"DSSE" bundle tests are really intoto bundle tests

[woodruffw]

Noticed this during #152: the *dsse* labeled tests are really tests of intoto entries (that use DSSE for the message part, but not the log entry part). As part of improving testcase categorization, we should probably adopt a scheme like this:

testcase_tlogkind_messagekind.bundleversion.good|bad.sigstore.json

so, for example, d.txt.good.sigstore.json would become:

d_intoto_dsse.v0.2.good.sigstore.json

(Where d should also really be a better name, but baby steps 🙂)

[haydentherapper]

@segiddins also pointed this out to me when reviewing the Rekor v2 doc. We should add some tests for dsse entries as well.

[loosebazooka]

@segiddins also pointed this out to me when reviewing the Rekor v2 doc. We should add some tests for dsse entries as well.

Adding some now

[haydentherapper]

Good to close?

[jku]

Good to close?

not really -- it's really misleading that the happy path dsse tests are dsse but the interesting (negative) tests are actually intoto: at least sigstore-python just fails to handle the kindversion so seemingly passes the tests... but has not actually been tested at all.

• should do some renaming of tests at a minimum (I'll do something in refactor bundle verification tests #217 )
• preferably we would have the interesting tests implemented with the commonly supported kindversions first
• similar tests should be constructed for rekor v2 types