You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@nutch.apache.org by "Mubey N." <mu...@gmail.com> on 2007/10/30 10:25:12 UTC
parse-pdf output is not pretty in cached.jsp
We are using parse-pdf to parse PDF documents. We modified cached.jsp
to display the parsed content instead of the link to the cached
document. We are using bean.getParseText(details) to get the parsed
text from the cached PDF document. But the output that comes on
cached.jsp is not pretty. Parsed text doesn't have any formatting
information. I am just wondering whether there is anything in Nutch
that could display cached PDF documents with proper formatting or at
least some formatting like headers, paragraphs, etc.?
Re: parse-pdf output is not pretty in cached.jsp
Posted by Andrzej Bialecki <ab...@getopt.org>.
Mubey N. wrote:
> We are using parse-pdf to parse PDF documents. We modified cached.jsp
> to display the parsed content instead of the link to the cached
> document. We are using bean.getParseText(details) to get the parsed
> text from the cached PDF document. But the output that comes on
> cached.jsp is not pretty. Parsed text doesn't have any formatting
> information. I am just wondering whether there is anything in Nutch
> that could display cached PDF documents with proper formatting or at
> least some formatting like headers, paragraphs, etc.?
It's possible to do this, but so far the demand has been low enough that
nobody implemented it in the public version ... The strategy is as follows:
* apply patches in https://issues.apache.org/jira/browse/NUTCH-466,
which provide support for multiple parts in a segment.
* implement a PDF parser based on pdftohtml.exe (using parse-ext
plugin?), which gives quite decent quality HTML output. Store this
output in an "html" segment part.
* modify the Cached.java to retrieve HTML content from the "html" part
of the segment, in case of PDF document.
--
Best regards,
Andrzej Bialecki <><
___. ___ ___ ___ _ _ __________________________________
[__ || __|__/|__||\/| Information Retrieval, Semantic Web
___|||__|| \| || | Embedded Unix, System Integration
http://www.sigram.com Contact: info at sigram dot com