Utilities and demo programs for Pascal on the Amstrad CPC Computers
HiSoft Pascal 4T has its own line editor and stores all files in a binary format to save space. Unfortunately the line editor is very painful to use and there is no option to import or export ASCII files from another text editor.
In fact, a good deal of the file content is actually ASCII text, but reserved words are tokenized and indenting spaces are compressed into a single byte per line. Actual compaction depends a lot on how much indenting is used, but in general the resulting binary files are around 25% smaller than the ASCII equivalent.
The CPM version was provided with a couple of utilities to convert files to and from ASCII text: FROMAMS.COM, TOAMS.COM. These aren't greatly convenient to use with the AMSDOS version or if editing on a modern machine and storing files on .dsk images for use with emulators or USB/SD card systems.
The two utilities here written in Python3 do the same task as TOAMS and FROMAMS on Linux or MacOS:
- asc2pas.py
- pas2asc.py
Both will translate from the AMSDOS file format to and from plain ASCII text. Optionally both can strip or add line numbers (necessary only for Hisoft's built in editor rather than the language itself) so that syntax colouring can be used in Emacs or other editors.
NAME
asc2pas.py - ASCII Text to HSP Translation
USAGE
asc2pas.py [-a|--ascii]<filename> [-p|--pascal]<filename> [-l|--addlinenum] [-v|--verbose] [-h|--help]
MANDATORY SWITCHES
-a|--ascii <filename> specify input file of ASCII text
OPTIONAL SWITCHES
-p|--pascal <filename> specify output filename. If omitted this defaults to the input filename with '.p' appended
-l|--addlinenum add line numbers
-v|--verbose writes information to stdout as translation progresses
-h|--help show this help message
EXAMPLES
python3 asc2pas.py -a sphere.asc -p sphere.p
NAME
pas2asc.py - ASCII Text to HiSoft Pascal 4T Translation
USAGE
pas2asc.py [-a|--ascii]<filename> [-p|--pascal]<filename> [-n|--nolinenum][-v|--verbose] [-h|--help]
MANDATORY SWITCHES
-p|--pascal <filename> specify input file
OPTIONAL SWITCHES
-a|--ascii <filename> specify output filename. If omitted this defaults to the input filename with '.asc' appended
-n|--nolinenum remove line numbers
-v|--verbose writes information to stdout as translation progresses
-h|--help show this help message
EXAMPLES
python3 pas2asc.py -a sphere.asc -p sphere.p -n
Create the Sphere demo program in your favourite Linux/MacOS editor
(* Sphere or Woolball demo after Acornsoft BASIC original *)
program sphere;
const
xo = 300;
yo = 200;
sc = 200;
var
n, x, y : integer;
i : real;
procedure scrsetmode(mode : integer);
begin
ra:=chr(mode);
user(#bc0e)
end;
procedure gramoveabs(x,y :integer );
begin
rde:=x; rhl:=y;
user(#bbc0)
end;
procedure gralineabs(x,y :integer );
begin
rde:=x; rhl:=y;
user(#bbf6)
end;
procedure graclearwindow;
begin
user(#bbdb)
end;
procedure klsettime;
begin
rde:=0; rhl:=0;
user(#bd10)
end;
begin
scrsetmode(0);
graclearwindow;
gramoveabs(xo, yo);
for n := 0 to 504 do
begin
i := n * 0.25;
x := round(sc * sin(i));
y := round(sc * cos(i) * sin(i*0.95));
gralineabs(x+xo,y+yo);
end
end.
Save it as sphere.p
Convert it from ASCII to Hisoft Pascal format, adding line numbers and then pop it oon a fresh DSK image
python asc2pas.py -a sphere.p -p sphere.hsp -l
iDSK new.dsk -n
iDSK new.dsk -i sphere.hsp
Fire up Hisoft Pascal on an emulator or better still real hardware and mount the disk image. Then compile and run
>g,,sphere.hsp
>C
>R
The file has the following general format
| Section | Comment |
|---|---|
| File size | 2 byte file size |
| [line][line]..[line] | Program body made up of multiple line sections |
| end of program | 2 byte end of program marker |
| padding | variable number of bytes to pad out to multiple of 128 bytes |
| end of file | 3 byte end of file |
This is a two byte field describing the total number of bytes in the file in 128byte sections.
The low byte (the first byte) holds the number of bytes in the final section.
The high byte (second byte) holds the number of 128 byte sections in the file, numbered starting at 1.
byte0 = file_len MOD 128
byte1 = (file_len DIV 128) + 1
Each line entry starts with a two byte field for the line number and ends with an 0x1D character to mark the end of line.
| Byte | Field | Comment |
|---|---|---|
| 0 | linenum lo byte | low byte = linenum MOD 256 |
| 1 | linenum hi byte | high byte = linenum DIV 256 |
| 2 | number of spaces indentation | one byte |
| 3..N | line text | multiple bytes |
| N+1 | 0x0d | end of line marker |
A single space is always assumed between line number and the first character of the line text, so the number of spaces in the field here is actually one less than will be shown on screen in the Hisoft editor.
Line text is mainly simple ASCII characters, except that reserved words are tokenized using the following table.
| Keyword | Token | Keyword | Token | Keyword | Token | Keyword | Token |
|---|---|---|---|---|---|---|---|
| PROGRAM | 0x81 | VAR | 0x8A | REPEAT | 0x93 | ARRAY | 0x9C |
| DIV | 0x82 | OF | 0x8B | CASE | 0x94 | FORWARD | 0x9D |
| CONST | 0x83 | TO | 0x8C | WHILE | 0x95 | RECORD | 0x9E |
| PROCEDURE | 0x84 | DOWNTO | 0x8D | FOR | 0x96 | TYPE | 0x9F |
| FUNCTION | 0x85 | THEN | 0x8E | IF | 0x97 | IN | 0xA0 |
| NOT | 0x86 | UNTIL | 0x8F | BEGIN | 0x98 | LABEL | 0xA1 |
| OR | 0x87 | END | 0x90 | WITH | 0x99 | NIL | 0xA2 |
| AND | 0x88 | DO | 0x91 | GOTO | 0x9A | PACKED | 0xA3 |
| MOD | 0x89 | ELSE | 0x92 | SET | 0x9B |
Reserved words in comments or quotes are not tokenized.
The end of program marker is made up of two zero bytes. Effectively these are a line number of zero as the parser progresses past the last actual line of program.
Each file is padded out to a multiple of 128 bytes. The actual padding character seems not to matter, but zeroes are used by asc2pas.py.
After the padding, three more bytes are added to mark the end of file. If the file (including padding) is 128 bytes long, then these will be bytes 129,130 and 131. The bytes are
[0x00] [0x00] [0x1A].