From c98bdb0edd5ebf5b222e1df1470a038a68f288f6 Mon Sep 17 00:00:00 2001 From: Jean-Francois Dockes Date: Fri, 8 Apr 2016 10:24:52 +0200 Subject: [PATCH] converted rcldjvu to python --- src/Makefile.am | 2 +- src/filters/rcldjvu | 180 ---------------------------------------- src/filters/rcldjvu.py | 107 ++++++++++++++++++++++++ src/sampleconf/mimeconf | 2 +- 4 files changed, 109 insertions(+), 182 deletions(-) delete mode 100755 src/filters/rcldjvu create mode 100755 src/filters/rcldjvu.py diff --git a/src/Makefile.am b/src/Makefile.am index 6b0f25b7..72133dd7 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -566,7 +566,7 @@ filters/rclcheckneedretry.sh \ filters/rclchm \ filters/rclconfig.py \ filters/rcldia \ -filters/rcldjvu \ +filters/rcldjvu.py \ filters/rcldoc.py \ filters/rcldvi \ filters/rclepub \ diff --git a/src/filters/rcldjvu b/src/filters/rcldjvu deleted file mode 100755 index 93210a71..00000000 --- a/src/filters/rcldjvu +++ /dev/null @@ -1,180 +0,0 @@ -#!/bin/sh -# @(#$Id: rcldjvu,v 1.6 2008-10-08 08:27:34 dockes Exp $ (C) 2005 J.F.Dockes - -# This program is free software; you can redistribute it and/or modify -# it under the terms of the GNU General Public License as published by -# the Free Software Foundation; either version 2 of the License, or -# (at your option) any later version. -# -# This program is distributed in the hope that it will be useful, -# but WITHOUT ANY WARRANTY; without even the implied warranty of -# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -# GNU General Public License for more details. -# -# You should have received a copy of the GNU General Public License -# along with this program; if not, write to the -# Free Software Foundation, Inc., -# 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. - -#================================================================ -# Extract text from a djvu file by executing djvused and djvutxt -# -# We use djvused to extract a possible title, djvutxt for the text -# -# Of course this only means anything if the djvu document actually has -# a text layer ! -# -# djvu utilities (04-2010) have a bug in which they try to interpret -# and convert file paths as character data, and fail miserably if the -# locale is not consistent with the actual encoding of the path (which -# could be arbitrary binary for all they know). We use a temporary -# symbolic link to get around this. -# -#================================================================ - -progname="rcldjvu" -filetype=dejavu - - -#RECFILTCOMMONCODE -############################################################################## -# !! Leave the previous line unmodified!! Code imported from the -# recfiltcommon file - -# Utility code common to all shell filters. This could be sourced at run -# time, but it's slightly more efficient to include the code in the -# filters at build time (with a sed script). - -# Describe error in a way that can be interpreted by our caller -senderror() -{ - echo RECFILTERROR $* - # Also alert on stderr just in case - echo ":2:$progname::: $*" 1>&2 - exit 1 -} - -iscmd() -{ - cmd=$1 - case $cmd in - */*) - if test -x $cmd -a ! -d $cmd ; then return 0; else return 1; fi ;; - *) - oldifs=$IFS; IFS=":"; set -- $PATH; IFS=$oldifs - for d in $*;do test -x $d/$cmd -a ! -d $d/$cmd && return 0;done - return 1 ;; - esac -} - -checkcmds() -{ - for cmd in $*;do - if iscmd $cmd - then - a=1 - else - senderror HELPERNOTFOUND $cmd - fi - done -} - -# show help message -if test $# -ne 1 -o "$1" = "--help" -then - echo "Convert a $filetype file to HTML text for Recoll indexing." - echo "Usage: $progname [infile]" - exit 1 -fi - -infile="$1" - -# check the input file existence (may be '-' for stdin) -if test "X$infile" != X- -a ! -f "$infile" -then - senderror INPUTNOSUCHFILE "$infile" -fi - -# protect access to our temp files and directories -umask 77 - -############################################################################## -# !! Leave the following line unmodified ! -#ENDRECFILTCOMMONCODE - -checkcmds djvutxt djvused awk - -# We need a temporary symlink to avoid path encoding issues -if test z"$RECOLL_TMPDIR" != z; then - ttdir=$RECOLL_TMPDIR -elif test z"$TMPDIR" != z ; then - ttdir=$TMPDIR -else - ttdir=/tmp -fi -tmplink=$ttdir/rcldjvu_tmp$$.djvu -rm -f $tmplink -ln -s "$infile" $tmplink || exit 1 - -cleanup() -{ - rm -f $tmplink -} - -trap cleanup EXIT HUP QUIT INT TERM - -# Title: we try to extract it from the annotations. djvused outputs string -# in C/awk \-escaped notation. Awk can only process this in string -# constants, so we have a first awk pass to create an awk program to parse -# the string as a constant (...). This is not exactly robust or nice -title=`djvused "$tmplink" -e 'select 1;output-ant' | \ -grep ' (title ' | sed -e 's/^.* (title //' -e 's/)$//' |\ -awk ' -{ - printf("BEGIN" " {s = %s; print s}\n", $0) -}' | awk -f -` - - -cat < - - $title - - - -
-EOF
-
-# The strange 'BEGIN' setup is to prevent 'file' from thinking this file
-# is an awk program
-djvutxt "$tmplink" | sed -e 's/[ 	][ 	]*$//' | \
-awk 'BEGIN'\
-' {
-  cont = ""
-}
-{
-    $0 = cont $0
-    cont = ""
-
-    if ($0 == "\f") {
-       print "
\n
\n
"
-       next
-    } else if ($0 ~ /[-]$/) {
-      # Break at last whitespace
-      match($0, "[ \t][^ \t]+$")
-      line = substr($0, 0, RSTART)
-      cont = substr($0, RSTART, RLENGTH)
-      $0 = line
-      gsub("-", "", cont)
-    }
-    gsub(/&/, "\\&", $0)
-    gsub(//, "\\>", $0)
-    print $0      
-}'
-
-cat <
-
-
-EOF
diff --git a/src/filters/rcldjvu.py b/src/filters/rcldjvu.py
new file mode 100755
index 00000000..46198a51
--- /dev/null
+++ b/src/filters/rcldjvu.py
@@ -0,0 +1,107 @@
+#!/usr/bin/env python
+# Copyright (C) 2016 J.F.Dockes
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the
+# Free Software Foundation, Inc.,
+# 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
+
+# Recoll DJVU extractor
+
+from __future__ import print_function
+
+import os
+import sys
+import re
+import rclexecm
+import subprocess
+
+class DJVUExtractor:
+    def __init__(self, em):
+        self.currentindex = 0
+        self.djvused = None
+        self.djvutxt = None
+        self.em = em
+
+    def extractone(self, params):
+        self.em.setmimetype('text/html')
+
+        # Extract metadata
+        if self.djvused:
+            try:
+                metadata = subprocess.check_output([self.djvused, self.filename,
+                                                    "-e", "select 1;print-meta"])
+            except Exception as e:
+                self.em.rclog("djvused failed: %s" % e)
+        author = ""
+        title = ""
+        metadata = metadata.decode('UTF-8', 'replace')
+        for line in metadata.split('\n'):
+            line = line.split('"')
+            if len(line) >= 2:
+                nm = line[0].strip()
+                if nm == "author":
+                    author = ' '.join(line[1:])
+                elif nm == "title":
+                    title = ' '.join(line[1:])
+
+        # Main text
+        try:
+            txtdata = subprocess.check_output([self.djvutxt, "--escape", self.filename])
+        except Exception as e:
+            self.em.rclog("djvused failed: %s" % e)
+            return (False, "", "", rclexecm.RclExecM.eofnow)
+        txtdata = txtdata.decode('UTF-8', 'replace')
+
+        data = '''''' + self.em.htmlescape(title) + ''''''
+        data += ''''''
+        if author:
+            data += ''''''
+        data += '''
'''
+
+        data += self.em.htmlescape(txtdata)
+        data += '''
'''
+        return (True, data, "", rclexecm.RclExecM.eofnext)
+
+    ###### File type handler api, used by rclexecm ---------->
+    def openfile(self, params):
+        self.filename = params["filename:"]
+        self.currentindex = 0
+        #self.em.rclog("openfile: [%s]" % self.filename)
+
+        if not self.djvutxt:
+            self.djvutxt = rclexecm.which("djvutxt")
+            if not self.djvutxt:
+                print("RECFILTERROR HELPERNOTFOUND djvutxt")
+                sys.exit(1);
+            self.djvused = rclexecm.which("djvused")
+
+        return True
+
+    def getipath(self, params):
+        return self.extractone(params)
+        return (ok, data, ipath, eof)
+        
+    def getnext(self, params):
+        if self.currentindex >= 1:
+            return (False, "", "", rclexecm.RclExecM.eofnow)
+        else:
+            ret= self.extractone(params)
+            self.currentindex += 1
+            return ret
+
+# Main program: create protocol handler and extractor and run them
+proto = rclexecm.RclExecM()
+extract = DJVUExtractor(proto)
+rclexecm.main(proto, extract)
diff --git a/src/sampleconf/mimeconf b/src/sampleconf/mimeconf
index 50b73ea0..7e3fbc04 100644
--- a/src/sampleconf/mimeconf
+++ b/src/sampleconf/mimeconf
@@ -128,7 +128,7 @@ image/jp2 = execm rclimg
 image/jpeg = execm rclimg
 image/png = execm rclimg
 image/tiff = execm rclimg
-image/vnd.djvu = exec rcldjvu
+image/vnd.djvu = execm rcldjvu.py
 image/svg+xml = execm rclsvg.py
 image/x-xcf = execm rclimg
 inode/symlink = internal