SI-4520 - reasonable exception instead of OOE in XML parser

[terma]

That's fix OutOfMemoryException for case when we parse valid XML, however user code request not existent content or not correct content type. Now parser instead of consume all memory just raise exception that request info not present.

Jira: https://issues.scala-lang.org/browse/SI-4339 and https://issues.scala-lang.org/browse/SI-4520

[terma]

Can't find reviewer exactly for XML lib in http://www.scala-lang.org/contribute/hacker-guide.html
So please review @phaller
Thx

[adriaanm]

Hi, @terma! I'm very sorry about the long wait for a review here. Might I suggest pinging one of the other recent submitters of PRs to review your fix? I'm not at all familiar with scala-xml, unfortunately. To facilitate the review, please expand your commit message with the reasoning behind the fix.

Also, SI-4520 is a duplicate of https://issues.scala-lang.org/browse/SI-4339.

I'm hoping to cut a 1.0.3 release soon and would like to include your fix.

[adriaanm]

(If not, there's always 1.0.4 :-))

[terma]

@adriaanm thx for feedback. Not clear for me what reason I need to provide as that's simple bug not feature so I just fixed it. Any way I added more info to pr description

@som-snytt please review

[adriaanm]

Thank you! Could you please rebase on current master? While you're at it, could you include the PR description in the commit message -- PR descriptions are not preserved in git history.

[terma]

OK Now I got your point will do
On Dec 3, 2014 10:06 PM, "Adriaan Moors" [email protected] wrote:

Thank you! Could you please rebase on current master? While you're at it,
could you include the PR description in the commit message -- PR
descriptions are not preserved in git history.

—
Reply to this email directly or view it on GitHub
#32 (comment).

[som-snytt]

The Mark Harrah patch addresses the change at

scala/scala@0536364#diff-8561ae5555eec4758d457614adc309bc

by putting the exception in xToken.

Not every syntax error is fatal. For instance, pubidLiteral complains about bad chars but it can still consume the quoted string and continue parsing.

So the patch is a little better: xToken says it better be what I ask for.

This fix doesn't address the loop in xTakeUntil that doesn't check eof. That seems a direct response.

[SethTisue]

anyone interested in moving forward on this...?

[ashawley]

Re-opened as #209 to see if these tests could be fixed, including some additional tests.