[
https://issues.apache.org/jira/browse/TIKA-2826?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16767412#comment-16767412
]
Tim Allison commented on TIKA-2826:
-----------------------------------
I'm attaching the output of tika-eval comparing output on .csv files before and
after adding the parser. When there aren't parse exceptions, the CSVParser
seems to do a much better job :D. However, there are now parse exceptions on
1-5% of .csv files. We may be able to improve this somewhat, but my initial
review shows that most (hand-waving...not formally measured) of the files that
caused exceptions were actually malformed csvs.
> Add a csv/tsv parser
> --------------------
>
> Key: TIKA-2826
> URL: https://issues.apache.org/jira/browse/TIKA-2826
> Project: Tika
> Issue Type: Task
> Reporter: Tim Allison
> Assignee: Tim Allison
> Priority: Minor
> Fix For: 2.0.0, 1.21
>
> Attachments: csvs_compared_reports.zip
>
>
> I'm not sure I want to wade into csv/tsv detection on this ticket, but it
> would be great to apply a csv parser at least for filenames ending in ".csv"
> or if a user "hints" that the file is a csv via the metadata.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
