[
https://issues.apache.org/jira/browse/LUCENE-1730?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12726596#action_12726596
]
Mark Miller commented on LUCENE-1730:
-------------------------------------
The trec data you are using is ISO-8859-1. Wouldn't it be conceivable that they
might change the encoding to UTF-8 at some point? Or that someone else has
created trec compatible data in another encoding? Or trec has data in different
encodings? If something reads a source of files, and the files could
technically be in any encoding, I would expect things to be configurable so
that I can specify what encoding the files are in. It just seems like a good
standard feature with something that reads what are essentially pluggable
files. The format is going to be consistent, but why would the encoding
necessarily be consistent?
> TrecContentSource should use a fixed encoding, rather than system dependent
> ---------------------------------------------------------------------------
>
> Key: LUCENE-1730
> URL: https://issues.apache.org/jira/browse/LUCENE-1730
> Project: Lucene - Java
> Issue Type: Bug
> Components: contrib/benchmark
> Reporter: Shai Erera
> Fix For: 2.9
>
> Attachments: LUCENE-1730.patch
>
>
> TrecContentSource opens InputStreamReader w/o a fixed encoding. On Windows,
> this means CP1252 (at least on my machine) which is ok. However, when I
> opened it on a Linux machine w/ a default of UTF-8, it failed to read the
> files. The patch changes it to use ISO-8859-1, which seems to be the right
> one (and http://mg4j.dsi.unimi.it/man/manual/ch01s04.html mentions this
> encoding in its example of a script which reads the data).
> Patch to follow shortly.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
