⟦e82a2d665⟧

TextFile

*** mytangle.web	Tue Mar  7 18:39:05 1989
--- tangle.web		Fri Jan  6 15:21:59 1989
***************
*** 12,18 ****
  % Version 2.7 fixed the definition of check_sum_prime (May, 1985).
  % Version 2.8 fixed a bug in change_buffer movement (August, 1985).
  % Version 2.9 allows nonnumeric macros before their definition (December, 1988).
- % RMD's attempts to improve error detection
  
  % Here is TeX material that gets inserted after \input webmac
  \def\hang{\hangindent 3em\indent\ignorespaces}
--- 12,17 ----
***************
*** 89,96 ****
  Other parts, delimited by |stat| and $|tats|$, are optionally included if
  statistics about \.{TANGLE}'s memory usage are desired.
  
! @d debug== {change this to `$\\{debug}\equiv\null$' when debugging}
! @d gubed== {change this to `$\\{gubed}\equiv\null$' when debugging}
  @f debug==begin
  @f gubed==end
  @#
--- 88,95 ----
  Other parts, delimited by |stat| and $|tats|$, are optionally included if
  statistics about \.{TANGLE}'s memory usage are desired.
  
! @d debug==@{ {change this to `$\\{debug}\equiv\null$' when debugging}
! @d gubed==@t@>@} {change this to `$\\{gubed}\equiv\null$' when debugging}
  @f debug==begin
  @f gubed==end
  @#
***************
*** 472,485 ****
  for i:=1 to @'37 do xchr[i]:=' ';
  
  @ The following system-independent code makes the |xord| array contain a
! suitable inverse to the information in |xchr|. The procedure |init_xord| is
! intended to set all entries of |xord| to |invalid_code|, so that out-of-range
! characters in the input can be detected.
  
- @d invalid_code= @'177
- 
  @<Set init...@>=
- init_xord( xord) ;
  for i:=first_text_char to last_text_char do xord[chr(i)]:=@'40;
  for i:=1 to @'176 do xord[xchr[i]]:=i;
  \f

@* Input and output.
--- 471,479 ----
  for i:=1 to @'37 do xchr[i]:=' ';
  
  @ The following system-independent code makes the |xord| array contain a
! suitable inverse to the information in |xchr|.
  
  @<Set init...@>=
  for i:=first_text_char to last_text_char do xord[chr(i)]:=@'40;
  for i:=1 to @'176 do xord[xchr[i]]:=i;
  \f

@* Input and output.
***************
*** 574,601 ****
  or |carriage_return|.
  
  @p function input_ln(var f:text_file):boolean;
! {inputs a line or returns |false|}
  var final_limit:0..buf_size; {|limit| without trailing blanks}
- c : text_char ;
  begin limit:=0; final_limit:=0;
!   if eof(f) then input_ln:=false
!   else  begin while not eoln(f) do
!     begin c := f^ ; get(f);
!       if xord[c] = invalid_code then
!       @<Complain about bad character, and discard it@>
!       else begin
!         buffer[limit] := xord[c] ;
!         incr(limit);
!         if buffer[limit-1]<>" " then final_limit:=limit;
!         if limit=buf_size then
!         begin while not eoln(f) do get(f);
!           decr(limit); {keep |buffer[buf_size]| empty}
!           print_nl('! Input line too long'); loc:=0; error;
!             @.Input line too long@>
!         end;
        end;
      end;
!     read_ln(f); limit:=final_limit; input_ln:=true;
    end;
  end;
  \f

@* Reporting errors to the user.
--- 568,589 ----
  or |carriage_return|.
  
  @p function input_ln(var f:text_file):boolean;
!   {inputs a line or returns |false|}
  var final_limit:0..buf_size; {|limit| without trailing blanks}
  begin limit:=0; final_limit:=0;
! if eof(f) then input_ln:=false
! else  begin while not eoln(f) do
!     begin buffer[limit]:=xord[f^]; get(f);
!     incr(limit);
!     if buffer[limit-1]<>" " then final_limit:=limit;
!     if limit=buf_size then
!       begin while not eoln(f) do get(f);
!       decr(limit); {keep |buffer[buf_size]| empty}
!       print_nl('! Input line too long'); loc:=0; error;
! @.Input line too long@>
        end;
      end;
!   read_ln(f); limit:=final_limit; input_ln:=true;
    end;
  end;
  \f

@* Reporting errors to the user.
***************
*** 1106,1114 ****
  @.Identifier conflict...@>
  for k:=byte_start[q] to byte_start[q+ww]-1 do print(xchr[byte_mem[w,k]]);
  error; q:=0; {only one conflict will be printed, since |equiv[0]=0|}
- print_nl( 'missing @@  ?' );
-   @.Missing \AT!  @>
- @<Show last checkpoint@>;
  not_found:
  end
  
--- 1094,1099 ----
***************
*** 1386,1401 ****
  "@@": print('@@@@');
  verbatim: print('@@=');
  force_line: print('@@\');
- and_sign: print('&');
- not_sign: print(' not ');
- set_element_sign: print(' in ');
- left_arrow: print(' := ');
- not_equal: print(' <> ');
- less_or_equal: print(' <= ');
- greater_or_equal: print(' >= ');
- equivalence_sign: print(' == ');
- or_sign: print(' or ');
- 
  othercases print(xchr[a])
  endcases
  \f

@* Stacks for output.
--- 1371,1376 ----
***************
*** 2053,2061 ****
    @!n:integer; {number being scanned}
  begin while stack_ptr>0 do
    begin cur_char:=get_output;
!   reswitch:
!   @<Semicolon-else test@>;
!   case cur_char of
    0: do_nothing; {this case might arise if output ends unexpectedly}
    @t\4@>@<Cases related to identifiers@>@;
    @t\4@>@<Cases related to constants, possibly leading to
--- 2028,2034 ----
    @!n:integer; {number being scanned}
  begin while stack_ptr>0 do
    begin cur_char:=get_output;
!   reswitch: case cur_char of
    0: do_nothing; {this case might arise if output ends unexpectedly}
    @t\4@>@<Cases related to identifiers@>@;
    @t\4@>@<Cases related to constants, possibly leading to
***************
*** 2546,2558 ****
  "<": control_code:=module_name; {beginning of a module name}
  "=": control_code:=verbatim; {beginning of \PASCAL\ verbatim mode}
  "\": control_code:=force_line; {force a new line in \PASCAL\ output}
! ",", ">", "!", "?", "|", "/", "#", "+", ";", "0","1","2":
!   control_code:=ignore; {Codes used by \.{WEAVE} but ignored by .{TANGLE}}
! othercases
! begin err_print('! Unknown control code: ', '@@', xchr[c]);
!             @.Unknown control code@>
!  control_code:=ignore;
! end;
  endcases;
  end;
  
--- 2519,2525 ----
  "<": control_code:=module_name; {beginning of a module name}
  "=": control_code:=verbatim; {beginning of \PASCAL\ verbatim mode}
  "\": control_code:=force_line; {force a new line in \PASCAL\ output}
! othercases control_code:=ignore {ignore all other cases}
  endcases;
  end;
  
***************
*** 2993,3001 ****
    a:=a mod @'400;
    end;
  module_name: if t<>module_name then goto done
!   else
!     begin @<Was an `@@ ' missed here? @>;
!     app_repl((cur_module div @'400)+@'250);
      a:=cur_module mod @'400;
      end;
  verbatim: @<Copy verbatim string from the buffer to |tok_mem|@>;
--- 2960,2966 ----
    a:=a mod @'400;
    end;
  module_name: if t<>module_name then goto done
!   else  begin app_repl((cur_module div @'400)+@'250);
      a:=cur_module mod @'400;
      end;
  verbatim: @<Copy verbatim string from the buffer to |tok_mem|@>;
***************
*** 3004,3013 ****
  @.\AT!p is ignored in Pascal text@>
  @.\AT!d is ignored in Pascal text@>
  @.\AT!f is ignored in Pascal text@>
!       ' is ignored in Pascal text: maybe @@  missing?');
!         @.Missing \AT!  @>
!       @<Show last checkpoint@>;
!       goto continue;
      end;
  new_module: goto done;
  
--- 2969,2975 ----
  @.\AT!p is ignored in Pascal text@>
  @.\AT!d is ignored in Pascal text@>
  @.\AT!f is ignored in Pascal text@>
!       ' is ignored in Pascal text'); goto continue;
      end;
  new_module: goto done;
  
***************
*** 3296,3435 ****
  error_message: print_nl('(Pardon me, but I think I spotted something wrong.)');
  fatal_message: print_nl('(That was a fatal error, my friend.)');
  end {there are no other cases}
- \f

@* Error detection.
- This section is an attempt to improve \.{TANGLE} by making it detect certain
- errors. Most of the additional code goes in here, so as not to disturb the
- module numbers of the rest of the program.
- 
- First: two consecutive modules in a \.{WEB} program must be separated by the
- string `@@ ' . If this is omitted, this error usually damages the \PASCAL
- output. If \.{TANGLE} does not detect this error, then the user will probably
- try to compile the damaged output. The likeliest result is that the compiler
- becomes confused, and prints lots of error reports that bear no clear relation
- to the cause of the trouble.
- 
- In order to try to detect this error, \.{TANGLE} now enforces an extra rule of
- syntax: if the replacement text of one module includes a call of a second
- module, then that call may not be immediately followed (on the same input
- line) by `=' or `+=' . This rule is not really valid for \.{WEB} programs: one
- might want to say:
- 
- $$ \hbox{if \AT!<the variable we are interested in \AT!>= 0 then...} $$
- 
- but this construction seems unlikely in \.{PASCAL}, and it can still be used
- if you put the equals sign on the next line.
- 
- @ Here is the code that performs the test. We have just read a module name
- within the replacement text of a module. |loc| should be pointing to the next
- character after the `@@>' that ends the module name. Note that we must look
- ahead without actually moving forward in the input file.
- 
- @<Was an `@@ ' missed here? @>=
- try_loc := loc;
- while (buffer[try_loc] = " " ) and (try_loc < limit) do incr(try_loc);
- if    (buffer[try_loc] = "+" ) and (try_loc < limit) then incr(try_loc);
- while (buffer[try_loc] = " " ) and (try_loc < limit) do incr(try_loc);
- if    (buffer[try_loc] = "=" ) then
-   begin
-     err_print('Nested named module definitions. Was an @@  missing?');
-       @. Nested named module @>@.Missing \AT!  @>
-     @<Show last checkpoint@>;
-   end;
- 
- @ This module will be called when \.{TANGLE} thinks it  has found a missing
- `@@ '.  It tries to help the programmer by indicating where the missing `@@ '
- should have been. It is not possible to tell the exact position, so \.{TANGLE}
- will only indicate a range of line numbers. The formal error is an illegal
- construction within the \.{PASCAL} code of the current module, and \.{TANGLE}
- says where it thinks that module started.
- 
- @<Show last checkpoint@>=
-     print_nl (' Pascal part of current module began at line ' , last_line:1 );
-     if was_changing then print(' in change file');
-     print_ln ( ' ');
- 
- @ The checkpoint is set at the start of the \.{PASCAL} code of each module,
- also at the beginning of the program.
- 
- @<Scan the \(definition part...@>=
-   last_line := line ;
-   was_changing := changing;
- 
- @ @<Set init...@>=
-   last_line := 0;     {place where current module started}
-   was_changing := false;
- 
- @ @<Glob...@>=
-   last_line, try_loc: integer ;
-   was_changing: boolean;
- 
- @ Another error is that of writing a program containing the sequence
- `|; else|',  which \.{PASCAL} does not allow. This is easy to do, especially
- if the semicolon and the |else| come from different modules. Although there is
- absolutely no reason why you cannot say `|; else|' in \.{WEB}, it is still
- desirable that \.{TANGLE} should detect this, as (on some machines) \.{TANGLE}
- runs much faster than the \.{PASCAL} compiler. To detect this error requires
- quite a bit of ``cookery''. The first step is to put an |else| into the
- |byte_mem|.
- 
- @<Glob...@>=
- point_else: name_pointer;
- semi_last: boolean ; {the last output was a semicolon}
- 
- @ @<Initialize the input system@>=
-     buffer[0] := "e" ; buffer[1] := "l" ;
-     buffer[2] := "s" ; buffer[3] := "e" ;
-     id_first := 0;
-     id_loc := 4 ;
-     point_else := id_lookup( normal);
-     buffer[0] := " " ;
-     semi_last := false ;
- 
- @ Then |send_the_output| tests for `|; else|' . This test is crude; it will
- not recognise `ELSE' or `Else' or the output from `el \@\& se'.
- 
- @<Semicolon-else test@>=
-   if    (cur_char = begin_comment)  or  (cur_char = join )
-     or  (cur_char = module_number)  or  (cur_char = 0 )
-     or  (cur_char = force_line)     or  (brace_level > 0 )
-   then do_nothing
-   else if cur_char = ";" then semi_last := true
-   else if semi_last and (cur_char = identifier) and (cur_val = point_else)
-     then err_print ('! semicolon-ELSE found')
-                 @.semicolon-ELSE found@>
-   else semi_last := false;
- 
- @ The next change is designed to catch out-of-range characters appearing in
- the input. On some systems (including VMS), characters in the range 128..255
- can cause very peculiar effects. Note that when a bad character is detected,
- the error message must not print that character, in case it messes up the
- terminal.
- 
- @<Complain about bad character...@>=
-   begin
-     if limit > 0 then loc := limit-1 ;
-     err_print
-     ('bad character in input, skipped, ordinal value is: ', ord(c): 1);
-       @.bad character in input...@>
-   end
- 
- @ A related difficulty is that \.{TANGLE} initialises only part of the |xord|
- array, namely the elements with ordinal values in the range 0..127. If the
- operating system allows characters with ordinal values outside that range,
- then the corresponding elements of |xord| will be  undefined. This procedure
- tries to prevent this happening by dynamically determining the bounds of the
- |text_char| type. If the local version of \PASCAL does not allow conformant
- array parameters, then the installer will need to find out the right values
- for |first| and |last| and write them into this procedure by hand.
-   @.system dependencies@>@.Conformant arrays@>
- 
- @<Error handling procedures@>=
-   procedure init_xord( var xx: array[first..last:text_char] of ASCII_code );
-   var c: text_char;
-   begin
-     for c := first to last do
-     xx[c] :=  invalid_code ;
-   end;
  \f

@* System-dependent changes.
  This module should be replaced, if necessary, by changes to the program
  that are necessary to make \.{TANGLE} work at a particular installation.
--- 3258,3263 ----
DataMuseum.dk

DKUUG/EUUG Conference tapes

⟦e82a2d665⟧ TextFile

Derivation

TextFile
