fix bzip2 magic detection

[alexanderadam]

Hi folks, 👋

this patch is actually a fix from @tomhughes.
See also the fix in OSM and Tom also opened issue #128 for that.

I added a minimal bzip file (echo -n '' | bzip2 > test/fixtures/magic/application/x-bzip2/bzip2.bz2) to the fixtures and the fix works just as expected.

PS: I'm looking for a new adventure in case anybody is looking for a Ruby/Rails/Crystal dev
PPS: would you be so kind and add the hacktoberfest-accepted label to this issue in case you find that PR helpful? 🥺

[tomhughes]

That's not really the right way to fix it though - what's really needed is regex support in the code and then translating type="regex" rules in tika.xml into regular expression matches.

What I did downstream is a hack to make this one specific case work by expanding the regex to a set of fixed strings.

[alexanderadam]

Oh, thank you for the super-quick response, @tomhughes 🙏 .

I'm sorry, I'll close this then.
But I'd assume that it should be better than before then at least, no? 🤔

[tomhughes]

I think those rules are just filling in gaps for things that the tika.xml file (which comes from https://github.com/apache/tika/blob/main/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml) doesn't have a rule for.

[alexanderadam]

I created #132 now. I think this might be the solution that you were looking for?