the file created by the PRINT statements to contain formatted line data if desired Commas spaces and end of line characters will all terminate numeric data strings and will then be bypassed PRINT Once a file has been opened for output the data is written to it by a special forn of the PRINT statement using a file number which corresponds to the number assigned in the OPEN statement All techniques usable in the normal form of the PRINT statment which outputs to the terminal may be used in the file form including PRINT USING for formatted data The data which is output to the file is in the exact format that would appear on the user terminal if the file number had been omitted The general format of the PRINT statement along with some valid examples follow PRINT lt file gt lt data list gt PRINT 1 A B C PRINT 4 USING AS A SOR A PRINT Q1 USING Al 10 PRINT 1 THIS IS A SINGLE LINE READ The READ statement is used to read a selected logical record from a file which has been opened for random access Processing The logical record READ lt file gt lt variable l gt lt variable 2 gt lt variable n gt The variables in the list may be any format but they obviously should match that of the designated record format The data will be read into the variables as unformatted bytes w
72. eturn numeric zero or null string values forever more This means that the normal sequence 38 for processing sequential input files would be to INPUT the data into the variables and then test the EOF X status before actually using the data in those variables EOF should only be tested for sequential input files Files open for output or for random processing will always return a zero value LEFT AS X or LEFT S AS x Returns the leftmost X characters of the string expression A RIGHT A X or RIGHTS AS X Returns the rightmost X characters of the String expression AS MID AS X Y or MIDS AS X Y Returns the substring composed of the characters of the string expression AS Starting at the Xth character and extending for Y characters A null string will be returned if x gt LEN AS LEN AS Returns the length of the string expression AS in characters INSTR X AS BS Performs a search for the substring BS within the string AS beginning at the Xth character position It returns a value of zero if BS is not in AS or the character position if B is found within AS Character position is measured from the start of the string with the first character position represented as one ASC AS Returns the ASCII decimal value of the first character in string AS CHR X or CHRS X Returns a single character string having the ASCII decimal value of x Only one character is generated for each CHR or CHRS function call
73. existing structures the data file must have an extension of IDA and all index files must have an extension of IDX There must be at 61 least one index file which is called the Primary index file There may also be additional index files called secondary index files which also link to the ETA primary data file The primary index file must always be opened in any program in order to gain access to the data file This is true even if you only intend to access the data file through one of the secondary index files in the current Program For information on file structures and operating the ISMBLD program refer to the ISAM System User s Guide SYMBOLIC AND RELATIVE KEYS Indexed files are accessed by one of two specific types of keys The relative key is already familiar to us since it is the same type of key used to access normal random files A relative key when used with an indexed file is used only to access a specific record in a data file The relative key must be a specific floating point variable which is specified in the OPEN statement The symbolic key is new to us and is used only with indexed files Symbolic keys are ASCII Secondary Symbolic keys are specified in the ISAM Statements when accessing the index file and are used to retrieve the relative key of the associated data record in the data file The concept of symbolic verses relative keys and their different uses is an important one and misuse of them will cause the ISAM syste
74. f intermediate results is always performed in full ll digit precision to minimize propagation of errors The significance of the system is initially set at 6 digits when the system is first started This is equivalent to standard single precision formats used in most of the popular versions of BASIC The significance is not reset by the RUN command and therefore may be set in immediate mode just prior to the actual running of a test program Of course any SIGNIFICANCE statements encountered during the execution of the program will reset the value STRSIZ The string size statement sets the default value for all strings which are encountered for the first time during the compilation phase Initially the default value of all strings in the absence of a STRSIZ statement is 10 bytes The statement STRSIZ 25 would cause all newly allocated strings which follow to have a maximum size of 25 bytes instead of 10 bytes This includes the allocation of string arrays The size value is evaluated at compilation time and therefore must be a single positive integer 35 RANDOMIZE Resets the random number generator seed to begin a new random number sequence Starting with the next RND x function call STOP Causes the program to suspend execution and print the message PROGRAM STOP AT LINE nnnn and then return control to the user in the interactive command mode The user may then continue with the next statement in sequence by executing a CONT command
75. f it is desired to read in the Sequence of a secondary file both the primary and secondary files must be opened Each record is retrieved by executing first an ISAM code 2 statement followed by a READ statement Remember that the READ statement must be to the primary file even though a secondary index file is being used for the sequential accesses After each ISAM code 2 statement the user should check for an end of file condition using the ERF X function to determine when no more data is left Refer to the section on error processing Close all files READING AN INDEXED FILE RANDOMLY BY KEY To read a file randomly by symbolic keys the files are opened as above for Sequential access As many secondary index files may be opened simultaneously as will be required for the random mode processing Each data record is located with an ISAM code l statement giving the symbolic key and the file number of the index file to which the symbolic key pertains Acheck should be made for a record not found error at this point indicating that the symbolic key was not located in the specified index file Assuming the key is valid a READ statement should be made to the primary file to get the correct data record This operation may be performed for each data record to be read Close all files 65 UPDATING DATA RECORDS Data records may be updated by locating the data record via either method above a and then updating the data buffer with the new data
76. g arrays may also be allocated such as DIM AS 5 The size of the array will depend on the current default string size in effect as specified by the last STRSIZ since each element in the array must be this number of bytes Arrays may be redimensioned during the execution of the program if desired The number of subscripts must remain the same but the number of elements in each level of the array may be changed Fach redimensioning of an existing array effectively erases the old array first and then allocates a new array with all elements zeroed out DIM A 10 DIM C 8 8 C 10 4 DIM TEST A B 4 DIM A B 4 SIGNIFICANCE The significance statement allows the user to dynamically change the default value of the numerical significance of the system for unformatted printing The significance value can be any value from 1 through 11 and will represent the maximum number of digits to be printed in unformatted numbers Rounding off to the specific number of digits will be performed only prior to the actual printing of the result The statement SIGNIFICANCE 8 will set the number of printable digits to 8 The value is interpreted at run time and therefore may be any valid numeric expression including variables if desired The current significance of the system is ignored when PRINT USING is in effect Note that the signifcance statement only affects the final printed result of all numeric calculations The calculations themselves and the storage o
77. he mapped variable for nesting purposes within other variables of a higher level It must be within the range of MAP1 through MAP16 As each mapped variable is assigned it will be automatically included as a part of those variables with a higher level Level numbers are actually backward with MAP being the highest level and MAP16 being the lowest or innermost level in the nesting scheme Levels need not be sequential as they are assigned That is a MAP5 statement may immediately follow a MAP2 statement without having dummy MAP3 and MAP4 statements intervening if desired This nesting scheme closely parallels the data division format in the COBOL language To eliminate potential allocation problems all MAP level variables will be forced to begin on an even address This allows insuring that certain binary and floating point variables will begin on word boundaries if desired for assembly language subroutine processing The AM 100 instruction set performs most efficiently when word data is aligned on word boundaries Also floating point variables will always be aligned to word boundaries VARIABLE NAME The variable name is the name to be given the variable for referencing within the program and must follow the rules for AlphaBasic variables Since the type may be explicitly specified the user need not follow the normal conventions such as requiring that the string variable names be followed by a dollar sign for identification If the variable n
78. he user memory The compiler itself may also be included in the monitor but due to the size of it this is not normally done unless the majority of the users will be doing program development and testing such as a school training program teaching BASIC To run in the interactive mode the mode most familiar to us due to other versions of BASIC the user types the command BASIC on his terminal and the compiler and runtime programs are both located and loaded into memory as required The LOAD command may then be used to load source programs into memory for editing compilation and test The SAVE command may be used to save either the source program or the campiled object program on disk for later use Source Programs have the extension BAS while compiled object programs have the extension RUN Note that there is no way to recover the source text from the compiled object code so the normal rule is to save both of them either on the same or different disks Running in the interactive mode always involves the compilation and running of a source program which is in memory and never includes running a saved object code program directly To run a compiled program the user must be in monitor mode Monitor mode may be entered from the interactive campiler mode above by exiting the compiler via the BYE command typed on the terminal Once the monitor mode is obtained the user enters the RUN command followed by the name of the object program to be run The
79. her single step functions Note that the single step function is performed by hitting the line feed key and not by actually entering the words single step BREAK This is a feature not normally found in other versions of BASIC which allows the user to set breakpoints on one or more line numbers in a program During execution if a line that has a breakpoint set on it is encountered the program will suspend execution and the message BREAK AT LINE nnnn will be printed The system will then be in interactive command mode to allow the inspection or 29 changing of variable values This suspension of execution occurs before the line that has the breakpoint set on it is executed There is no limit to the number of breakpoints that may be set in one program There is no additional overhead paid in execution speed when breakpoints are set Breakpoints may be cleared by placing a minus sign in front of the line number or by compiling the program which always clears all breakpoints If no line numbers follow the BREAK command all current breakpoints will be listed on the user terminal For example BREAK Lists all currently set breakpoints BREAK 120 Sets a breakpoint at line 120 BREAK 120 Clears the breakpoint at line 120 BREAK 120 130 40 500 Sets breakpoints at lines 120 130 40 and 500 BREAK 50 60 Clears the breakpoint at 50 and sets one at 60 Once a breakpoint has been reached the user may optionally continue the execution of the program by
80. iable AS BS C 2310 ABLE A 0QQ34 4 10 Q Q The mode independence feature allows substring modifiers to be applied to numeric expressions A string is returned but if the destination is a numeric variable another conversion will be made on the substring to return a numeric value Q A B 2 5 PRINT X 3 4 Substring modifiers may be applied to subscripted variables or expressions containing subscripted variables The substring modifiers are independent functions not to be confused with subscripts Q Qs AS 3 4 2 5 AS 1 B 3 5 3 oil Substring modifiers return a string value These may be used the same as strings in expressions QS AS BS 2 5 AS 2 2 C 5 3 Substring modifiers may be applied to the left side of an assignment in order to alter a Substring within a string variable Only that portion of the string defined by the substring modifier will be changed The other characters in the string will not be altered This may not be applied to numeric variables AS 2 4 QRS In the above example if A had contained ABCDEF before the assignment was executed the result in AS would be AQRSEF 17 MEMORY MAPPING SYSTEM AlphaBasic derives much of its versatility and speed from the fact that it is a true compiler in its execution mode Memory storage is allocated during compilation for all defined variables in an area that is contiguous and predictable All variables
81. id for the particular function l Illegal line number The specified line number is invalid e g not between 1 and 65534 Illegal or undefined variable in overlay l The variable specified in a MAP statement overlay via has not been previously defined or is not a mapped variable Illegal record number The relative record number specified in a random file processing statement i e READ or WRITE is either less than the current FILEBASE or outside of the file Illegal size for variable type The specified variable size is not valid for the particular variable type Floating point variables must be size 6 and binary variables must have size 1 5 Illegal subroutine name The name specified as a subroutine is not valid Illegal subscript The subscript expression is not valid Illegal type code The variable type code specified in a MAP statement is not one of the valid types Illegal user code The specified PPN was not found on the specified device or is not in a valid format Insufficient memory to load program xxxxxx The RUN program did not find enough free memory to be able to load the specified program Invalid filename The specified filename was not a legal filename Invalid syntax code An internal error has occurred in BASIC Please notify Alpha Micro of this error Provide an example of what caused it Line number must be from 1 65534 The line number entered was not in the range of legal line numbers Li
82. instance assume the logical record size is 100 and you need a maximum of 252 logical records in your file Each physical disk record is 512 bytes and therefore will contain 5 logical records You need 252 logical records so dividing 252 by 5 gives 50 full physical records plus 2 logical records remaining Since the file must be allocated in whole physical records you will need 53 physical records which will give you a maximum of 255 logical records These logical records will be referenced in your program as records 0 through 254 since the first record of any random file is numbered record 0 The logical record size is specified dynamically in the OPEN statement when the file type is RANDOM so it is possible to get things fouled up if you do not have the record size correct No logical record size is maintained within the file structure itself This fact does make it nice in one respect and that is that a file which is accessed by many programs can have its record size expanded without recompiling all programs Heres how Assume you have a file which is considered the parameter descriptor file for all other files in the entire system This file gives the record size as 100 bytes for the vendor name and address file as an example All programs which reference the vendor file first read this parameter file to get the size of the vendor file logical record The programs then set the size into a variable and use this variable in the OPEN statement for th
83. is is a powerful feature which can save much programing effort when used correctly There is an ambiguous situation which arises from this mode independence The plus symbol is used both as an addition operator for numeric operations and as a concatenation operator for string operations The value of 34 5 is equal to 39 but the value of 34 5 is equal to the string 345 The operation of the plus symbol is unambiguaous in its operation but may take a little thought to figure out its exact usage in all situations A few examples might help If the first operand is numeric and the second is string we convert the second to numeric and perform addition 34 5 equals 39 If the first operand is string and the second operand is numeric we convert the second to string and perform concatenation 34 5 equals 345 The above two examples apply only when we are not expecting a particular type of variable or term This generally occurs only in a first level PRINT expression At other times we are expecting a specific type of variable and the conversion of the first variable will be performed prior to inspecting the operator plus sign The operation of the plus sign is then implicitly specified by the result of the first first variable Take the following example 5 34 4 The multiplication operator forces us to expect a numeric term to follow The 34 string is therefore immediately converted to numeric 34 an
84. ithout regard to variable type The data will be transferred into each variable until the variable has been completely filled and then the next variable in the list will be filled and so on If the record is longer than the variable list specifies all excess data in the record will not be 49 transferred An attempt to transfer more data than is in the logical record size will result in an error message The most efficient use of the random files comes when the variable or variables used are mapped by the MAP statement to the exact picture of the record format in use WRITE The WRITE statement is used to write a selected logical record into a file which has been opened for random access processing The logical record which is trnasferred by the system I O is that whose record number is currently in the variable mentioned in the OPEN statement The format of the WRITE statement is WRITE lt file gt lt variable l gt lt variable 2 gt lt variable n gt The variables in the list may be any format but they obviously should match that of the designated record format The data will be written into the logical record from the user variables as unformatted bytes without regard to variable type The data will be transferred from each variable until the variable has been canpletely emptied and then the next variable in the list will be used and so on If the record is longer than the variable list specifies all excess data in the
85. l 512 byte block boundaries in the file Index files are arranged in a complex balanced tree structure and contain one symbolic key for each active data record plus a link to that data record in the data file This link is the relative record number and is used in the same manner as its counterpart in a normal random access file The index file also contains an array of internal links which comprise the sequential access tree structure Two references used in this manual may be confusing if they are not understood References to an indexed file are made when speaking of the entire file structure in general including the data file and one or index files References to an index file are used when specifically speaking of the portion of the structure which contains only the symbolic keys and the tree links Index files may be primary or secondary and have the extension IDX All indexed sequential files must be created by the ISMBLD program prior to access by any AlphaBasic program There is no method for the creation of a new indexed file within the AlphaBasic language since this would require a prohibitive amount of seldom used code for execution The user may however create indexed files in a system structure by using the feature that allows a BASIC program to create and then execute a command file This command file could set up parameters and then call the ISMBLD program to perform the actual creation of the files For compatibility with
86. le and rim programs without having to reread the source program by saving the compiled code on disk and using the runtime package to run that code at any later time There are two main components in the compiler system named BASIC PRG and RUN PRG These programs reside in the system library DSKO 1 4 and are called into memory as they are required BASIC PRG is the compiler and contains the code which scans the source program in memory and creates the runnable object program also in memory All system commands such as LIST LOAD SAVE etc are also processed in this Program RUN PRG is the runtime package and contains the code and subroutines which perform the actual execution of the compiled object code from memory This object code can be fram the compilation phase during interactive program development or it can be from the direct loading of a saved object code file on the disk The runtime package RUN PRG is free standing and may run without referencing the compiler program BASIC PRG The compiler however will first insure that the runtime package is present in memory and load it if necessary since it draws upon several of the internal runtime routines during compilation The user may optionally build his monitor with the runtime package resident for sharing by all users Those users which then run in the interactive mode will also benefit since BASIC PRG will locate RUN PRG in the monitor memory and not have to load it again into t
87. leted CONT The execution of the program is continued from wherever it last left off This is normally done after a control c interrupt a program STOP statement a breakpoint interrupt or a single step sequence A program may not be continued after it has come to the end of the statements CONTROL C Depressing the control and c keys simultaneously will interrupt any program that is Currently running and return to command interpretive mode If a program was being executed the line number about to be executed will be printed vai the message INTERRUPTED AT LINE nnnn The program may be continued by the CONT or single step commands or it may be restarted from the beginning by the RUN command SINGLE STEP line feed key The single step function is a feature not found in other versions of BASIC but is very useful in debugging programs and in teaching the principles of BASIC programing to newcomers The single step function causes the current line in the program to be listed on the user terminal and then executed Any output generated by the execution of a PRINT statement will then follow on the next line After the line has been executed the execution pointer is advanced to the next line and control returns to the user in the interactive command mode Successive single step commands may be used to follow the program through its paces Single step is legal after program STOP statements breakpoint interrupts control c interrupts and ot
88. ll the rules of the standard PRINT statement All characters in the formatting string will be printed as they appear except for the special formatting characters which will be described below The string is continually scanned over and over until the list of print items is exausted The formatting characters and their usage in the string are described in the following paragraphs EXCLAMATION MARK An exclamation mark in the format string causes the first character of the corresponding string variable to be printed in the corresponding space The rest of the string variable if it exceeds one character will be ignored BACKSLASHES Two backslashes in the format string define a string field whose size equals the number of characters enclosed by the brackets plus the brackets themselves Normal BASIC syntax dictates that the characters between the backslashes be spaces but AlphaBasic will accept any characters If the string variable to be printed exceeds the size of the specified string field the excess characters will be ignored If the string variable is shorter than the specified string field trailing blanks will be added to fill out the correct field size NUMERIC FIELDS Digit positions within formatted numeric fields are specified in the format string by the pound sign using one for each position desired both in front of and behind the decimal point One decimal point is allowed to define the explicit alignment of the digits
89. llocations which pertain to various areas in use by the compiler and may or may not be of use to you The byte counts returned for the various values of X are Free memory space remaining in current user partition Total size of current user partition Size of source code text area Size of user label tree Size of user symbol tree variable names and user function names Size of campiled object code area Size of data pool resulting from all compiled DATA statements Size of array index area dynamic links to variable arrays Size of variable storage area excluding arrays Size of file I O linkage and buffer area DONDAN PWNYOHO Note that the parameters for values above 1 may change as new versions of the compiler are developed Also some of these values will be meaningless when running the runtime object module in compiled mode EOF X The EOF function returns a value giving the status of a file whose file number is X The file is assumed to be open for sequential input processing The values returned by the EOF function are l if the file is not open or the file number X is zero 0 if the file is not yet at end of file during input calls l if the file has reached the end of file condition Due to the method used by the AMOS operating system for processing files the EOF status will not be valid until after an INPUT statement which reaches the end of file condition Any INPUT statements which reach EOF will r
90. lows data to be entered from the user terminal and loaded into specific variables at execution time The INPUT statement contains one or more variables Separated by commas When the INPUT statement is executed a single question mark is printed on the user terminal to signal the request for data entry Numeric variables require the data to be in one of the acceptable floating point formats String variables require the data to be an ASCII string of characters If multiple variables are used in one INPUT statement the user is expected to enter multiple values separated by commas sufficient to Satisfy the number of variables called out If insufficient data is entered a double question mark will be printed to signal the need for additional data INPUT Al INPUT BS CS Q 8 32 If the user enters a blank line carriage return only in response to a request for input the previous values of all variables will remain unchanged and the program will proceed to the next statement This effectively bypasses the entire input statement or the remainder of a partial data request If a control c is entered in response to a request for data the input statement is bypassed as with a carriage return and the program is interrupted at the next statement following the input statement The program may be resumed by the CONT or single step commands The user may cause his own prompt character or character string to be printed in place of the standard question mark b
91. ls deep but practicality dictates some reasonable limit of 20 or so Each level is referenced by a subscript value Starting with element 1 and extending to element N Once an array has been dimensioned by a DIM statement it may not be redimensioned by a subsequent DIM statement in the same program At ho time may the number of subscripts vary in any of the references to any element in the array The number of subscripts in each element reference must also match the number of subscripts in the corresponding DIM statement which defined the array size A tieg 10 DATA FORMATS AlphaBasic has been designed with the goal of flexibility in interfacing with other language processors within the AMOS operating system and primarily aimed towards assembly language programs where the data from a BASIC program must be manipulated in a way that is either infeasible or inefficient for another BASIC Program to do In order to accamplish this goal the data formats must be clearly defined and understandable by the user who wishes this information Data will normally be output to one or more disk files to be passed on to another BASIC program or to an assembly language program for further reduction or processing Data may also be output in a print image format for printing by the operating system print spooler job It is our intention to provide all required information on the internal formats and workings of AlphaBasic so that the user will find the task of int
92. m to malfunction in a number of ways Symbolic keys are used with the ISAM THE ISAM STATEMENT Indexed files are accessed by a special statement in AlphaBasic called the ISAM statement This statement has the general form ISAM lt file gt lt code gt lt symbol ic key gt Find a record by symbolic key Find the next sequential record in file Add a symbolic key to an index file Delete a symbolic key from an index file Locate next free data record in data file Delete a record from data file and return to free list DO Aa wN An error will result if an ISAM statement is executed with the value of lt code gt not equal to one of the above valid numbers The code may be any legal numeric 62 expression which is resolved at runtime OPENING AN INDEXED FILE As with other types of files an indexed file must be opened with a specific file number prior to any references to the file by other statements The OPEN statement follows the same format as that used by the normal random files OPEN lt file gt lt filename gt INDEXED lt record size gt lt relative key gt The filename must refer to the name given to the index file during the ISMBLD creation If this is a call to open a secondary index file the user must have already previously opened the corresponding primary index file on another file number so that the data file may be accessed AS an example assume that an indexed file structure consists of the
93. m to load a source program and then compile it for the CHAIN statement s use The object program module may however be already in memory or it may be on disk in which case it will be loaded automatically If the chained Program is already in memory there will be nho wait time due to loading from the disk The object program module may be loaded into memory by the monitor LOAD command different fran the AlphaBasic LOAD command Note also that the compiled object program is totally reentrant and sharable between several users so that any programs that are commonly used as either complete programs or chained links may be stored in monitor memory for high speed access ALPHABASIC VARIABLES Variable names may be any number of unique alphanumeric characters and are not limited to a single character or character digit pair as in other versions of BASIC The first character of the name must be alphabetic and the variable name must not begin with any reserved word used in BASIC Apostrophes may also be used in variable names to improve Clarity Mapped variables may take on any type format regardless of the name terminator Unmapped variables assume the type code as in other versions of BASIC String variables are specified by appending a dollar sign to the name and integer variables are specified by appending a percent sign to the name Refer to the section on data formats for more detail on these options NUMERIC VARIABLES The normal mode of
94. mmand clears out all current source code object code user symbols and variables It effectively initializes the compiler to accept new source program statements or immmediate mode statements LIST The source program if any is listed in numerical sequence on the user terminal If no line numbers follow the LIST command the entire program will be listed The listing may be aborted by entering control c on the terminal which will return the user to interactive command mode If one line number follows the LIST command only that line number will be listed If the command is followed by two line numbers separated by a comma space or other non numeric character only those lines which fall within the range bounded by the two numbers will be listed DELETE The DELETE command is used to delete groups of source lines from the program text If the command is followed by a single line number only that line will be deleted If the command is followed by two line numbers separated by a comma space or other non numeric character all lines of text which fall between the two line numbers inclusive will be deleted from the text 27 SAVE c The entire source program is saved on the disk in the user s file area The 4 user must enter the name of the program 1 6 characters following the SAVE command The program will be saved with the extension BAS and will be in ASCII format which may be listed or edited with the normal system programs outside of
95. n a one for one basis allowing direct editing of the source program in memory and testing of that program in a debug mode This is the mode that most BASIC interpreters operate in and the one that most users of BASIC are already familiar with In addition to the normal commands there are some unique features that make AlphaBasic more useful in a debug mode In interactive mode the entire compiler as well as the user source program must be stored in memory for operation Editing of the source program takes place in the conventional manner by typing each line with its line number first Lines are kept in sequence automatically by the internal editing routines Each time a change is made in the source program a switch is set which indicates that the program must be recompiled before it can be executed Any time the user enters a command which results in program execution RUN CONT or single step this switch is tested and if set the program is compiled in memory and the object code generated in its own special area of memory also It is this object code that is then executed by the run time package which also must be in memory during interactive operations l l i Commands which do not have a line number are considered immediate mode commands and are executed immediately Actually the statement is compiled and execution is applied against the currently existing set of defined variables New variables are defined as required by inmediate mode commands C
96. ne x not found The specified line was not found for a DELETE LIST etc operation July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 9 NEXT without FOR i A NEXT statment was encountered without a matching FOR statement No breakpoints set f BASIC is telling you that there are currently no breakpoints set in your program No source program in text buffer You tried to compile when there was no program in memory Operator interrupt You typed a Control C to interrupt program execution Out of data A READ statement was encountered after the data in all DATA statements had been used Out of memory op BASIC is telling you that it has run out of memory in which to execute your program Out of memory Compilation aborted COMPIL is telling you that it does not have enough free memory to finish compiling your program Program name You tried to SAVE or LOAD a program without providing a filename Enter the filename at this point Protection violation Your program tried to write into another account where you do not have write priviliges RESUME without error A RESUME statement was encountered but no error has occurred RETURN without GOSUB A RETURN statement was encountered but no corresponding GOSUB has been executed Record size overflow Your program tried to read a file record into a variable larger than the file record size Redimensioned array You tried to redimension an array
97. ning and ending positions or in terms of its beginning position and its length Substrings are defined by referencing the desired string followed by the substring modifier The substring modifier is two numeric arguments enclosed within square brackets The substring modifier takes on two distinct formats beg inning position ending position beg inning position substr ing length The first format defines the substring in terms of its beginning and ending positions within the string and uses a comma to separate the two arguments The second format defines the substring in terms of its beginning position within the string and its length using a semicolon to separate the arguments The second format basically performs the same function as the MIDS function The beginning and ending positions may are defined as character positions within the string relative to either the left or right end A positive value represents the character position relative to the left end of the string with character position 1 representing the first leftmost position A negative value represents the character position relative to the right end of the string with character position l representing the last rightmost position Assume the following string has the letters ABCDEF in it The positions are defined in terms of positions 1 through 6 left relative or positions l through 6 right relative A B C DEF 6 characters within main string 123
98. nitial work memory is 2310 bytes Phase 2 Adjust object Eile and process errors Illegal MAP level 350 MAP FILL 7 S 2 Syntax error 980 SLSMTD SLSMTD SSLAMT Memory usage Total work space 4712 bytes Label symbol tree 322 bytes Variable symbol tree 1186 bytes Data statement pool 0 bytes Variable indexi area 274 es Compiler work stack 140 bytes Excess available memory 11918 bytes Note that any error messages are reported during pass two and that the source line containing the error is typed on your terminal The Excess available memory line is useful for letting you know how close you are to running out of memory If you do run out of memory during a compilation July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 5 you see the message Out of memory compilation aborted and COMPIL returns your terminal to AMOS command level 9 3 Line Numbers Because AlphaBASIC allows the use of labels and because COMPIL assumes that you are using one of the text editors VUE or EDIT to maintain your source code line numbers are optional in source code given to COMPIL By omitting line numbers and with judicious use of indentation you can give source code a much more structured look than is normally possible in BASIC 9 4 Continuation Lines COMPIL allows the use of continuation lines within the source program This is especially useful for giving source code a structured appearance Specify
99. nner Integer constants are generated as their equivalent floating point values and are included only for compatibility with existing program structures Since integer variables are effectively floating point variables with an additional INT conversion performed they are actually slower than regular floating point variables This is opposite from most other BASIC s which usually store integer variables as 2 byte signed values and perform special integer arithmetic on them True integer variables may be defined by using the MAP statement and the B binary type code UNFORMATTED defines a fixed size area of storage used to contain absolute unformatted data which may be in any of the above formats This format will normally be used in the mapping system to define contiguous storage which is subdivided into multiple variables of different formats No conversion ever takes place when moving data to and from this format Unformatted variables are treated as string variables when used in expressions Note that dynamic length string variables are not yet implemented 12 ALPHABASIC EXPRESSIONS Expressions in AlphaBasic follow the same format used by other popular versions of BASIC Parentheses are used to designate hierarchy within expression terms and the normal mathematical hierarchy prevails in the absence of parentheses The following mathematical operators are recognized by AlphaBasic unary plus or addition unary minus or subtraction
100. no PPN the search sequence is as follows Default disk User P PN Default disk User P 0 DSKO 7 6 Note that earlier versions of BASIC pre 4 2 used a different search algorithm that was in reverse of the one outlined above 7 0 AUTOMATIC SUBROUTINE LOADING When a BASIC program calls a subroutine via an XCALL statement BASIC attempts to locate the subroutine in user or system memory If it is unable to do so it attempts to load the subroutine off the disk following the search pattern outlined above If a BASIC fetches a subroutine from disk BASIC loads it into memory only for the duration of its execution Once the subroutine has completed its execution it is removed from memory Therefore if a subroutine is to be called a large number of times it is ia load it into memory to avoid the overhead of fetching the subroutine rom disk 8 0 ADDITIONAL ERROR MESSAGES In addition to the error codes defined on page 56 of the AlphaBASIC manual two more error codes exist We give a complete list of all BASIC messages in Section 9 0 32 Invalid filename 33 Stack overflow 9 0 DISK COMPILER PROGRAM COMPIL To enable you to compile programs that are too large to fit into memory we provide a disk based compiler COMPIL COMPIL is a two pass compiler that gains memory space by omitting the interactive features of BASIC COMPIL produces RUN modules that are completely compatible with those produced by the interactive system J
101. ocation es of variables particularly in more canplex mapping schemes A few minutes at the terminal with immediate mode MAP statements followed by commands will help you see how the mapping scheme works 26 INTERACTIVE COMMAND SUMMARY Whenever AlphaBasic is not either campiling or executing a program it will be in interactive command mode which means it is waiting for a command from the user terminal to initiate some action The action taken depends on the type of input entered by the user which will fall into one of the following main categories 1 Edit command which begins with a valid source line number 2 Program statement for immediate execution no line number 3 Interactive system command resulting in a controlled action Edit commands will allow the creation and editing of a source program in memory on a single line basis The line number must be first followed by the program statement which will add the line to the source format in proper numerical sequence If the line contains only the line number the line will be deleted from the program No fancy character editing commands are implemented in this version Program statements without line numbers result in the immediate compilation and execution of the statement entered These program statements will be covered in another section The remainder of this section will briefly list the available interactive commands and the corresponding action performed NEW This co
102. ommand has the general format of an atsign followed by a variable name If the variable name is not followed by a subscript the system will search for the requested variable and print out all parameters about the variable for you on the terminal This may actually be two definitions since the variable A may actually be two different variables one which is a single floating point number and one which is a subscripted array The information returned about the variable will be the type of variable string binary etc the dimensions of the array if the variable is indeed an array the size of the variable in bytes and the offset to the variable from the base of the memory area used to allocate all variables If you enter a reserved word such as PRINT the system will tell you that the name is a reserved word Remember that the current system will not allow you to begin a variable name with a reserved word so that PRINTSUM is also considered a reserved word at this time 24 The general format of the definition line which is returned by the system is memory type variable type dimensions SIZE n location Memory type is the method of allocation used for the variable being defined FIXED variables are those which have not been defined by a MAP statement and were allocated automatically by the compiler when they were referenced in the program This is the normal method used by other BASIC versions to allocate variables DYNAMIC varia
103. or trapping by specifying a line number of zero or leaving the line number off completely ON ERROR GOTO 0 ON ERROR GOTO After executing the above form the program will print the standard error message and abort the program run A special case exists when the above statement is encountered within an error recovery routine prior to executing the RESUME statement In this instance the error trapping is disabled and the existing error is forced to be processed as if no error trapping were ever enabled It is recommended that all error trapping routines execute the ON ERROR GOTO 0 statement for all errors which have no special recovery processing Note that if an error occurs within the error trapping routine itself that error will be forced and the standard error message will occur There is no method to detect errors within the error recovery routine 54 RESUME STATEMENT The RESUME statement is used to resume execution of the program after the error recovery procedure has been performed The Statement takes on two forms similar to the ON ERROR GOTO statement The first form specifies a line number or label wihtin the program at which point the execution is to be resumed RESUME 410 RESUME TRY AGAIN The second form specifies a line number of zero or no line number at all and causes the execution to be resumed with the statement that caused the error to occur RESUME 0 RESUME Both forms cause the error condition to be cleared
104. perate properly with AlphaBasic with no modifications to either the compiler or to your programs Note that most terminals do not Support all of the above commands Commands that are not supported on the terminal in use will Merely be ignored by the driver 59 FORMATTING NUMERIC DATA VIA THE USING MODIFIER Often it is desirable to format a numeric value without having to immediately print it on the terminal or output it to a file The PRINT USING performs the formatting but the output must be to the terminal or a file which can cause extra code to be generated Formatted numeric data is handy for Creating print image lines and headers Some more exotic operations would allow pre inspection of numeric data and string manipulation of that data for specialized applications AlphaBasic allows formatting numeric data into its string equivalent with the USING expression modifier Basically the format for doing this is AS B USING C complex statement is legal AS ABCD B HOURS USING B 2 4 END Although it is academic in nature a restriction exists in that the USING modifier is not recursive i e the mask string itself cannot be the result of a USING modified expression 60 PROCESSING INDEXED SEQUENTIAL FILES AlphaBasic has the ability to process indexed sequential files by linking to the ISAM assembly language package which must reside wither in system memory or individual user memory Multiple directory files
105. point number For example the most common application is in a business environment where the scaling factor of 2 would be used to give absolute 11 place accuracy which extends 2 places to the right of the decimal point This means that the value of 50 12 will be multiplied by the scaling factor of 2 digits 100 and stored as the floating point value of 5012 Since this value is an integer it has absolute accuracy Just before printing this number it will be divided by the scaling factor to reduce it to its intended value of 50 12 and everybody is happy Other little conversions had to be included into the system to take care of all the little subtle effects of storing scaled numbers For example when converting scaled numbers to integer or binary format the number must be unscaled first before conversion When two scaled numbers are multiplied together the result is a number which must be unscaled once Division of two scaled numbers creates exactly the opposite problem Dealing with scaled numbers for exponential logarithmic and trigonometric functions creates even more exotic problems All these conversions are done automatically by AlphaBasic and so the programmer is relieved of the task of keeping track of them Scaled arithmetic will normally be entered at the start of a program and will continue in effect throughout the program The statement for setting the program into scaled mode is SCALE n The scaling factor n must be a decimal
106. port all features of conventional BASIC which is currently in use in the microprocessor field In those few areas where no mention is made here of a feature which is normally considered standard it is probably due to lack of space or possibly lack of time during the initial writing of this description The reader is encouraged to contact his local dealer to clear up any specific questions that may arise until such time as more comprehensive documentation is generated on AlphaBasic ALPHABASIC GENERAL INFORMATION This section will attempt to explain the general differences encountered in the AlphaBasic compiler system and to list the currently implemented features or problems and limitations that are known to exist at this time COMPILER OPERATION The user initiates the AlphaBasic compiler function by entering the command BASIC while in monitor command mode Once the compiler has been located and loaded into RAM it will print the READY message and await user input Although the system is a campiler in actual operation it has been designed to look as much like the popular interpreters that are currently on the market The system is interactive in nature and the user may enter his source program and edit it on a line number basis just like the interpretive versions No fancy editing techniques are yet available and each line must be changed by entering the line number and the entire new line Line numbers must be in the range of 1 65534 to be vali
107. processing mathematical variables as opposed to string variables is in ll digit accuracy which might be termed single and one half precision compared to normally accepted standards This is due to the hardware floating point instructions which are implemented in the WD 16 microprocessor chip set used in the AM 100 computer Integer and binary variables are also considered numeric variables but are always converted to floating point format prior to performing mathematical Operations on them All printing of math variables is done under normal BASIC format with the significance being variable under user control from l to 11 digits The SIGNIFICANCE statement is used to set up this value STRING VARIABLES AlphaBasic supports string variables in both single and array modes The Memory that is allocated for each string variable is the number of bytes representing the maximum size that the string is allowed to expand to Each string is variable in size within this maximum limit and a null byte is stored at the end of each string to indicate its current actual size At the start of each compilation the default size to be used for Strings is 10 characters maximum The STRSIZ statement may be used within the Program to alter the value to be used for all new string variables which follow Future releases of the compiler syntax will allow individual string definitions to have their own size specification without the need to change this default value
108. rmat defined by Alpha Microsystems for this date and it may be stored in any format you choose Some common methods are to pick a base date say 1 1 60 and store the date as the number of days since that date Another method is to store the Julian date with the year being offset by the appropiate integer amount The date will store a positive value of 2 32 or greater than 4 billion DATE A lt expr gt sets system date to expr DATE returns system date into A 1 2 Demo of use of Alpha Basic DATE routine with DATE formatted date 3 10 MAFI DATE HOLDER B 3 11 MAFI MMDDYY HATE HOLDER 12 MAP2 MM B 1 13 MAP2 DD 8 1 14 MAF2 YY B 1 15 PRINT t 20 NATE KOLDER PATE 21 PRINT USING Current date is A 4 H MM TD YY 22 PRINT 30 32 INPUT Enter new date MM DB YY MM DO YY 34 DATE DATE HOLDER 34 PRINT 38 GOTO 26 58 EXPANDED TAB FUNCTIONS The TAB function in AlphaBasic has been expanded past the normal usage to include terminal screen handling such as cursor control and other special functions To be used only in a PRINT statement the TAB function operates in the traditional manner when supplied with only a single numeric argument such as TAB X In this case the function Causes the carriage to be positioned over to the X column on the current line When supplied with two arguments such as TAB X Y however the TAB function performs special CRT functions If the value of X is positive the
109. s as do the level 3 items which comprise UNITX 1 thru UNITX 5 Following the above mess in memory come two more variables defined in statements 180 and 190 THING is a normal floating point variable which occupies 6 bytes and WORK1 is an unformatted area whose size is 40 bytes Note that since WORKI was not subdivided into one or more level 2 items a size clause was required to explicitly define its storage requirements Total storage used by the above series of statements 100 190 is 252 bytes Note that the variable UNITX 1 refers to the 40 byte item comprised of the variables in order SIZA 1 SIZB 1 NTOT 1 FEAG 1 1 thru FEAG 1 10 and CNAME 1 Moving the variable UNITX 1 to another area such as WORK will 23 transfer the entire 40 bytes with no conversions of any data Ib can be seen that although this mapping system includes advanced programming practices it provides the user with a degree of flexibility never before offered in a BASIC language implementation USING THE MAP STATEMENTS Map statments may be used in immediate mode as a learning tool to see how the variables are allocated They are not designed to be practical in the inmediate mode however and are best used by putting them into a program file and compiling the program In the immediate mode if an error occurs in the syntax of the statement the variable will have already been added to the tree and you will not be able to repeat the map statement again
110. se are never modified regardless of the scaling factor currently in use 44 ALPHABASIC FILE I O SYSTEM AlphaBasic supports both Sequential and random access disk files Data may optionally be written in ASCII or packed binary formats Files created by AlphaBasic programs are compatible with all other system utility formats and may be interchangeably introduced into and Manipulated by programs written in other languages Conversely files created by other languages and system utilities may be read and manipulated by programs written in AlphaBasic Files are created and referenced by the general statements OPEN CLOSE INPUT PRINT READ and WRITE All file references are done by a file number which may be any legal integer value from 1 to 65535 There is no absolute limit to the number of files that may be open at any given time in a program but since each file requires a certain amount of memory there is a practical limit to this number based on available memory The file number always follows the verb in any file I O statement and may be any legal numeric expression which is preceded by a pound sign File number zero is defined as the user terminal and is legal in file statements to allow generalized programs to be written which may selectively output to either a file or to the terminal at run time desired All file statements are valid in immediate mode but any open files are automatically closed before each new RUN command is executed
111. t You have specified a DELETE command without specifying what line s are to be deleted Device does not exist 4 The device you specified in a file operation OPEN LOOKUP etc does not exist Device driver must be loaded into user or system memory If you are accessing a non DSK device the appropriate device driver must be loaded into user or system memory Device error An error has occurred on the referenced device Device full The specified device has run out of room during a WRITE CLOSE or ALLOCATE operation Remember that an ALLOCATE requires contiguous disk space so that a Device full error may occur when there are still a number of non contiguous blocks available Device in use The specified device is currently assigned to another user Device not ready The specified device is not ready for use Disk not mounted The specified disk has not been mounted Mount it via the MOUNT monitor command or via the XMOUNT subroutine Divide by zero Your program attempted to perform a division by zero Duplicate label Your program has defined the same label name more than once End of Program You have reached the end of the program during single stepping July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 7 Enter lt CR gt to continue You have reached a STOP statement in your program You may continue from the STOP statement via a carriage return or may abort the run via a Control Cc File alrea
112. t describes the use of the PRINT USING and USING statements which allow you to format output into specific character positions by use of editing masks In addition to the masks mentioned there exists one type of mask that you can use to generate a number with leading zeros This mask takes the form of one standard numeric mask character followed by a series of Zs The total size of the output string will be the number of Zs plus the one For example PRINT 123 USING 22222 yields 000123 3 0 FILERASE During normal operation the first record in a random file is referred to as record number zero i e you set the record number variable to zero to access the first record in the file In some applications it is desirable to have this first record referred to by some number other than zero This is often done to allow you to use zero to flag some special condition such July 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 2 as a deleted record The FILEBASE command allows you to set the number used to refer to the first record to any value For example FILEBASE 1 tells BASIC that the first record in the file is record number one not record number zero You may use any numeric argument with FILEBASE Note that FILEBASE does not associate its value with a file but is only in effect in the program where it is executed If one program uses a FILEBASE command when referencing a file all other programs which refer
113. ta in the range of 0 65535 inclusive Any unused bits will be ignored with no error Message Note that these commands are not protected and it is possible to poke the operating system other users or yourself to death with improper use BYTE X lt expr gt writes the low byte of expr into memory loc X WORD X lt expr gt Iwrites the low word of expr into memory loc X A BYTE X treads memory loc X and places the byte into A A WORD X reads memory loc X and places the word into A TIME The TIME system function requires no argument and is used to set and retrieve the time of day as stored in the System monitor communications area The time is stored as a two word integer representing the number of clock ticks since midnight The Programmer is responsible for conversions to printable format in those cases where it is required One clock tick represents one interrupt from the CPU line clock which is usually 60 hz for domestic systems and 50 hz for overseas systems Dividing the time by the clock rate will give the number of seconds since midnight Converting this to current time is then accomplished by Successive division by 60 to get minutes and again by 60 to get hours 57 TIME lt expr gt sets time of day in system to expr TIME treturns time of day in clock ticks into A p gt DATE The DATE system function is identical to the TIME function except that it sets and returns the two word system date There is no current fo
114. te or merge several programs or subroutines together to be saved as a single program The separate routines Must not duplicate line numbers in the other routines that they are to be merged with or else the new line numbers will overlay the old ones just as if the file had been edited in from the user terminal The NEW conmand should be used prior to any load command if it is desired to insure that the text buffer is clear COMPILE The current source program in memory is compiled and the object code built up in another area of memory Control then returns to interactive command mode The compiled program is not executed Campilation effectively sets all variables to zero and deletes all variables that may have been generated as a direct result of immediate mode commands 28 RUN This is the normal way to initiate the running of the existing program in memory A check is first made to see if the program has been compiled since the last editing change to the source code and if it has not been an automatic compilation phase takes place to insure the object code is up to date All variables are reset to zero strings are reset to null and the compiled object code is then executed Execution may be interrupted at any time by typing a control c on the user terminal This control lt c status is tested only at the beginning of each new source line so multiple statement lines will not be interrupted until all statements on that line have been comp
115. thereby preventing files which were opened by an immediate statement to be written or read by statements within an executable program Under the current version of AlphaBasic each open file requires about 580 bytes of free memory for buffers and control blocks Future releases will allow more than one file to share the same buffer area thereby reducing the execution memory requirements SEQUENTIAL ASCII FILES Sequential disk files are the easiest to understand and to implement in AlphaBasic Data is written in ASCII format and all numerics are stored as ASCII string values Carriage returns and line feeds are included in the output file as a result of the print statement formatting but are bypassed when the file is read by another program The data files created by AlphaBasic sequential I O functions normally have the extension DAT unless otherwise explicitly stated in the OPEN Statement These files are normal ASCII sequential files in all respects and may be manipulated by the text editor the print spooler or any of the other system utilities Data is written to sequential ASCII files by using the PRINT statement with a file number non zero following the verb in standard I O statement format Data is read back from sequential ASCII files by using the INPUT statement ina Similar manner 45 RANDOM ACCESS FILES Random access files are more complex than sequential files but offer a much more fo flexible method for storing and retrie
116. uly 1979 ADDENDUM TO THE ALPHABASIC USER S MANUAL Page 4 9 1 COMPIL Operation To use COMPIL enter COMPIL filespec switches P where filespec selects the BAS file you want to compile and switches select one or both of the optional COMPIL options COMPIL allows the compilation of source files located on any device or in any account COMPIL always places the resultant RUN module in the account and device you are currently logged into for BASIC versions 4 2 and later BASIC does not support wildcarded filespecs 9 1 1 Operation Switches COMPIL allows the use of two switches O and T The O switch is the same as the O switch in the interactive compiler That is it tells COMPIL to omit line number references from the compiled code This reduces the total object code size but it prevents COMPIL error messages from reporting the line numbers where errors occurred The T switch is primarily designed for debugging purposes If you specify the T switch as COMPIL scans each source line of your program it displays that line on your terminal If a problem occurs during the compilation you can use the T switch to determine the line in which the problem is occurring You may also use this switch to gauge the compilation speed of various statements 9 2 Compiler Messages COMPIL reports a number of statistics on your terminal as it compiles A typical compilation might look something like this COMPIL ACMSLS Phase 1 I
117. value in the equivalent positive binary variable For instance a l byte binary may be loaded with positive numbers from 0 thru 255 decimal or negative numbers from l thru 128 but the negative numbers will be returned as the positive values of 128 thru 255 respectively Only 5 byte binary variables will return the original sign and value when loaded with a negative number AP AS Mi Binary variables may be used in expressions but they are slower than floating point variables because they are always converted first to floating point format before any mathematical operations are performed on them Binary variables are useful in integer and logical Boolean operations or for storing values in small amounts of memory floating point numbers always take 6 bytes of memory regardless of their values All logical operations performed within expressions AND OR XOR NOT etc cause the values to be first converted to signed 5 byte binary format before the logical operation is performed The value l represents a 40 bit mask of all ones and it is this value which is returned as the result of any relational comparison between two expressions or variables INTEGER integer variables and constants are specified by appending the variable name with a percent sign which is the standard convention in use by other BASIC s AlphaBasic generates floating point variables and performs automatic integer truncation for all integer variables specified in this ma
118. ventions of other BASIC s by enclosing the subscripts within parenthesis Some examples of legal variables follow A AS NUMBER STRINGS MASTER INVENTORY RECORD HEADER MOM ALWAYS LIKED YOU BEST 21234567 PROGRAM LABELS AlphaBasic allows the use of program labels to identify points in the program A program label is camposed of 1 or more alphanumeric characters of which the first must be alpha A Z Apostrophes may also be used within labels for Clarity Labels when used must be the first itme on a line and must be terminated by a colon A label may be followed by a program statement on the same line or it may be the only item on the line Labels Operate similar to line numbers for GOTO and GOSUB statements and make the program easier to document An example of label usage follows 10 START PROGRAM 20 INPUT ENTER TWO NUMBERS TO GET SUM A B 30 PRINT As sBs 24R 40 IF A B lt gt 0 GOTO SUM NOT ZERO 50 PRINT SUM IS ZERO 60 GOTO START PROGRAM 70 SUM NOT ZERO PRINT SUM IS NOT ZERO 80 GOTO START PROGRAM 90 END MEMORY ALLOCATION The compiler system allocates memory dynamically during editing compiling and execution of the user program Currently there is no check made to insure that the user is not running out of memory in his allocated partition and chances are pretty good that memory overrun will result in a system crash We realize that this is a grave restriction and there are several technic
119. ving data in different formats Random ae files are written in what is considered unformatted or packed data mode All program accesses to random files are made via the logical record approach A logical record is defined as a fixed number of bytes whose format is explicitly under control of the program performing the access Physical records on the disk are each 512 bytes long and each random file must be preallocated as some given number of these 51l2 byte physical records Logical records may be any length from 1 byte to 512 bytes in length The AlphaBasic I O system will automatically compute the number of logical records that will fit into one physical record and perform the blocking and unblocking functions for you For example if your logical record size is defined as 100 bytes then each physical record on the disk will contain 5 logical records with the last 12 bytes of each physical record being lost Therefore the most efficient use of random files comes when the logical record size will evenly divide into 512 bytes 32 64 128 etc a Random access files are preallocated once using the ALLOCATE statement and giving the number of physical 512 byte records to be allocated It is up to the Programmer to calculate the maximum number of logical records required in the file and then using the above description calculate how many physical records will be required to completely contain the number of logical records desired For
120. within the numeric field format Normally numeric fields are right justified with leading blanks being used to fill in for digit positions that are not required in front of the decimal point Unused digit positions behind the decimal point are filled with trailing zeros If the numeric field specified in the format string is too small to contain the numeric variable to be printed the field will be printed with a leading percent sign indicating the overflow This will be followed by the number in standard BASIC format 41 If the format field specifies any digit positions in front of the decimal point at least one digit is always output before the decimal point itself If necessary this digit is a zero pN Note that other special characters described in the following paragraphs also define numeric digit positions in addition to performing special formatting functions ASTERISK FILL If a numeric field in the format string begins with a double asterisk any leading spaces that would normally be output in front of a number will be replaced by asterisks This is quite useful in printing checks for instance The double asterisk also defines two digit positions FLOATING DOLLAR SIGN If a numeric field in the format string begins with a double dollar sign a dollar sign will be printed immediately preceding the first digit of the number The double asterisk also defines two digit positions one of which is taken up by the dollar
121. y enclosing the string in quotes immediately following the INPUT verb INPUT ENTER YOUR NAME A INPUT ENTER 3 VALUES FOR X Y AND 2 X Y Z The INPUT statement is also used for reading data fran sequential files For details on this refer to the section describing the file I O system INPUT LINE The INPUT LINE statement operation is identical to that of the INPUT statement with the exception that input into a string variable will accept the entire line up to but not including the carriage return and line feed This allows commas quotes blank lines and other special characters to be input without the need for quotes around them The INPUT LINE statement may be used in sequential file processing as well as the standard terminal input statement The question mark prompt character is never printed for an INPUT LINE statement but the user may include his own prompt string as in the INPUT statement above Some examples of the statement are INPUT LINE A INPUT LINE ENTER YOUR FULL NAME PLEASE NAME INPUT LINE 2 LINE OF INPUT PRINT The print statement performs the same as other versions of BASIC and will not be detailed extensively in this first printing Multiple variables or literals are printed on the same line separated by commas or semicolons Commas cause the next variable to be printed in the next zone while semicolons cause the variables to be printed with no Separating spaces If the line ends with a semicolon the carriag

AMOS AlphaBasic (old)

Contents

Download Pdf Manuals

Related Search

Related Contents