not indexing docx documents

[EKC]

hi,
i'm using seeddms 5.1.19 on a centos 7.6 (php 7.1 / mariadb 10.2)

the docx document are not indexed.

i try with the docx2txt.pl but it doen't solve the problem.

extract of my settings.xml

    <converters target="fulltext">
        <converter mimeType="application/pdf">pdftotext -nopgbrk %s - | sed -e 's/ [a-zA-Z0-9.]\{1\} / /g' -e 's/[0-9.]//g'</converter>
        <converter mimeType="application/msword">catdoc %s</converter>
        <converter mimeType="application/vnd.openxmlformats-officedocument.wordprocessingml.document">perl /appli/seeddms/www/seeddms51x/pear/docx2txt.pl %s</converter>
        <converter mimeType="application/vnd.ms-excel">ssconvert -T Gnumeric_stf:stf_csv -S %s fd://1</converter>
        <converter mimeType="audio/mp3">id3 -l -R %s | egrep '(Title|Artist|Album)' | sed 's/^[^:]*: //g'</converter>
        <converter mimeType="audio/mpeg">id3 -l -R %s | egrep '(Title|Artist|Album)' | sed 's/^[^:]*: //g'</converter>
        <converter mimeType="text/plain">cat %s</converter>
        <converter mimeType="text/rtf">cat %s</converter>
    </converters>

i also have a pb with the accented characters (éèàê etc ... ) in the rtf documents

edit : i enable "Override MimeType:" in the settings

best regards
Emmanuel

[Uwe Steinmann]

Does your perl script output the converted document document to stdout? Do the other converters work?

[EKC]

it's ok for .pdf (and .rtf without accented characters)
how can i try the perl script in command line ?

[Michael]

I have the same issue. PDF and Text work, but no docx documents. I use 6.0.13
I testet it on command line. Doc files work while docx seem not to be supported by catdoc. What is the alternative?

[Michael]

I fixed it. In my case I needed to add a converter with mimeType="application/vnd.openxmlformats-officedocument.wordprocessingml.document"
and use docx2txt

this is the line I added:
<converter mimetype="application/vnd.openxmlformats-officedocument.wordprocessingml.document">docx2txt %s -</converter>