Darwin (macOS) port for the udffsck tool

[gmerlino]

Hi,

I'm trying to get a workable setup for repairing UDF filesystems under macOS (darwin) as well as Linux and, as discussed in pali#13, possibly merged upstream as well.

With this PR I'm able to get at least the udffsck (not other udftools at the moment) compiled (and launched) without errors. This does not imply I've tested the tool, but as you can verify from the diffs, it is basically standard fixes to the autotools stuff, plus adding some basic header files (taken from other projects verbatim) which are required under macOS but otherwise missing.

To complete the picture, if you (or anyone else) would like to easily test that the PR works as expected, you can use my Homebrew Formula for it:

https://github.com/gmerlino/homebrew-udffsck

At the moment it points to my fork, but I would like to update it accordingly to your repo, if the PR gets merged here, and later on down the road to Pali's repo, once he accepts your udffsck upstream (not to mention, possibly pushing the formula upstream to Homebrew).

Regards,
Giovanni

[argorain]

Hi,

thanks for your PR.

I can see your changes are more or less standard fixes, but in fact those changes broke building of udftools project, as you can see here: https://travis-ci.org/argorain/udftools/builds/325280450?utm_source=github_status&utm_medium=notification

udftools is by default for linux and it should compile under linux without any aditional parameter. Therefore it is manadatory to isolate all macOS stuff under some switch or it should compile with default parameters on both platforms.

It would be great if I can run some tests for macOS. Do you know about some service like Travis CI for it?

Thanks,
Vojtech

[gmerlino]

I noticed the failed builds actually, but at first look it's not clear to me the relationship between my modifications and the failures, I'm a bit of a Travis newbie.

I'll give it a deeper look, and try to fix of course.

With regard to macOS in Travis it seems supported actually, see:

https://docs.travis-ci.com/user/multi-os/

[gmerlino]

Fixed building, and thanks for triggering again the Travis tests that previously failed due to transients.

As you may have noticed already, I've also tackled Travis testing under macOS.

This is a quite different PR with respect to the non-fsck-enabled work in: pali#14

I've tried to keep things as separate (and as simple) as possible, because I think it is valuable to have the macOS port merged even before the fsck is fully upstreamed.
Nevertheless I believe this PR is absolutely not invasive, and has it merits whatever the timing (or fate) of the fsck upstreaming.

Please let me know your opinion.

Thanks again for your help.

[argorain]

Why is testing, x86/x86_64 and gcc disabled for osx? All those tests shall be run.

Keeping osx as allowed failure is good idea, since it is primarily linux package.

[argorain]

With respect to pali#14, this needs to be sorted out in same way to prevent future conflict.