#!/bin/bash
#
# Syntax:  tcheck [-h] [filename [...]]
#
# Purpose: Checks LaTeX files for the repetitions of words, for words that
#          occur in the file ~/info/priv/errors.lst, and for words or
#          parts of words that appear in ~/info/priv/danger_full.lst
#          and ~/info/priv/danger_part.lst respectively.
#          If no argument is given, the most recent tex file will be processed.
#          The -h flag shows the help info for this command.

#
# Notes: . This command file can easily be edited such that instead
#          of words also regular expressions are allowed (see below:
#          remove -F option of grep)
#          Note however that this will slow down the search.
#        . If your background is white then replace 1;32 by 0;32
#

# Created: Jul  6, 1994       by Bart De Schutter
# Last revised: Jan  2, 2011  by Bart De Schutter
#
# See http://www.deschutter.info/util/scripts.html for the latest version of
# this script.

# Status: public
# Category: latex

if [  $# -ge 1   -a   "$1" = "-h"  ];
then
   sed '/^$/q;/^#!/d;s/^# //;s/^#//' $0 >&2
   exit 1
fi

#
# Check for existence of awk and presence of lst files.
#

AWK=$(type -p gawk)
if [ -z "$AWK" ];
then
   AWK=$(type -p nawk)
   if [ -z "$AWK" ];
   then
      AWK=$(type -p awk)
   fi
fi
if [ -z "$AWK" ];
then
   echo >&2
   echo >&2 "WARNING: awk is not installed on this system."
   echo >&2 "         Check for double words will be skipped!"
   echo >&2
fi

ERRORS_LST=~/info/priv/errors.lst
DANGER_FULL_LST=~/info/priv/danger_full.lst
DANGER_PART_LST=~/info/priv/danger_part.lst

if [ ! -e $ERRORS_LST ]
then
   echo >&2
   echo >&2 "WARNING: $ERRORS_LST does not exist."
   echo >&2 "         Check for errors will be skipped!"
   echo >&2
fi
if [ ! -e $DANGER_FULL_LST ] || [ ! -e $DANGER_PART_LST ]
then
   echo >&2
   echo >&2 "WARNING: $DANGER_FULL_LST or "
   echo >&2 "         $DANGER_PART_LST does not exist."
   echo >&2 "         Check for 'dangerous' words will be skipped!"
   echo >&2
fi


# Perform tcheck

declare -a list   # Declare list as an array, necessary in order to be able
                  # to deal with files that have a space in their name.
if [ $# -eq 0 ];
then
   texfile=$( (ls -t *.tex | head -1) 2>/dev/null)
   if [ -n "$texfile" ];
   then
      list=$texfile
   else
      echo "No tex files found."
   fi
else
   list=("$@")
fi

firsttime=1
for fname in "${list[@]}"
do
   if [ ! -e "$fname" ];
   then
      fname="${fname%.tex}.tex"               # Append .tex extension if
                                              # necessary.
   fi
   if [ $firsttime -eq 0 ];
   then
      echo
   else
      firsttime=0
   fi
   echo
   echo "CHECKING file $fname"
   echo
   if [ ! -e "$fname" ]
   then
      echo
      echo "ERROR: File $fname not found."
      echo
   else
      if [ ! -z $AWK ]
      then
         echo
         echo "Checking for repetition of words."
         echo
$AWK '
BEGIN {
   previous_linenr = 0
   previous_line=""
   previous_word=""
   previous_fieldnr = 0
}
$0 !~ /^[ \t]*%/ {
   for ( i = 1; i <= NF; i++ )
   {
      if ( $i ~ /^%/ )
         next
      if ( $i !~ /^[a-zA-Z0-9{]/ )  #}
      {
         previous_word=""
         continue
      }
      current_word=$i
      gsub(/{|}|,$|\.$|\?$|!$/,"",current_word)
      if ( length(current_word) == 0 )
         continue
      if ( current_word == previous_word )
      {
         if ( previous_line != NR )
            print_line(previous_linenr,previous_line,previous_fieldnr)
            print_line(NR,$0,i)
            printf("\n")
         }
         previous_word=current_word
         previous_fieldnr=i
         previous_linenr=NR
         previous_line=$0
      }
   }

   function print_line(nr,string,fieldnr){
#
#  Commented out by BDS on Oct 26, 1998:
#                       nwords,words,space,label,defpos,pos,j){ #}
#
      nwords=split(string,words)
      words[fieldnr]="" words[fieldnr] ""
      space = ""
      label = sprintf("Line %d : ",nr);
   defpos = length(label)+1
   printf("%s", label)
   gsub(/./," ",label)
   pos = defpos
   for ( j = 1; j <= nwords; j++ )
      if ( pos + length(words[j]) < 81 )
      {
         pos += length(words[j])+1
         if ( j == fieldnr )
            pos -= 8
         printf("%s ", words[j])
      }
      else
      {
         pos = defpos
         printf("\n%s%s ",label,words[j])
      }
   printf("\n");
}
END{}' "$fname"
         fi
         if [ -e $ERRORS_LST ]
         then
            echo
            echo "Checking for words in ${ERRORS_LST##*/}"
            echo
            export GREP_COLOR='1;31'  # b;fg(30-39);bf(40-49)
            # Remove -F if you want to search on regular expressions
            # also.
            grep -F --color=always -n -i -f $ERRORS_LST "$fname"
         fi
         if [ -e $DANGER_FULL_LST ] && [ -e $DANGER_PART_LST ]
         then
            echo
            echo "Checking for words in ${DANGER_FULL_LST##*/} and"\
                 "${DANGER_PART_LST##*/}"
            echo
##
##            export GREP_COLOR='1;32'  # b;fg(30-39);bf(40-49)
##
            # Remove -F if you want to search on regular expressions
            # also.
            (export GREP_COLOR='1;32';\
             grep -F --color=always -n -w -i -f $DANGER_FULL_LST "$fname";\
             export GREP_COLOR='1;33';\
             grep -F --color=always -n -i -f $DANGER_PART_LST "$fname") \
            | sort -n
         fi
   fi
   echo
done