CWEB - release 4.10, 2023-08-19.

Fixes:

* Fix issue #43: 'out_buf' overflow on malign user input
  - Handle border case from bug report
  - Discard rest of input lines exceeding 'buf_size' characters
  - Purge 'ungetc' from the TeX part as well
* Fix typo in 'examples/wordtest.w' (reported by DEK)
* 'Ilk' is of type 'eight_bits'
* Avoid over-/underflow of section depth with malicious user input

Internal changes:

* Unroll 'init_p' helper function in COMMON
  - it is used for CWEAVE only (and CTWILL downstream)
* Match 'common.w' and 'common.h'
  - Reshuffle declarations and some code sections
  - Add starred section titles
* Demote all starred sections in 'common.w' (single star)
* Code cleanup
  - Make 'pop_level' a simple macro in CWEAVE
  - Use structure assignment (as in CTANGLE)
  - More compact 'stack' handling
  - Make use of type 'mode'
  - Purge a few 'goto' statements (and add some others)
  - Replace some 'sprintf' with 'snprintf'
  - Syntactic sugar for functional macros
  - Straighten 'byte_start' arithmetic
  - Desolve former section 19 of CWEAVE with internal array sizes
  - Hide 'print_text' analytics function from C compiler with DEBUG
  - Streamline 'phase_two' of CTANGLE
  - Refurbish 'phase_three' of CWEAVE
* Insert blank line after '\ch' in TeX output (cf. CWEB 3.0)

Macro changes

* Cleanup some '\acro' macros in 'cwebmac.tex'

.gitattributes

@@ -6,3 +6,5 @@ testthings.w export-ignore
 indentation.w export-ignore
 if.w export-ignore
 grün.w export-ignore
+out_buf-overflow.w export-ignore
+section_depth.w export-ignore

Makefile

@@ -1,6 +1,6 @@
 # This file is part of CWEB.
 # It is distributed WITHOUT ANY WARRANTY, express or implied.
-# Version 4.9 --- May 2023
+# Version 4.10 --- August 2023
 
 # Copyright (C) 1987,1990,1993,2000 Silvio Levy and Donald E. Knuth
 
@@ -205,6 +205,6 @@ tarball:
 	tar zcvhf /tmp/cweb.tgz $(ALL) examples
 
 ctan:
-	git archive -o ~/cweb-4.9.zip --prefix=cweb/ cweb-4.9
+	git archive -o ~/cweb-4.10.zip --prefix=cweb/ cweb-4.10
 	make PDF=pdf usermanual
-	cd ..; zip -r ~/cweb-4.9.zip cweb/cwebman.pdf
+	cd ..; zip -r ~/cweb-4.10.zip cweb/cwebman.pdf

README.md

@@ -1,4 +1,4 @@
-# cweb 4.9
+# cweb 4.10
 
 people have been accumulating several decades of experience with
 [cweb](http://www-cs-faculty.stanford.edu/~uno/cweb.html), and we know that
@@ -15,7 +15,7 @@ details in the
 [releases description](https://github.com/ascherer/cweb/releases).
 
 you can either compile these sources as directed in the
-[readme file](README) or stick the recent version “cweb-4.9” into my
+[readme file](README) or stick the recent version “cweb-4.10” into my
 [cwebbin](https://github.com/ascherer/cwebbin), which introduces convenient
 extensions and prepares cweb for inclusion in the “tex live” distribution.
 
@@ -27,8 +27,11 @@ incomplete – repository of original release versions of levy/knuth
 
 ## call for help
 
-recently i spotted versions 2.7 and 2.8 in
-[utah](http://ftp.math.utah.edu/pub/tex/pub/cweb/) – and there's even "the"
+recently i spotted versions 0.5, 2.7, and 2.8 in
+[utah](http://ftp.math.utah.edu/pub/tex/pub/cweb/),
+version 2.0/2.1 in
+[Fred Fish's Disk 551](https://ftp.funet.fi/pub/amiga/fish/501-600/ff551/),
+– and there's even "the"
 [original cweb 2.8](http://ftp.cs.stanford.edu/pub/cweb2.8) in stanford –, but
 the intermediate versions 3.0 and 3.1 are still missing in my collection. i
 could scrape them from

comm-bs.ch

@@ -15,8 +15,8 @@ by using "huge" pointers.
 (Update attempt by Andreas Scherer, 31 Jan 2021.  Good luck!)
 
 
-@x Section 9.
-  for section names */
+@x Section 10.
+@d ilk dummy.Ilk /* used by \.{CWEAVE} only */
 
 @<Common code...@>=
 typedef struct name_info {
@@ -25,7 +25,7 @@ typedef struct name_info {
   union {
     struct name_info *Rlink; /* right link in binary search tree for section
       names */
-    char Ilk; /* used by identifiers in \.{CWEAVE} only */
+    eight_bits Ilk; /* used by identifiers in \.{CWEAVE} only */
   } dummy;
   void *equiv_or_xref; /* info corresponding to names */
 } name_info; /* contains information about an identifier or section name */
@@ -34,7 +34,7 @@ typedef name_pointer *hash_pointer;
 extern char byte_mem[]; /* characters of names */
 extern char *byte_mem_end; /* end of |byte_mem| */
 @y
-  for section names */
+@d ilk dummy.Ilk /* used by \.{CWEAVE} only */
 
 @f huge extern
 
@@ -45,7 +45,7 @@ typedef struct name_info {
   union {
     struct name_info *Rlink; /* right link in binary search tree for section
       names */
-    char Ilk; /* used by identifiers in \.{CWEAVE} only */
+    eight_bits Ilk; /* used by identifiers in \.{CWEAVE} only */
   } dummy;
   union {
     void huge* equiv_member;
@@ -59,7 +59,7 @@ extern char huge* byte_mem_end; /* end of |byte_mem| */
 @z
 
 
-@x Section 39.
+@x Section 36.
     cur_file_name[l]='/'; /* \UNIX/ pathname separator */
 @y
     cur_file_name[l]='/'; /* A valid {\mc MSDOS} pathname separator */

comm-mac.ch

@@ -4,14 +4,14 @@ No changes to CTANGLE or CWEAVE are needed.
 (Contributed 13 Oct 2000 by [email protected]; slightly edited by Don Knuth)
 
 @x in limbo, change the title page document to specify Mac version
-  \centerline{(Version 4.9)}
+  \centerline{(Version 4.10)}
 @y
-  \centerline{(Version 4.9 for MacOS)}
+  \centerline{(Version 4.10 for MacOS)}
 @z
 
 @x section 23: Make input_ln accept \n, \r, \n\r, or \r\n as line endings
 @ In the unlikely event that your standard I/O library does not
-support |feof|, |getc|, and |ungetc| you may have to change things here.
+support |feof| and |getc| you may have to change things here.
 @^system dependencies@>
 
 @c
@@ -24,9 +24,9 @@ FILE *fp) /* what file to read from */
   limit = k = buffer; /* beginning of buffer */
   while (k<=buffer_end && (c=getc(fp)) != EOF && c!='\n')
     if ((*(k++) = c) != ' ') limit = k;
-  if (k>buffer_end)
-    if ((c=getc(fp))!=EOF && c!='\n') {
-      ungetc(c,fp); loc=buffer; err_print("! Input line too long");
+  if (k>buffer_end) {
+    while ((c=getc(fp))!=EOF && c!='\n'); /* discard rest of line */
+    loc=buffer; err_print("! Input line too long");
 @.Input line too long@>
     }
   if (c==EOF && limit==buffer) return false; /* there was nothing after
@@ -61,7 +61,7 @@ FILE *fp) /* what file to read from */
       return true;
     }
     else if (k>buffer_end) {
-      ungetc(c,fp); loc=buffer; err_print("! Input line too long");
+      while ((c=getc(fp))!=EOF && c!='\n' && c!='\r'); /* discard rest of line */
       return true;
 @.Input line too long@>
     }
@@ -81,17 +81,7 @@ FILE *fp) /* what file to read from */
   @<Skip over comment lines in the change file; |return| if end of file@>@;
 @z
 
-@x section 36, don't try to open a change file if none was specified
-if ((change_file=fopen(change_file_name,"r"))==NULL)
-       fatal("! Cannot open change file ", change_file_name);
-@y
-if (change_file_name[0] == '\0')  /* no change file specified */
-        change_file = NULL; /* reset at least the |change_file| */
-else if ((change_file=fopen(change_file_name,"r"))==NULL)
-       fatal("! Cannot open change file ", change_file_name);
-@z
-
-@x section 39, declare colon as Mac's path separator
+@x section 36, declare colon as Mac's path separator
 (Colon-separated paths are not supported.)
 The remainder of the \.{@@i} line after the file name is ignored.
 
@@ -104,12 +94,22 @@ The remainder of the \.{@@i} line after the file name is ignored.
 @d PATH_SEP ':'   /* MacOS pathname separator */
 @^system dependencies@>
 @z
-@x section 39, use the path separator constant
+@x section 36, use the path separator constant
     cur_file_name[l]='/'; /* \UNIX/ pathname separator */
 @y
     cur_file_name[l]=PATH_SEP; /* pathname separator */
 @z
 
+@x section 41, don't try to open a change file if none was specified
+if ((change_file=fopen(change_file_name,"r"))==NULL)
+       fatal("! Cannot open change file ", change_file_name);
+@y
+if (change_file_name[0] == '\0')  /* no change file specified */
+        change_file = NULL; /* reset at least the |change_file| */
+else if ((change_file=fopen(change_file_name,"r"))==NULL)
+       fatal("! Cannot open change file ", change_file_name);
+@z
+
 @x section 75, explain the convention for omitted change files
 An omitted change file argument means that |"/dev/null"| should be used,
 @y
@@ -134,7 +134,7 @@ An omitted change file argument means that no change file should be used,
 @z
 
 @x section 85, insert an extra module before the index
-@** Index.
+@* Index.
 @y by putting the new module here, we preserve all the previous section numbers
 @ We assume an interface to \CEE/ command-line emulation as supplied by
 the |ccommand| function of Metrowerks CodeWarrior, as defined in
@@ -144,5 +144,5 @@ the header file \.{console.h}.
 #include <console.h>
 @^system dependencies@>
 
-@** Index.
+@* Index.
 @z

comm-ql.ch

@@ -10,24 +10,24 @@ ex <dev_>cc;'-v -h -c -=500000 -DCWEBINPUTS=flp2_ common_c'
 @x
 \def\v{\char'174} % vertical (|) in typewriter font
 
-\def\title{Common code for CTANGLE and CWEAVE (Version 4.9)}
+\def\title{Common code for CTANGLE and CWEAVE (Version 4.10)}
 \def\topofcontents{\null\vfill
   \centerline{\titlefont Common code for {\ttitlefont CTANGLE} and
     {\ttitlefont CWEAVE}}
   \vskip 15pt
-  \centerline{(Version 4.9)}
+  \centerline{(Version 4.10)}
   \vfill}
 \def\botofcontents{\vfill
 \noindent
 @y
 \def\v{\char'174} % vertical (|) in typewriter font
 
-\def\title{Common code for CTANGLE and CWEAVE (QL Version 4.9)}
+\def\title{Common code for CTANGLE and CWEAVE (QL Version 4.10)}
 \def\topofcontents{\null\vfill
   \centerline{\titlefont Common code for {\ttitlefont CTANGLE} and
     {\ttitlefont CWEAVE}}
   \vskip 15pt
-  \centerline{(Version 4.9)}
+  \centerline{(Version 4.10)}
   \vfill}
 \def\botofcontents{\vfill
 \noindent
@@ -212,17 +212,15 @@ the routine as if it would work.
     @<Complain about argument length@>@;
   if (dot_pos==NULL) {
     sprintf(tex_file_name,"%s.tex",*argv);
-    sprintf(idx_file_name,"%s.idx",*argv);
-    sprintf(scn_file_name,"%s.scn",*argv);
     sprintf(C_file_name,"%s.c",*argv);
   } else {
     strcpy(tex_file_name,*argv);
     strcpy(C_file_name,*argv);
-    if (make_xrefs) { /* indexes will be generated */
-      *dot_pos='\0';
-      sprintf(idx_file_name,"%s.idx",*argv);
-      sprintf(scn_file_name,"%s.scn",*argv);
-    }
+    *dot_pos='\0'; /* string now ends where the dot was */
+  }
+  if (make_xrefs) { /* indexes will be generated */
+    sprintf(idx_file_name,"%s.idx",*argv);
+    sprintf(scn_file_name,"%s.scn",*argv);
   }
   found_out=true;
 }
@@ -233,17 +231,15 @@ the routine as if it would work.
     @<Complain about argument length@>;
   if (dot_pos==NULL) {
     sprintf(tex_file_name,"%s_tex",*argv);
-    sprintf(idx_file_name,"%s_idx",*argv);
-    sprintf(scn_file_name,"%s_scn",*argv);
     sprintf(C_file_name,"%s_c",*argv);
   } else {
     strcpy(tex_file_name,*argv);
     strcpy(C_file_name,*argv);
-    if (make_xrefs) { /* indexes will be generated */
-      *dot_pos='\0';
-      sprintf(idx_file_name,"%s_idx",*argv);
-      sprintf(scn_file_name,"%s_scn",*argv);
-    }
+    *dot_pos='\0'; /* string now ends where the dot was */
+  }
+  if (make_xrefs) { /* indexes will be generated */
+    sprintf(idx_file_name,"%s_idx",*argv);
+    sprintf(scn_file_name,"%s_scn",*argv);
   }
   found_out=true;
 }

comm-vms.ch

@@ -75,7 +75,7 @@ null device |"NL:"| should be used, when no changes are desired.
 @z
 
 @x section 85 (01-FEB-1992 ST)
-@** Index.
+@* Index.
 @y
 @* VAX/VMS specific code.
 We have used |SS$_NORMAL| and |SS$_ABORT| as return codes,
@@ -84,5 +84,5 @@ so we have to include the system message codes.
 @<Include files@>=
 #include ssdef /* we need VAX/VMS system messages */
 
-@** Index.
+@* Index.
 @z

comm-w32.ch

@@ -34,16 +34,10 @@ header files to be safe.
 Admittedly, this is not vital for cweb, except that I build the whole texlive
 set of programs using the __fastcall convention.
 
-@x
-extern boolean names_match(name_pointer,const char *,size_t,eight_bits);@/
+@x section 11
+extern boolean names_match(name_pointer,const char *,size_t,eight_bits);
 @y
-extern boolean __cdecl names_match(name_pointer,const char *,size_t,eight_bits);@/
-@z
-
-@x
-extern void init_p(name_pointer,eight_bits);@/
-@y
-extern void __cdecl init_p(name_pointer,eight_bits);@/
+extern boolean __cdecl names_match(name_pointer,const char *,size_t,eight_bits);
 @z
 
 @x section 75