Subject: texindex -- make an index from a LaTeX .idx file
Newsgroups: mod.sources
Approved: jpn@panda.UUCP
Mod.sources:  Volume 4, Issue 106
Submitted by: seismo!turtlevax!weitek!robert (Robert Plamondon)
Below is a shar file of my texindex program, which takes the (relatively
useless) .idx file that LaTeX produces, and sorts it on major minor entries,
reformats it, turns lists of consecutive page numbers into ranges, and 
kicks out a file that LaTeX can actually make an index from.
It consists of two awk files, a shell script to tie them together, and a
man page.
	Robert Plamondon
	UUCP: {turtlevax, cae780}!weitek!robert
	FidoNet: 143/12 robert plamondon
# This is a shell archive.  Remove anything before this line,
# then unpack it by saving it in a file and typing "sh file".
# Wrapped by weitek!robert on Fri May  2 09:39:36 PDT 1986
# Contents:  texindex.n texindex index.awk index1.awk
echo x - texindex.n
sed 's/^@//' > "texindex.n" <<'@//E*O*F texindex.n//'
@.TH TEXINDEX N "25 April 1986"
@.SH NAME
texindex \- Create an index for a LATEX document
@.SH SYNOPSIS
texindex
@.SH DESCRIPTION
@.I LATEX
produces
@.I .idx
files that contain the information that goes into the index.
@.I LATEX
also has a set of macros that are used to format an index. For some reason,
@.I .idx
file that
@.I LATEX
produces has no resemblance to the input file that it requires. In addition,
@.I .idx
file isn't sorted, doesn't have multiple page numbers per entry line, and
doesn't arrange subentries under the corresponding main entry.
@.I Texindex
takes a 
@.I LATEX \|.idx
file and converts it into a format that
@.I LATEX
recognize as an index. It puts multiple page numbers on the same line, and
handles subentries properly,
@.I i.e.,
	Gnus, 5-6, 25, 111
	\ \ \ Habits of, 5
	\ \ \ Smell, 25
Output is directed to standard out.
@.I Texindex
uses two
@.I awk
script and calls
@.I sort
to sort the index.
@.SH FILES
/usr/new/texindex	main program (a csh script)
/usr/lib/tex/index.awk	first awk script
/usr/lib/tex/index1.awk	second awk script
/tmp/texindex.xx	temporary file
@.SH BUGS
Blank space is left between entries for each letter of the
alphabet. If there are no entries for that letter, blank space
is left anyway, leaving larger gaps.
Should probably put large capitals before the entries for each letter, but
doesn't.
@//E*O*F texindex.n//
chmod u=rw,g=r,o=r texindex.n
echo x - texindex
sed 's/^@//' > "texindex" <<'@//E*O*F texindex//'
#!/bin/csh -f
# texindex -- create an index from a LaTeX .idx file
# uses the file index.awk
set INDEXAWK = /usr/lib/tex/index.awk
set INDEXAWK1 = /usr/lib/tex/index1.awk
set TEMP = /tmp/texindex.$$
cat $1 > $TEMP
cat >> $TEMP <<xxx
%\indexentry{%AZZZZZZ}{}
%\indexentry{%BZZZZZZ}{}
%\indexentry{%CZZZZZZ}{}
%\indexentry{%DZZZZZZ}{}
%\indexentry{%EZZZZZZ}{}
%\indexentry{%FZZZZZZ}{}
%\indexentry{%GZZZZZZ}{}
%\indexentry{%HZZZZZZ}{}
%\indexentry{%IZZZZZZ}{}
%\indexentry{%JZZZZZZ}{}
%\indexentry{%KZZZZZZ}{}
%\indexentry{%LZZZZZZ}{}
%\indexentry{%MZZZZZZ}{}
%\indexentry{%NZZZZZZ}{}
%\indexentry{%OZZZZZZ}{}
%\indexentry{%PZZZZZZ}{}
%\indexentry{%QZZZZZZ}{}
%\indexentry{%RZZZZZZ}{}
%\indexentry{%SZZZZZZ}{}
%\indexentry{%TZZZZZZ}{}
%\indexentry{%UZZZZZZ}{}
%\indexentry{%VZZZZZZ}{}
%\indexentry{%WZZZZZZ}{}
%\indexentry{%XZZZZZZ}{}
%\indexentry{%YZZZZZZ}{}
%\indexentry{%ZZZZZZZ}{}
sort -o $TEMP -bdfu -t\{ +1 -2 +2n $TEMP
awk -f $INDEXAWK $TEMP | awk -f $INDEXAWK1
rm -f $TEMP
@//E*O*F texindex//
chmod u=rwx,g=rx,o=rx texindex
echo x - index.awk
sed 's/^@//' > "index.awk" <<'@//E*O*F index.awk//'
# index.awk -- take a sorted LaTeX index, and produce \item and
# \subitem entries for it
# Robert Plamondon, March 1986
BEGIN	{FS = "{"
	print("\\begin{theindex}")}
# leave spaces on comment lines
$2 ~ /^%/		{print ""
			 print("\\indexspace")
			 print ""
			 next}
# Replace ! with \
#$2 ~ /^[\\]*[!]/		{while(index($2,"!") > 0)
#				 {x = index($2,"!")
#				 $2 = (substr($2,1,x-1) "\\" substr($2,x+1))}}
			{	$2 = substr($2,1,length($2)-1)
				newentry = $2
				newpage = substr($3,1,length($3)-1)
# Handle subentries (entries with commas in them)
		{comma = index($2,",")
		if (comma > 0)
		  subentry = substr($2,comma+1)
		  mainentry = substr($2,1,comma-1)
		  {
		   if (mainentry != substr(oldentry,1,comma-1))
# make new major entry
		    printf("\n\\item %s %s", mainentry, newpage)
		  else
		    if (oldentry == newentry)
	  	      {
			if (oldpage != newpage)
			printf(", %s", newpage)
		      }
		    else
		       printf("\n	\\subitem %s %s", subentry,newpage)
		   } 
		else	# no comma -- this is a major entry
			if (oldentry == newentry)
			  {if (oldpage != newpage)
				printf(", %s", newpage)}
			else
				printf("\n\\item, %s %s", newentry,newpage)
		{oldpage = newpage}
		{oldentry = newentry}
END	{print("\\end{theindex}")}
@//E*O*F index.awk//
chmod u=rw,g=r,o=r index.awk
echo x - index1.awk
sed 's/^@//' > "index1.awk" <<'@//E*O*F index1.awk//'
# index1.awk -- takes index entries in the form:
# entry, number, number, number...
# and turns consecutive numbers into ranges, i.e., 
# gnus, 5, 6, 7, 10, 11, 15
# becomes
# gnus, 5-7, 10-11, 15
# Robert Plamondon, April 25, 1986
BEGIN	{FS = ","}
 { if (NF > 1)
	{toprange = 0; field = 2; botrange = $field; field1 = 1+field}
	{ printf("%s",$1)
	  while (field < NF)
		  while ($field1 == 1+$field)
			{ toprange = $field1; field++; field1++}
		if (toprange != 0)
			{printf(", %d-%d",botrange,toprange)
			 lasttop = toprange
		else
			printf(", %d",$field)
		toprange = 0
		botrange = $field1
		field ++; field1++
	if (lasttop != $NF)
		if ($NF != 0)
			printf(", %d", $NF)
	printf("\n")
  else print}
@//E*O*F index1.awk//
chmod u=rw,g=r,o=r index1.awk
exit 0