tris/recoll
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

*** empty log message ***

Browse Source
This commit is contained in:
dockes 2005-02-02 17:57:08 +00:00
parent d8b2533102
commit 69f56860e8
1 changed files with 73 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

73
src/filters/rclps Executable file
View File

@ -0,0 +1,73 @@
#!/bin/sh
# @(#$Id: rclps,v 1.1 2005-02-02 17:57:08 dockes Exp $ (C) 2004 J.F.Dockes
# Parts taken from Estraier:
#================================================================
# Estraier: a personal full-text search system
# Copyright (C) 2003-2004 Mikio Hirabayashi
#================================================================
#================================================================
# rclps
# Extract text from a postscript file by executing pstotext or ps2ascii.
#
# The default is to use pstotext which can deal with accents.
#
# OTOH, ps2ascii is much faster, comes with ghostscript, and sometimes work
# better (ie: on some openoffice output files).
#
#================================================================
# set variables
LANG=C ; export LANG
LC_ALL=C ; export LC_ALL
progname="rclpdf"
decoder=pstotext
#decoder=ps2ascii
# show help message
if test $# -ne 1 -o "$1" = "--help"
then
printf 'Convert a postscript file to unformatted HTML text.\n'
printf 'Usage: %s [infile]\n' "$progname"
exit 1
fi
infile="$1"
# check the input file existence
if test ! -f "$infile"
then
printf '%s: %s: no such file\n' "$progname" "$infile"
exit 1
fi
# output the result
$decoder "$infile" |
awk '
BEGIN {
printf("<html><head><title></title>\n")
printf("<meta http-equiv=\"Content-Type\" content=\"text/html;charset=UTF-8\">\n")
printf("</head>\n<body><p>");
esc = 1
}
{
if ($0 ~ /-$/) {
sub(/-$/, "", $0)
printf("%s", $0);
} else if($0 == "\f") {
printf("</p>\n<hr>\n<p>")
} else {
if(esc > 0) {
gsub(/&/, "\\&amp;", $0)
gsub(/</, "\\&lt;", $0)
gsub(/>/, "\\&gt;", $0)
}
print $0
}
}
END {
printf("</p></body></html>\n");
}' | iconv -f iso-8859-1 -t UTF-8 -c -s
# exit normally
exit 0