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

added abiword + some oofice cleanup

Browse Source
This commit is contained in:
dockes 2007-06-15 11:41:50 +00:00
parent 2098c7afad
commit fcf027b22f
5 changed files with 47 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/filters/rclabw
View File

@ -1,5 +1,5 @@
#!/bin/sh
# @(#$Id: rclabw,v 1.1 2007-06-15 09:25:23 dockes Exp $ (C) 2004 J.F.Dockes
# @(#$Id: rclabw,v 1.2 2007-06-15 11:41:50 dockes Exp $ (C) 2004 J.F.Dockes
# Parts taken from Estraier:
#================================================================
# Estraier: a personal full-text search system
@ -97,7 +97,9 @@ encoding=`sed -e '/<?xml version=/s/"?>$//' \
if test X$encoding = X ; then encoding=UTF-8;fi
# Note: there can be newlines inside the description field, we don't want
# them...
# them... Have 2 use 2 different selectors for the single-line and
# multiple-line cases because of the generic tag end (</m> for all meta
# tags)
descsedprog='
/<m key="dc.description">\([^<]*\)<\/m>/ {
s//\1/
@ -149,7 +151,7 @@ creator=`sed -n -e "$metasedprog" "$infile"`
# a single line in the xml (no multiple <p> per line, no embedded newlines
# in text).
contentsedprog='
/<p /{
/<p[ >]/{
s/<[^>]*>/ /g
p
}

40
src/filters/rclsoff
View File

@ -1,5 +1,5 @@
#!/bin/sh
# @(#$Id: rclsoff,v 1.10 2007-06-08 13:51:09 dockes Exp $ (C) 2004 J.F.Dockes
# @(#$Id: rclsoff,v 1.11 2007-06-15 11:41:50 dockes Exp $ (C) 2004 J.F.Dockes
# Parts taken from Estraier:
#================================================================
# Estraier: a personal full-text search system
@ -127,26 +127,45 @@ cd $tmpdir/rclsofftmp
descsedprog='/<dc:description>/,/<\/dc:description>/{
s!.*<dc:description>!!
s!</dc:description>.*!!
H
${
g
s/\n/ /g
p
}
}
'
description=`sed -n -e "$descsedprog" < meta.xml | tr '\n' ' '`
description=`sed -n -e "$descsedprog" meta.xml`
#echo description "$description"
subject=`sed -e "s/\"/'/" -e 's/.*<dc:subject>\([^<]*\).*/\1/p;d' < meta.xml`
# Takes tag name as parameter and creates sed program to extract single
# line meta tags values.
setmetasedprog() {
metasedprog="s/\"/'/g"'
/.*<'"$1"'>\([^<]*\).*/s//\1/p
'
}
title=`sed -e "s/\"/'/" -e 's/.*<dc:title>\([^<]*\).*/\1/p;d' < meta.xml`
setmetasedprog dc:subject
subject=`sed -n -e "$metasedprog" meta.xml`
#echo subject: $subject
keywords=`sed -e "s/\"/'/" -e 's/.*<meta:keyword>\([^<]*\).*/\1/p;d' \
< meta.xml`
setmetasedprog dc:title
title=`sed -n -e "$metasedprog" meta.xml`
#echo title: $title
setmetasedprog meta:keyword
keywords=`sed -n -e "$metasedprog" meta.xml`
#echo keywords: $keywords
setmetasedprog dc:creator
creator=`sed -n -e "$metasedprog" meta.xml`
#echo creator: $creator
# Note: next expr inserts a newline at each end of paragraph (for preview)
content="`sed -e 's!</text:p>!\\
!g' -e 's/<[^>]*>/ /g' < content.xml`"
#echo description "$description"
#echo subject "$subject"
#echo title "$title"
#echo keywords "$keywords"
#echo content "$content"
# output the result
@ -155,6 +174,7 @@ echo '<title>' "$title" '</title>'
echo '<meta http-equiv="Content-Type" content="text/html;charset=UTF-8">'
echo '<meta name="description" content="' "$description $subject" '">'
echo '<meta name="keywords" content="' "$keywords" '">'
echo '<meta name="author" content="' "$creator" '">'
echo '</head><body><p>'
# The strange 'BEGIN' setup is to prevent 'file' from thinking this file

10
src/sampleconf/mimeconf
View File

@ -1,4 +1,4 @@
# @(#$Id: mimeconf,v 1.27 2007-06-08 13:51:09 dockes Exp $ (C) 2004 J.F.Dockes
# @(#$Id: mimeconf,v 1.28 2007-06-15 11:41:50 dockes Exp $ (C) 2004 J.F.Dockes
# Recoll : associations of mime types to processing filters.
# There are different sections for decompression, 'interning' for indexing
@ -41,6 +41,7 @@ application/vnd.sun.xml.math = exec rclsoff
application/vnd.sun.xml.writer = exec rclsoff
application/vnd.sun.xml.writer.global = exec rclsoff
application/vnd.sun.xml.writer.template = exec rclsoff
application/x-abiword = exec rclabw
application/x-dvi = exec rcldvi
application/x-lyx = exec rcllyx
application/x-kword = exec rclkwd
@ -73,6 +74,8 @@ application/vnd.sun.xml.impress.template = presentation
application/vnd.sun.xml.writer = wordprocessing
application/vnd.sun.xml.writer.global = wordprocessing
application/vnd.sun.xml.writer.template = wordprocessing
application/x-abiword = wordprocessing
application/x-kword = wordprocessing
application/x-fsdirectory = folder
application/x-dvi = document
audio/mpeg = sownd
@ -88,13 +91,16 @@ text/x-c = source
[categories]
texts = application/msword \
texts = \
application/msword \
application/pdf \
application/postscript \
application/vnd.sun.xml.writer \
application/vnd.sun.xml.writer.global \
application/vnd.sun.xml.writer.template \
application/x-abiword \
application/x-dvi \
application/x-kword \
application/x-lyx \
application/x-scribus \
image/vnd.djvu \

3
src/sampleconf/mimemap
View File

@ -1,4 +1,4 @@
# @(#$Id: mimemap,v 1.22 2007-06-08 13:51:09 dockes Exp $ (C) 2004 J.F.Dockes
# @(#$Id: mimemap,v 1.23 2007-06-15 11:41:50 dockes Exp $ (C) 2004 J.F.Dockes
# Recoll: associations of file name extensions to mime types
.txt = text/plain
@ -54,6 +54,7 @@
.sxg = application/vnd.sun.xml.writer.global
.stw = application/vnd.sun.xml.writer.template
.abw = application/x-abiword
.lyx = application/x-lyx
.sla = application/x-scribus
.scd = application/x-scribus

3
src/sampleconf/mimeview
View File

@ -1,4 +1,4 @@
# @(#$Id: mimeview,v 1.5 2007-06-08 13:59:55 dockes Exp $ (C) 2004 J.F.Dockes
# @(#$Id: mimeview,v 1.6 2007-06-15 11:41:50 dockes Exp $ (C) 2004 J.F.Dockes
## ##########################################
# External viewers, launched by the recoll GUI when you click on a result
@ -9,6 +9,7 @@
application/x-all = xdg-open %f
application/x-kword = kword %f
application/x-abiword = abiword %f
application/msword = openoffice %f
application/pdf = xpdf %f