Club Amiga de Montreal - CAM/CAM_CD_1.iso/files/448a.lha/DECISION_PLUS_Demo/Syntax.Doc

Text File | 1990-10-29 | 10.9 KB | 274 lines

                         DECISION PLUS (TM) SYNTAX
----------------------------------------------------------------------


Syntax Notation:

      The '!' character must appear as shown. It indicates a declaration
      statement follows as opposed to general logic statements.
      The '[' and ']' (square braces) indicate optional items .
      The '<' and '>' (triangular braces) indicate manditory items .



Normal Sequence of events:

      The first step of a totaly new problem is to design and describe
      any new files to be used. This is done with the 'describe' program.
      No real trick here if you have already grasped how 'Decision Plus'
      describes fields in a file or the temporary fields in a program.

      If not the explanation follows:

      there are 5 main field types of 3 different sizes, but only certain
      cominations are allowed. The normal form of the declaration is :

        "name,[heading],type &size,format"


declaration in program                       meaning
-----------------------                     ------------
!temp xx_name,Name,a20,u               a 20 byte character(a=alpha) field
                                       called xx-name with a heading of
                                       'Name' forced to upercase(u)

!temp xx_date,Last Update,d4,dd/mm/yy  a date field called xx-date with a
                                       heading of 'Last Update' to be
                                       displayed/printed as day/month/year.
                                       (type&size for dates are always 'd4')
!temp xx_dtime,,m4,00.00               a time field called xx-time without a
                                       heading & displayed/printed as 'hh.mm'
                                       (type&size for times are always 'm4')

!temp xx_dollars,Credit,m4,####.##     a money field (must be m4 or m8) with
                                       heading Credit' and displayed with 2
                                       decimals up to "9999.99"

!temp xx_onorder,On Order,i1,##        a 1 byte integer showing 2 digits
                                       and can hold a number from 0 to 255

!temp xx_stock,In Stock,i2,####        a 2 byte integer showing four digits
                                       and can hold a number = +- 32767

!temp xx_sold,Total Sold,i4,&&&&&&#### a four byte integer called xx-sold
                                       displaying up to 10 digits and a
                                       number aprox +- 1.4 billion. NOTE
                                       this format blank fills any  leading
                                       zeros except for last 4 digits.

!temp xx_margin,Margin,r8,&&&###.####  a real number called xx_margin with
                                       heading of 'Margin' and displaying
                                       four  decimals Can coantain a number
                                       of approx +- 2.4 billion.

!temp xx_window,,i2[20],,100-2000      an array of 20, 2 byte integers with
                                       no heading or format but if used in
                                       sage program would limit input values
                                       for this field to numbers between
                                       100 and 2000.




      The describe program follows the same idea by asking you to enter
    values for each field in the file in teh same manner as the above
    examples for a program, except you will leave out the '!temp' part.
    The only trick to remember (cuz you dont have a proper manual) is the
        1) all key fields must declared first and in the order that you
           want the index to be properly sorted
        2) You must hit reurn the 'filed-name' entry point after the last
           key field to indicate to describe that you atre switching to
           the data field section of the record and the same after the last
           data field to indicate all fields are entered and you want to
           get to the command menu for further operations on your data or
           to save and then exit

     The best way to see this hamlessly is to open a newcli or ashell,
     whatever and (with the path of these commands added to your list)
     type in "describe". You'll see the usage message. Then try "describe
     sxx.d". when it comes up read the message and notice the "Key Fields"
     line. Enter a return and then notice "Data Fields". Enter a return
     and you should see the command line. Type for exit. Next type in
     'describe +l /demo/diary' to see a list of the 'diary' file
     organization. ('describe >PRT: +l /demo/diary.d' to output on your
     printer) . Next try 'describe /demo/diary.d' to bring up the program.
     When used on an existing descriptor it will go straight to the menu
     edit mode. Try the 'list', then the 'abandon' options.


     Once you have a descriptor file you and describe like, you need to
     create the actual data and key (index) files by using the 'newkf'
     utility. So type in "newkf <your-filename>" and it will come back with
     "'your-filename' created". Now try some of the other utilites on
     your file to see what they return such as the following:

     kfcheck <your-filename> - checks the index for corruption
     kfdet   <your-filename> - tells you the keylenght amd record length
                               of your file.
     kfri    <your-filename> - rebuilds the index of your file


  Other program syntax
  ---------------------

     describe filename.d              (descriptor file editor)

     reformat old-filename new-filename  (assumes new-filename is created
                                        but empty


     cf filename[.f]                  (produces a pcode file filename.g)
     cr filename[.r]                  (produces a pcode file filename.q)

     sg filename[.d]                  (produces a source file filename.f)
     rg filename[.d]                  (produces a source file filename.r)

     sage filename.g    [arg1 arg2]   ( run a sage program)

     sagerep filename.q [<printer> arg1 arg2] .....

     menu menuname.m                  (display a textfile called menuname.m
                                       and execute commands from it.
                                       NOTE preceed all comnmands with a '-'
                                       as in the examples without
                                       redirectiion (see BUGS))



    Ok now we have a database file and teh simplest way to create a
    program to maintain this file is to type in "sg <your-filename>".
    The 'sg' program will create the source code file for an editor of
    your-filename and then complie it using 'cf'. Now you typer in 'sage
    your-filename" to run the editor program. Similarily the 'rg' program
    will create a list program for 'your-filename' and you may run it by
    typing in "sagerep your-filename".  While 'sg' and especially 'rg'
    do not create elegant programs they are a very speedy way of getting
    all the field names (and their headings in the case of rg) into a text
    file that you can edit and recomplie as much as your ol' head desires.


    Note if your file contains so many fields that 'sg' can not figure out
      how to fit them on the vaailbale screen you can use the '-a' option to
    interactively design what you will display and enter for this program.
    This option is also available for 'rg'.

    Whether you build the program via 'sg/rg' or from scratch with an editor
    'Decsion Plus' (TM) allows you to overwite the '.d' descriptor heading
    and/or format within any particular program.


    REVIEW the examples in '/DEMOS', they will help but if you find this
    language as much fun and as useful as I, then a manual is a must for
    building greater things. It is not practical to reproduce all of here.





------------------------ tear here ----------------------------------




Sage/Sagerep Syntax Synopsis:  ( statements starting with '!' appear in the
                                 same order as they should in a program -
                                 not absolutely necessary but good style,
                                 and all temps declared within a program
                                 are global ie accessable even from a i
                                 subroutine)

!file   logical-filename/number [path/]name (driving file in sagerep)
!cfile  logical-filename/number [path/]name (closed file after startup)
!xfile  logical-filename/number [path/]name [key=field_name]
                           cross_reference file or "joined file"

!record file-name/number [path/]name (alternate record layout of file )

!temp   name,[title],type&size,[format],[limits]

!init   expression (occurs before first record read (sagerep only))
!title  expression (occurs after  first record read (sagerep only))

!scroll x,y   starting scroll array area an number of lines (sage only)

!depth  n : depth of page in lines (sagerep)
!width  n : width of page in columns (sagerep)
!gap    n :

!constant name,[heading],type&size[,format]='value'

!input 'string-prompt',field-name (sagerep only)

!startrec key=field-list
!startrec key=field-list

!select  if (field-name    <conditional expression>....)
!exclude if (if field-name <conditional expression> .....)

!on starting field-name (in driving file, sagerep only)<expression> ......
!on ending   field-name (in driving file, sagerep only)<expression> ......

!footnote <statement [: statement] ...... >(sagerep only)
!final    <statement [:statements] ......> (sagerep only)

. ... Any line starting with a period is considered a comment line
. Any regular statement ending with a backslash ('\') is considered
.  to be continued on the next line.

!display (message-string ) sagerep only

 Logic Flow
   if ..then ... [else ....]   (same line or continued line with a '\')
   gosub [label],
   goto  [label]

 Keyed File opearations
   check    logical-file [nrs=label] (nrs - no record currently selected )
   close    logical-file
   delete   logical-file [nrs=label nrs=label]
   insert   logical-file [re =label ( re - record already exists on insert)
   next     logical-file [nsr=label riu=label] (riu-record in use(locked))
   open     logical-file
   prev     logical-file
   preserve logical-file (preserve current record data on clear)
   rewind   logical-file [nrs=label]
   trynext  logical-file [nrs=label]
   readkey  logical-file [nsr=label riu=label]
   testkey  logical-file [nsr=label] (nsr-no such key found in file)
   write    logical-file [nrs=label] (nrs-no current rec selected for update)

 Common
   exit  - exits program cleanly.
   abort - exits program immediately.
   end   - indicates the end of a major logic section or option
   sleep - cause program to sleep or suspend for n seconds.

 Screen only

   check logical-file :do we currently have record in hand from file 'file'?
   display  field-list
   highlite field-list
   clear    field-list
   input    field-list [eoi=label ni=label] (eoi= end of input char &
                                   ni= no input)
   message "string expression" -displays message at bottom of screen)
   error   "string expression" - displays error message at bottom of screen
   scroll n                    - move to position 'n' in array
   prompt  "string expression" -displays prompt at bottom of screen and
             waits for 'y' or 'n' input

 Report Only

    print field-list or quoted string
    keep n lines               - do 'top of form' an "depth-n" lines

 Arithmetic

    +(add), -(subtract), /(divide), *(mulitply), %(modulo)

 Comparitive
     <, >, =, <>, <=, >=

------------------------- the end -----------------------------------