PPT filter: use mso-dump

src/filters/ppt-dump.py

@@ -0,0 +1,141 @@
+#!/usr/bin/env python2
+########################################################################
+#
+#  Copyright (c) 2010 Kohei Yoshida, Thorsten Behrens
+#  
+#  Permission is hereby granted, free of charge, to any person
+#  obtaining a copy of this software and associated documentation
+#  files (the "Software"), to deal in the Software without
+#  restriction, including without limitation the rights to use,
+#  copy, modify, merge, publish, distribute, sublicense, and/or sell
+#  copies of the Software, and to permit persons to whom the
+#  Software is furnished to do so, subject to the following
+#  conditions:
+#  
+#  The above copyright notice and this permission notice shall be
+#  included in all copies or substantial portions of the Software.
+#  
+#  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+#  EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
+#  OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+#  NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
+#  HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
+#  WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+#  FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+#  OTHER DEALINGS IN THE SOFTWARE.
+#
+########################################################################
+
+import sys, os.path, getopt
+sys.path.append(sys.path[0]+"/msodump.zip/src")
+import ole, pptstream, globals, olestream
+
+from globals import error
+
+def usage (exname):
+    exname = os.path.basename(exname)
+    msg = """Usage: %s [options] [ppt file]
+
+Options:
+  --help        displays this help message.
+  --no-struct-output suppress normal disassembly output
+  --dump-text   print the textual content
+"""%exname
+    print msg
+
+
+class PPTDumper(object):
+
+    def __init__ (self, filepath, params):
+        self.filepath = filepath
+        self.params = params
+
+    def __printDirHeader (self, dirname, byteLen):
+        dirname = globals.encodeName(dirname)
+        globals.outputln("")
+        globals.outputln("="*68)
+        globals.outputln("%s (size: %d bytes)"%(dirname, byteLen))
+        globals.outputln("-"*68)
+
+    def dump (self):
+        file = open(self.filepath, 'rb')
+        strm = pptstream.PPTFile(file.read(), self.params)
+        file.close()
+        strm.printStreamInfo()
+        strm.printHeader()
+        strm.printDirectory()
+        dirnames = strm.getDirectoryNames()
+        result = True
+        for dirname in dirnames:
+            if len(dirname) == 0 or dirname == 'Root Entry':
+                continue
+
+            try:
+                dirstrm = strm.getDirectoryStreamByName(dirname)
+            except Exception, err:
+                error("getDirectoryStreamByName(%s): %s\n" % (dirname,str(err)))
+                # The previous version was killed by the exception
+                # here, so the equivalent is to break, but maybe there
+                # is no reason to do so.
+                break
+            self.__printDirHeader(dirname, len(dirstrm.bytes))
+            if  dirname == "PowerPoint Document":
+                if not self.__readSubStream(dirstrm):
+                    result = False
+            elif  dirname == "Current User":
+                if not self.__readSubStream(dirstrm):
+                    result = False
+            elif  dirname == "\x05DocumentSummaryInformation":
+                strm = olestream.PropertySetStream(dirstrm.bytes)
+                strm.read()
+            else:
+                globals.dumpBytes(dirstrm.bytes, 512)
+        return result
+
+    def __readSubStream (self, strm):
+        # read all records in substream
+        return strm.readRecords()
+
+
+def main (args):
+    exname, args = args[0], args[1:]
+    if len(args) < 1:
+        print("takes at least one argument")
+        usage(exname)
+        return
+
+    params = globals.Params()
+    try:
+        opts, args = getopt.getopt(args, "h",
+                                   ["help", "debug", "show-sector-chain",
+                                    "no-struct-output", "dump-text"])
+        for opt, arg in opts:
+            if opt in ['-h', '--help']:
+                usage(exname)
+                return
+            elif opt in ['--debug']:
+                params.debug = True
+            elif opt in ['--show-sector-chain']:
+                params.showSectorChain = True
+            elif opt in ['--no-struct-output']:
+                globals.muteOutput(1)
+                params.noStructOutput = True
+            elif opt in ['--dump-text']:
+                params.dumpText = True
+            else:
+                error("unknown option %s\n"%opt)
+                usage()
+
+    except getopt.GetoptError:
+        error("error parsing input options\n")
+        usage(exname)
+        return
+
+    dumper = PPTDumper(args[0], params)
+    if not dumper.dump():
+        error("FAILURE\n")
+    if params.dumpText:
+        print(globals.textdump.replace("\r", "\n"))
+
+if __name__ == '__main__':
+    main(sys.argv)

src/filters/rclppt

@@ -17,11 +17,9 @@
 
 #================================================================
 # Handle powerpoint files for recoll. 
-# Uses catppt from the catdoc utilities
+# Use unoconv, this is very slow, but catppt just can't handle the majority
-# (http://ftp.45.free.net/~vitus/software/catdoc/)
+# of semi-modern ppt files
-# In my experience, this sometimes fail to extract text, printing "Default
+
-# Design" ou "format par defaut" instead and only.
-#
 #================================================================
 
 # set variables
@@ -96,72 +94,17 @@ umask 77
 # !! Leave the following line unmodified !
 #ENDRECFILTCOMMONCODE
 
-havecappt=no
+filtersdir=`dirname $0`
-iscmd cappt && havecappt=yes
+checkcmds $filtersdir/ppt-dump.py
-haveunoconv=no
-iscmd unoconv && haveunoconv=yes
-iscmd pdftotext || haveunoconv=no
 
-if test X$havecatppt = Xno -a X$haveunoconv = Xno ; then
+mso="$filtersdir/ppt-dump.py --no-struct-output --dump-text"
-    # checkcmds will exit with the appropriate salutations
-    checkcmds catppt unoconv pdftotext
-fi
 
-# This needs a temp dir because we first output pdf (outputting html
+cat <<EOF
-# would produce one file per page), and pdftotext can't read from
+<html><head>
-# stdin
+<meta http-equiv="Content-Type" content="text/html;charset=UTF-8">
-if test z"$RECOLL_TMPDIR" != z; then
+</head><body><pre>
-    ttdir=$RECOLL_TMPDIR
+EOF
-elif test z"$TMPDIR" != z ; then
-    ttdir=$TMPDIR
-else
-    ttdir=/tmp
-fi
 
-tmpdir=$ttdir/rclppt_tmp$$
+$mso "$infile"| sed -e 's/</&lt;/g' -e 's/&/&amp;/g' 
-mkdir $tmpdir || exit 1
-mkdir $tmpdir/rclppttmp || exit 1
 
-# We have to use a directory as output parameter to unoconv. Up to
+echo '</pre></body></html>'
-# version 0.5, it could not use a file name for this
-unopdf=$tmpdir/rclppttmp
-cattxt=$tmpdir/rclppttmp/output.txt
-cleanup()
-{
-    # Note that we're using a constant part (rclppttmp), which
-    # hopefully guarantees that we can't do big mistakes here.
-    rm -rf $tmpdir/rclppttmp
-    rmdir $tmpdir
-}
-trap cleanup EXIT HUP QUIT INT TERM
-
-# Try catppt. If the output looks too small and unoconv is available,
-# use this instead. unoconv is very slow but it handles newer files
-# that catppt will not convert.
-#
-# I'm not sure of the right test for detecting catppt failure. On the
-# sample I have, it outputs Azure\n1_Azure\n\n. I don't know if Azure
-# is a good marker of failure. Anyway, it seems unlikely that a real
-# ppt would have fewer than 5 lines
-
-catppt -d utf-8 "$infile" > $cattxt
-lines=`wc -l < $cattxt`
-
-if test $lines -lt 5 -a X$haveunoconv = Xyes; then
-    unoconv -f pdf -o "$unopdf" "$infile"
-    sinfile=`basename "$infile"`
-    `dirname $0`/rclpdf "$unopdf/${sinfile%.*}.pdf"
-else
-    # output the catppt result
-    echo '<html><head>'
-    #echo '<title>' "$title" '</title>'
-    echo '<meta http-equiv="Content-Type" content="text/html;charset=UTF-8">'
-    echo '</head><body>'
-    echo '<pre>'
-
-    catppt -d utf-8 "$infile" | \
-        sed -e 's/</&lt;/g' -e 's/&/&amp;/g' < $cattxt
-
-    echo '</pre>'
-    echo '</body></html>'
-fi