[
https://issues.apache.org/jira/browse/JENA-699?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14132957#comment-14132957
]
Andy Seaborne commented on JENA-699:
------------------------------------
Thank you for the contribution. Looks good.
The CSV parser in Jena has an element of "technical debt". There needed to be
a parser to get through the W3C SPARQL-WG tests (W3C process : 2
implementations needed). All that was required was strict RFC 4180 though UTF-8
handling was on my list of good things to have. There are a lot of CSV parsers
around but looking at licenses, wanting just a parser, not object model for
handing CSV-ish data, and the correct handlign of "" escapes, it became easier
to write one, Apache Common CSV not being out of snapshot at the time. Given
Jena has a parsing infratstructure, cooking up a CSV parser is was the easiest
thing to do at the time but that does accumulate some "technical debt". Real
CSV is a whole lot messier than the RFC defines and a solid basis to CSV
parsing is better than the current one. The parser API should not be seen as
fixed; it did it's job at the time but it's not really a public contract. The
new jena-csv is better off with a solid parser.
> Replace the CSV/TSV parsing with Apache Commons CSV
> ---------------------------------------------------
>
> Key: JENA-699
> URL: https://issues.apache.org/jira/browse/JENA-699
> Project: Apache Jena
> Issue Type: Improvement
> Reporter: Andy Seaborne
> Priority: Minor
> Attachments: JENA-699.patch
>
>
> When Apache Commons CSV is released, use that and remove the current parsers
> in favour of a properly written and designed component.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
