Elasticsearch can index media files by using external tools/parsers to extract the text content of the files.

The parsers can be command line tools or web services. They must return plain text or a JSON structure.

Define parsers for each file extension that should be processed. Use ''%in%'' to specify the input file for CLI tools. Web services must accept the input file as POST data.

Here's a short example:

<code>
pdf    /usr/bin/pdftotext %in% -
docx   http://givemetext.okfnlabs.org/tika/rmeta
</code>