[
https://issues.apache.org/jira/browse/TIKA-4029?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Tim Allison updated TIKA-4029:
------------------------------
Description:
In the CSVParser, we had a catch statement in case the csv file was malformed.
In the catch statement we read the rest of the stream and wrote it as one big
paragraph. We then threw the exception.
In commons-csv 1.10.0, it looks like they changed the IllegalStateException to
an UncheckedIOException. So, we're now missing a bunch of text that we used to
extract from csv files.
This issue was discovered during our regression tests in preparation for the
release of 2.8.0-rc1
was:
In the CSVParser, we had a catch statement in case the csv file was malformed.
In the catch statement we read the rest of the stream and wrote it as one big
paragraph. We then threw the exception.
In commons-csv 1.10.0, it looks like they changed the IllegalArgumentException
to an UncheckedIOException. So, we're now missing a bunch of text that we used
to extract from csv files.
This issue was discovered during our regression tests in preparation for the
release of 2.8.0-rc1
> commons-csv 1.10 changed IllegalStateException to IOException
> -------------------------------------------------------------
>
> Key: TIKA-4029
> URL: https://issues.apache.org/jira/browse/TIKA-4029
> Project: Tika
> Issue Type: Task
> Reporter: Tim Allison
> Priority: Minor
>
> In the CSVParser, we had a catch statement in case the csv file was
> malformed. In the catch statement we read the rest of the stream and wrote
> it as one big paragraph. We then threw the exception.
>
> In commons-csv 1.10.0, it looks like they changed the IllegalStateException
> to an UncheckedIOException. So, we're now missing a bunch of text that we
> used to extract from csv files.
>
> This issue was discovered during our regression tests in preparation for the
> release of 2.8.0-rc1
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)