tools/textviewer.sh

   1 #!/usr/bin/env bash
   2
   3 # textviewer.sh
   4 # Copyright 2003 Luke Plant <L.Plant.98@cantab.net>
   5 # and Johann Koenig <johann@mental-graffiti.com>
   6
   7 # This program is free software; you can redistribute it and/or
   8 # modify it under the terms of the GNU General Public License
   9 # as published by the Free Software Foundation; either version 3
  10 # of the License, or (at your option) any later version.
  11 #
  12 # This program is distributed in the hope that it will be useful,
  13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15 # GNU General Public License for more details.
  16 #
  17 # You should have received a copy of the GNU General Public License
  18 # along with this program; if not, write to the Free Software
  19 # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
  20 # 02111-1307, USA.
  21
  22 ##############################################################################
  23 #
  24 # This script is a text viewer designed to be used with claws-mail actions
  25 # Set up an action with the command line:  textviewer.sh %p |
  26 #
  27 # The script will try to detect file type automatically, and then
  28 # invokes a relevant program to print the file in plain text to
  29 # the standard output.
  30 #
  31 # From v 0.9.7claws7, claws-mail sets the temporary file
  32 # of a part to XXXXXX.mimetmp.[filename of attachment]
  33 # This means we can use the extension of the filename for checking.
  34 # Also use the program 'file' if that fails.
  35 #
  36 # To extend the script just follow the patterns that already exist, or
  37 # contact the author if you have problems.
  38
  39 ##############################################################################
  40 #
  41 # Change Log
  42 #
  43 # 2003-03-25
  44 #       - make extension matching case insensitive
  45 #
  46 # 2003-03-23
  47 #       - Support for MS Excel (xlhtml) and Powerpoint (ppthtml)
  48 #
  49 # 2004-03-09
  50 #       - Support for HTML (html2text)
  51 #
  52 # 2004-02-13
  53 #       - added support for perl and shell scripts, and recognize that
  54 #         'file' will always return 'text' somewhere in its output for
  55 #         files that, well, contain text
  56 #
  57 # 2004-01-25
  58 #       - added brief messages describing whats going on
  59 #
  60 # 2004-01-23
  61 #       - added support for 'pdftotext,' from xpdf-utils debian package
  62 #
  63 # 2004-01-05
  64 #       - added matcher and action for OpenOffice Writer documents
  65 #         (requires ooo2txt)
  66 #
  67 # 2004-01-05
  68 #       - changed page width parameter for antiword
  69 #       - fixed matcher for 'diffs'
  70 #       - added a matcher and action for bzip2 - bzip2 files
  71 #         are decompressed and textviewer.sh run on the result
  72 #       - similarly decompress gzip files and run textviewer.sh
  73 #         on the result, insteading of doing 'gzip -l'
  74 #
  75 # 2003-12-30
  76 #       added the script to claws-mail/tools
  77 #
  78 # 2003-12-30
  79 #       - use 'fold' after 'unrtf' to wrap to a nice width
  80 #       - added basic file sanity checks
  81 #
  82 # 2003-12-29
  83 #       Added recognition for "Zip " from 'file' output
  84 #
  85 # 2003-12-19
  86 #       Initial public release
  87 #
  88 ###############################################################################
  89
  90 if [ $# -eq 0 ]
  91 then
  92         echo "No filename supplied." >&2
  93         echo "Usage: textviewer.sh FILE" >&2
  94         exit 1
  95 fi
  96
  97 [ -f "$1" ] ||
  98 {
  99         echo "File \"$1\" does not exist or is not a regular file." >&2
 100         exit 1
 101 }
 102
 103 [ -r "$1" ] ||
 104 {
 105         echo "Cannot read file \"$1\"." >&2
 106         exit 1
 107 }
 108
 109 FILETYPE=`file --brief "$1"` ||
 110 {
 111         echo "Please install the command 'file' to use this script." >&2
 112         exit 1
 113 };
 114
 115 FNAME=`echo "$1" | tr [A-Z] [a-z]`
 116 case "$FNAME" in
 117         *.doc)  TYPE=MSWORD     ;;
 118         *.ppt)  TYPE=POWERPOINT ;;
 119         *.zip)  TYPE=ZIP        ;;
 120         *.tar.gz|*.tgz) TYPE=TARGZ ;;
 121         *.tar.bz2|*.tar.bz)     TYPE=TARBZ ;;
 122         *.gz)   TYPE=GZIP       ;;
 123         *.bz2|*.bz)     TYPE=BZIP       ;;
 124         *.tar)  TYPE=TAR        ;;
 125         *.diff) TYPE=TEXT       ;;
 126         *.txt)  TYPE=TEXT       ;;
 127         *.rtf)  TYPE=RTF        ;;
 128         *.sxw)  TYPE=OOWRITER   ;;
 129         *.pdf)  TYPE=PDF        ;;
 130         *.sh)   TYPE=TEXT       ;;
 131         *.pl)   TYPE=TEXT       ;;
 132         *.html|*.htm) TYPE=HTML ;;
 133         *.xls)  TYPE=EXCEL      ;;
 134 esac
 135
 136 if [ "$TYPE" = "" ]
 137 then
 138         case $FILETYPE in
 139                 *"HTML"*)       TYPE=HTML ;;
 140                 *"text"*)       TYPE=TEXT ;;
 141                 gzip*)          TYPE=GZIP ;;
 142                 bzip2*)         TYPE=BZIP ;;
 143                 "POSIX tar archive"*)   TYPE=TAR        ;;
 144                 "Zip "*)        TYPE=ZIP  ;;
 145                 "Rich Text Format"*)
 146                                 TYPE=RTF  ;;
 147         esac
 148 fi
 149
 150 case $TYPE in
 151         TARGZ)  echo -e "Gzip'd tarball contents:\n"    ;
 152                 tar -tzvf "$1"                          ;;
 153
 154         TARBZ)  echo -e "Bzip'd tarball contents:\n"    ;
 155                 tar -tjvf "$1"                          ;;
 156
 157         BZIP)   TMP=`mktemp "$1".temp.XXXXXXX` || exit 1;
 158                 bunzip2 -c "$1" > "$TMP"  || exit 1;
 159                 echo -e "Re-running \"$0\" on bunzip'd contents of \"$1\":\n";
 160                 "$0" "$TMP";
 161                 rm "$TMP"                                       ;;
 162
 163         GZIP)   TMP=`mktemp "$1".temp.XXXXXXX` || exit 1;
 164                 gunzip -c "$1" > "$TMP"  || exit 1;
 165                 echo "Re-running \"$0\" on gunzip'd contents of \"$1\":\n";
 166                 "$0" "$TMP";
 167                 rm "$TMP"                                       ;;
 168
 169         TAR)    echo -e "Tar archive contents:\n"       ;
 170                 tar -tvf "$1"                           ;;
 171
 172         ZIP)    echo -e "Zip file contents:\n"          ;
 173                 unzip -l "$1"                           ;;
 174
 175         RTF)    which unrtf > /dev/null  2>&1 ||
 176                 {
 177                         echo "Program 'unrtf' for displaying RTF files not found" >&2
 178                         exit 1
 179                 };
 180                 echo -e "Displaying \"$1\" using \"unrtf\":\n";
 181                 unrtf -t text "$1" 2>/dev/null | egrep  -v '^### ' | fold -s -w 72  ;;
 182
 183         TEXT)   cat "$1"                                ;;
 184
 185         MSWORD) which antiword  > /dev/null  2>&1 ||
 186                 {
 187                         echo "Program 'antiword' for displaying MS Word files not found" >&2
 188                         exit 1
 189                 };
 190                 echo -e "Displaying \"$1\" using \"antiword\":\n";
 191                 antiword -w 72 "$1"                             ;;
 192
 193         POWERPOINT) which ppthtml > /dev/null 2>&1 ||
 194                 {
 195                         echo "Program 'ppthtml' for displaying Powerpoint files not found" >&2
 196                         exit 1
 197                 };
 198                 which html2text > /dev/null 2>&1 ||
 199                 {
 200                         echo "Program 'html2text' for displaying Powerpoint files not found" >&2
 201                         exit 1
 202                 };
 203                 echo -e "Displaying \"$1\" using \"ppthtml\" and \"html2text\":\n";
 204                 ppthtml "$1" | html2text                        ;;
 205
 206         EXCEL) which xlhtml > /dev/null 2>&1 ||
 207                 {
 208                         echo "Program 'xlhtml' for displaying Excel files not found" >&2
 209                         exit 1
 210                 };
 211                 which html2text > /dev/null 2>&1 ||
 212                 {
 213                         echo "Program 'html2text' for displaying Excel files not found" >&2
 214                         exit 1
 215                 };
 216                 echo -e "Displaying \"$1\" using \"xlhtml\" and \"html2text\":\n";
 217                 xlhtml "$1" | html2text                        ;;
 218
 219         OOWRITER) which ooo2txt > /dev/null 2>&1 ||
 220                 {
 221                         echo "Program 'ooo2txt' for converting OpenOffice Writer files not files not found" >&2
 222                         exit 1
 223                 };
 224                 echo -e "Displaying \"$1\" using \"ooo2txt\":\n";
 225                 ooo2txt "$1"                                    ;;
 226
 227         PDF) which pdftotext > /dev/null 2>&1 ||
 228                 {
 229                         echo "Program 'pdftotext' for converting Adobe Portable Document Format to text not found" >&2
 230                         exit 1
 231                 };
 232                 echo -e "Displaying \"$1\" using \"pdftotext\":\n";
 233                 pdftotext "$1"  -                               ;;
 234
 235         HTML) which html2text > /dev/null 2>&1 ||
 236                 {
 237                         echo "Program 'html2text' for converting HTML files not found" >&2
 238                         exit 1
 239                 };
 240                 html2text -nobs "$1" ;;
 241
 242         *)      echo "Unsupported file type \"$FILETYPE\", cannot display.";;
 243 esac