This contribution does not follow the conventions set by the Google Java style guide. Please run the following command line at the root of the project to fix formatting errors: ./gradlew goJF.

✅ Rule acceptance tests passed.
New Errors: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
Dropped Errors: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
New Warnings: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
Dropped Warnings: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
0 out of 1426 sources (~0 %) are corrupted.
Commit: 7f10fd5
Download the full acceptance test report here (report will disappear after 90 days).
✅ Rule acceptance tests passed.

One question that is probably out of scope for this PR:
During the contributor meeting, we talked about merging the two concepts of "files affected" and "references" into one category.
In this PR it looks like we are going to keep the two categories via @FileRefs and @UrlRef, correct?

Also, we might want to have @FileRef and @UrlRef or @FileRefs and @UrlRefs for consistency?

@isabelle-dr a couple points:

• @FileRefs is used to capture all GTFS file links, whether they were in "files affected" or "references".
• @UrlRef is used to capture links to arbitrary webpages not already covered by the existing spec. You can see it used in this PR for a link to the "Original Python validator implementation". We don't want someone using it to link to a section of the GTFS spec or a GTFS file reference (we could probably enforce that with a unit-test?)

Does that make sense?

There is a bit of an edge-case where for the "https://gtfs.org/best-practices/#dataset-publishing--general-practices" link. I suppose I could update that to be a section ref?

As for @FileRefs vs @UrlRef, the difference is intentional.

Namely, @FileRefs can contain an array of values, such that you can specify:

files = @FileRefs({GtfsStopTimeSchema.class, GtfsStopSchema.class})

I didn't want to go with @FileRef here because it seemed more verbose to have to do something like:

files = {@FileRef(GtfsStopTimeSchema.class), @FileRef(GtfsStopSchema.class)}

Compare that to @UrlRef, which has two distinct arguments:

@UrlRef(
            label = "Original Python validator implementation",
            url = "https://github.com/google/transitfeed")

It doesn't make as much sense to do @UrlRefs because it's not possible to construct multiple different URLs in the specification. Thus, I did go with individual @UrlRef instead.

Looking back, I think I had imagined adding more fields to @FileRefs but instead, that info ended up in @SectionRefs instead. We could probably refactor the code to look something like:

files = {GtfsStopTimeSchema.class, GtfsStopSchema.class}

getting rid of @FileRefs entirely.

✅ Rule acceptance tests passed.
New Errors: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
Dropped Errors: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
New Warnings: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
Dropped Warnings: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
0 out of 1426 sources (~0 %) are corrupted.
Commit: fface89
Download the full acceptance test report here (report will disappear after 90 days).
✅ Rule acceptance tests passed.

✅ Rule acceptance tests passed.
New Errors: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
Dropped Errors: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
New Warnings: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
Dropped Warnings: 0 out of 1426 datasets (~0%) are invalid due to code change, which is less than the provided threshold of 1%.
0 out of 1426 sources (~0 %) are corrupted.
Commit: f27a179
Download the full acceptance test report here (report will disappear after 90 days).
✅ Rule acceptance tests passed.