Fix broken uri-use page links in updated reference docs of develop branch

[aj-stein-nist]

Committer Notes

This PR should fix failures on various PRs pointed to develop branch that are failing after the merge of #1263. It should fix website link checking on the following PRs:

• Bump peter-evans/create-issue-from-file from 4.0.0 to 4.0.1 #1514
• Bump peter-evans/create-or-update-project-card from 2.0.0 to 2.0.1 #1515
• Bump actions/setup-go from 3.3.0 to 3.3.1 #1516
• Bump actions/setup-java from 3.5.1 to 3.6.0 #1517

All Submissions:

• Have you selected the correct base branch per Contributing guidance?
• Have you set "Allow edits and access to secrets by maintainers
  "?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?
• Have you squashed any non-relevant commits and commit messages? [instructions]
• Do all automated CI/CD checks pass?

By submitting a pull request, you are agreeing to provide this contribution under the CC0 1.0 Universal public domain dedication.

Changes to Core Features:

• Have you added an explanation of what your changes do and why you'd like us to include them?
• Have you written new tests for your core changes, as applicable?
• Have you included examples of how to use your new feature(s)?
• Have you updated all OSCAL website and readme documentation affected by the changes you made? Changes to the OSCAL website can be made in the docs/content directory of your branch.

[aj-stein-nist]

I talked with Dave about this yesterday afternoon and it really seems we do not want to ignore non-existent pages, but revive the attempt to use the remapping feature and circle back with the upstream developer re lycheeverse/lychee#683 (similar use case, just different example as it pertains to our CI/CID checking). Will work towards quickly fixing that. Also removing relative paths that will confuse this remapping (the erroneous file:/// URI links that failed are because of a few relative paths that slipped into the source Metaschema definitions that are being checked).

[aj-stein-nist]

@david-waltermire-nist I am hopefully running this one for the last time. Even with updated configuration to look at local files (and avoid this inception problem with staged new files that are check-in but not "published" in nist-pages branch and triggered to get sent to pages.nist.gov) the path remapping is not so smart so saying "remap https://pages.nist.gov/OSCAL to /path/to/github/workspace/docs/public" (with updated site build) errors with the URL path not having OSCAL in it:

file:///home/runner/work/OSCAL/OSCAL/docs/public/OSCAL/concepts/uri-use: Failed: Cannot find file (status code: ERR)

From this last run before force-pushed commit on Metaschema @href changes: observe that /OSCAL/ is in the middle of paths in the local filesystem but only matter on the deployed server, despite the remap. I think this it the best we can do for now. I had to switch back to relative paths, but the kicker is they must be super-relative and I checked main Metaschema defs: it must be href="/concepts/..." and not href=/OSCAL/concepts/...", that breaks it like how we will continue getting this error ☝️. Oh well!

[aj-stein-nist]

@david-waltermire-nist this is finally ready for review. It did not occur to me what you were suggesting earlier today, but now think I understand. To avoid monkeying around further to override branch and remote protections. I use the local docker container to not only change Metaschema url definitions to relative, and update the config to have lychee scan locally mapped file:///path/to/website urls. What I hadn't done was force updates of the model reference content accordingly for the develop branch. I apologize I had not figured this out sooner.

This is good to go. I know it is unfortunately I must manually add in the model reference docs (from running /oscal/build/ci-cd/generate-model-documentation.sh -b develop -v in the container), but this will resolve the issues with the remaining 36 files with the source files changed and the rendered target docs which were missing.

Ironically, in a way this relates to #1447 in an edge case we usually don't need, but I will keep that in mind for the future.

[david-waltermire]

@aj-stein-nist I'd really like to find a way to not change the absolute links in the metaschema from https://pages.nist.gov/OSCAL/ to /. This make the links unusable for someone browsing a metaschema. They have no way of knowing where they point.

[aj-stein-nist]

@aj-stein-nist I'd really like to find a way to not change the absolute links in the metaschema from https://pages.nist.gov/OSCAL/ to /. This make the links unusable for someone browsing a metaschema. They have no way of knowing where they point.

Done and done. It appears that the issues I thought I had doing that before were the other issues you had pointed out to me while I was out on leave.

Also, @david-waltermire-nist, as discussed in the other PR, all is working except for the other issue that I fixed with bad, old Markdown links (to keep it organized and logically separate) in #1534. If you want to point there here instead of main, let me know and I will adjust accordingly.