[
https://issues.apache.org/jira/browse/MIME4J-60?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12659103#action_12659103
]
Oleg Kalnichevski commented on MIME4J-60:
-----------------------------------------
Stefano,
Thank you so much for making me measuring the impart of these changes on
performance of the parser. To my embarrassment, the patch causes a major
performance degradation (almost 3 fold)
Parsing a mulripart message 53'920 bytes (50'000 repetitions)
---
trunk: 23.364 ms
patched: 79.333 ms
patched (MimeMode is an int): 62.453 ms
---
Folks,
I think it is just not worth it. Performance degradation is huge, benefits are
questionable. I suggest we just make all classes behave consistently with
regards to line delimiter handling and be done with it.
I am going to commit the code I was using to benchmark performance, if I hear
no objections, so that others could re-run the tests locally.
Oleg
> Configurable strategy for line delimiters
> -----------------------------------------
>
> Key: MIME4J-60
> URL: https://issues.apache.org/jira/browse/MIME4J-60
> Project: JAMES Mime4j
> Issue Type: Wish
> Affects Versions: 0.4
> Reporter: Stefano Bagnara
> Fix For: 0.6
>
> Attachments: mime4j-60-phase1.patch,
> MIME4J-60-readLine-returns-no-newline.patch, newlines-tests.zip
>
>
> There is an ongoing discussion about how we should deal with non canonical
> line endings (isolated LF and/or isolated CR).
> This issue is to track discussion results and proposed patches.
--
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]
