Skip to content

WIP: Source Links Fix#440

Draft
MaximilianSoerenPollak wants to merge 27 commits intoeclipse-score:mainfrom
MaximilianSoerenPollak:MSP_fix_sourcelinks_2
Draft

WIP: Source Links Fix#440
MaximilianSoerenPollak wants to merge 27 commits intoeclipse-score:mainfrom
MaximilianSoerenPollak:MSP_fix_sourcelinks_2

Conversation

@MaximilianSoerenPollak
Copy link
Contributor

@MaximilianSoerenPollak MaximilianSoerenPollak commented Mar 11, 2026

Still in progress, do not merge.
It works but need to fix tests and clean it up a lot.

Still TODO:

  • Achieve same or better test coverage as currently (see last comment)
    • Adapt all current tests to work with new structure
    • Add new tests to enable testing of new functionality
  • Update Documentation if needed
  • Clean up implementation (naming / debug prints etc. )
  • Proof it works in ref-integration & locally (docs-as-code & other local repos)

📌 Description

🚨 Impact Analysis

  • This change does not violate any tool requirements and is covered by existing tool requirements
  • This change does not violate any design decisions
  • Otherwise I have created a ticket for new tool qualification

✅ Checklist

  • Added/updated documentation for new or changed features
  • Added/updated tests to cover the changes
  • Followed project coding standards and guidelines

@github-actions
Copy link

github-actions bot commented Mar 11, 2026

License Check Results

🚀 The license check job ran with the Bazel command:

bazel run --lockfile_mode=error //src:license-check

Status: ⚠️ Needs Review

Click to expand output
[License Check Output]
Extracting Bazel installation...
Starting local Bazel server (8.3.0) and connecting to it...
INFO: Invocation ID: a498a856-a238-4879-b4fc-4995b788a549
Computing main repo mapping: 
Computing main repo mapping: 
Loading: 
Loading: 0 packages loaded
Loading: 0 packages loaded
Loading: 0 packages loaded
    currently loading: src
Loading: 0 packages loaded
    currently loading: src
Analyzing: target //src:license-check (1 packages loaded, 0 targets configured)
Analyzing: target //src:license-check (1 packages loaded, 0 targets configured)

Analyzing: target //src:license-check (57 packages loaded, 9 targets configured)

Analyzing: target //src:license-check (68 packages loaded, 9 targets configured)

Analyzing: target //src:license-check (76 packages loaded, 9 targets configured)

Analyzing: target //src:license-check (128 packages loaded, 1672 targets configured)

Analyzing: target //src:license-check (129 packages loaded, 2146 targets configured)

Analyzing: target //src:license-check (140 packages loaded, 2564 targets configured)

Analyzing: target //src:license-check (140 packages loaded, 2576 targets configured)

Analyzing: target //src:license-check (140 packages loaded, 2576 targets configured)

Analyzing: target //src:license-check (143 packages loaded, 4465 targets configured)

Analyzing: target //src:license-check (144 packages loaded, 4589 targets configured)

INFO: Analyzed target //src:license-check (145 packages loaded, 4715 targets configured).
[12 / 16] [Sched] JavaToolchainCompileClasses external/rules_java+/toolchains/platformclasspath_classes
[13 / 16] JavaToolchainCompileClasses external/rules_java+/toolchains/platformclasspath_classes; 1s disk-cache, processwrapper-sandbox
[14 / 16] JavaToolchainCompileBootClasspath external/rules_java+/toolchains/platformclasspath.jar; 0s disk-cache, processwrapper-sandbox
INFO: Found 1 target...
Target //src:license.check.license_check up-to-date:
  bazel-bin/src/license.check.license_check
  bazel-bin/src/license.check.license_check.jar
INFO: Elapsed time: 23.343s, Critical Path: 2.84s
INFO: 16 processes: 12 internal, 3 processwrapper-sandbox, 1 worker.
INFO: Build completed successfully, 16 total actions
INFO: Running command line: bazel-bin/src/license.check.license_check src/formatted.txt <args omitted>
usage: org.eclipse.dash.licenses.cli.Main [-batch <int>] [-cd <url>]
       [-confidence <int>] [-ef <url>] [-excludeSources <sources>] [-help] [-lic
       <url>] [-project <shortname>] [-repo <url>] [-review] [-summary <file>]
       [-timeout <seconds>] [-token <token>]

@github-actions
Copy link

The created documentation from the pull request is available at: docu-html

Copy link

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR reworks how “source code links” are generated and injected into Sphinx needs, aiming to support both local builds (derive Git hash from the current repo) and combo/ref-integration builds (use per-module metadata such as repo URL + commit hash).

Changes:

  • Add metadata support (module name / repo URL / commit hash) to NeedLink JSON handling and Bazel CLI generation/merge scripts.
  • Introduce grouping of needs by module and a new module-grouped cache JSON.
  • Update Sphinx extension flow to read the new caches and generate GitHub links using either git-derived or metadata-derived refs.

Reviewed changes

Copilot reviewed 10 out of 10 changed files in this pull request and generated 15 comments.

Show a summary per file
File Description
src/helper_lib/additional_functions.py Changes GitHub link generation API to require module metadata.
src/extensions/score_source_code_linker/needlinks.py Extends NeedLink with metadata fields; adds new metadata-aware JSON format/load path.
src/extensions/score_source_code_linker/need_source_links.py Moves group_by_need into this module.
src/extensions/score_source_code_linker/module_source_links.py New module-grouped cache format and grouping logic.
src/extensions/score_source_code_linker/metadata.py New TypedDict + TypeGuard for metadata records.
src/extensions/score_source_code_linker/generate_source_code_links_json.py Refactors extraction helper signature and logging (currently inconsistent call sites).
src/extensions/score_source_code_linker/init.py Adds module-linker stage and changes injection to use module-grouped cache + metadata-based link generation.
scripts_bazel/merge_sourcelinks.py Merges per-module sourcelinks and enriches with known-good repo/hash metadata.
scripts_bazel/generate_sourcelinks_cli.py Emits sourcelinks JSON with a leading metadata dict and uses updated extraction helper signature.
docs.bzl Adds optional known_good wiring into the sourcelinks merge genrule and the public docs() macro.

You can also share your feedback on Copilot code review. Take the survey.

Comment on lines +106 to +108
store_source_code_links_with_metadata_json(
file=args.output, metadata=metadata, needlist=all_need_references
)
Copy link

Copilot AI Mar 12, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This switches the generated JSON format from a plain list of NeedLinks to a list whose first element is a metadata dict. Any existing consumers/tests that expect the old schema will now fail. Consider either keeping the old format as default (with an opt-in flag for metadata), or updating all in-repo consumers and tests in the same PR to avoid a partially-migrated state.

Copilot uses AI. Check for mistakes.
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@a-zw Valid issue here .

Do you think it would be better to rename this specific cache that comes from here a bit so it is clear it is with metadata?
Like 'scl_metadata_cache.json' or whatever? SO that the name makes it clear to use the metadata reader?

Comment on lines +30 to +39
metadata: moduleInfo,
link: NeedLink | DataForTestLink | DataOfTestCase | None = None,
) -> str:
if link is None:
link = DefaultNeedLink()
if not metadata.hash:
# Local path (//:docs)
return get_github_link_from_git(link)
# Ref-Integration path (//:docs_combo..)
return get_github_link_from_json(metadata, link)
Copy link

Copilot AI Mar 12, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

get_github_link now requires a metadata argument, but there are existing call sites in the repo that still call get_github_link(link) (e.g. in tests and xml_parser.py). This is a breaking API change that will raise TypeError. Consider keeping backwards compatibility by making metadata optional with a sensible default (e.g. infer local/git behavior when metadata is omitted) or providing a new function name for the metadata-based behavior.

Suggested change
metadata: moduleInfo,
link: NeedLink | DataForTestLink | DataOfTestCase | None = None,
) -> str:
if link is None:
link = DefaultNeedLink()
if not metadata.hash:
# Local path (//:docs)
return get_github_link_from_git(link)
# Ref-Integration path (//:docs_combo..)
return get_github_link_from_json(metadata, link)
metadata: moduleInfo
| NeedLink
| DataForTestLink
| DataOfTestCase
| None = None,
link: NeedLink | DataForTestLink | DataOfTestCase | None = None,
) -> str:
"""
Return a GitHub link for the given location.
Backwards compatible calling conventions:
- New style: get_github_link(metadata, link)
- Old style: get_github_link(link) # metadata omitted
"""
# Distinguish between new-style and old-style calls.
if isinstance(metadata, moduleInfo):
actual_metadata: moduleInfo | None = metadata
actual_link = link
else:
# Old-style: first argument is actually the link (or None)
actual_metadata = None
# If both metadata and link are provided but metadata is not a moduleInfo,
# prefer the explicitly provided link argument.
actual_link = link if link is not None else metadata
if actual_link is None:
actual_link = DefaultNeedLink()
# If no metadata is available or metadata.hash is falsy, fall back to git-based link.
if actual_metadata is None or not getattr(actual_metadata, "hash", None):
# Local path (//:docs)
return get_github_link_from_git(actual_link)
# Ref-Integration path (//:docs_combo..)
return get_github_link_from_json(actual_metadata, actual_link)

Copilot uses AI. Check for mistakes.
Copy link

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Copilot reviewed 9 out of 9 changed files in this pull request and generated 8 comments.


You can also share your feedback on Copilot code review. Take the survey.

@MaximilianSoerenPollak
Copy link
Contributor Author

Just realized that testlinks also need the module name earlier than source_code_linker extension end state.
As well as the edgecase that it could be that source_code_links are empty but testlinks exists. Then we have 0 idea about the module state.

I need to re-think this approach a bit and see what I can adapt to make this more convenient. I think I have an idea, but unsure.

@MaximilianSoerenPollak
Copy link
Contributor Author

Still a bit to do here, but the architecture is 'working', though not pretty.
Will work on fixing the tests soon, and fixing the finaly bug in the testlinks that the actuall github link is put together wrong

Copy link

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Copilot reviewed 18 out of 18 changed files in this pull request and generated 14 comments.


You can also share your feedback on Copilot code review. Take the survey.

# ─────────────────────[ Roundtrip Tests ]───────────────────


def test_roundtrip_standard_format(tmp_path: Path):
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Wait, How is this different than this function:

test_store_and_load_source_code_links

=> Double check, this might have been copy paste errors (copying from older version where top tests was not yet in)

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

They indeed were duplicates .Deleted test_store_and_load...



def test_parse_info_from_known_good_no_module_in_json(tmp_path: Path):
"""Test that assertion works when module not in top level keys"""
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
"""Test that assertion works when module not in top level keys"""
"""Test that assertion works when 'module' not in top level keys"""

@@ -21,7 +21,6 @@
# This whole directory implements the above mentioned tool requirements
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't like the namings in this file anymore.
As it was build with 'SourceCodeLiniks' as the final step before.
Now that isn't the case anymore so most likely can make this nicer / more viewable again.

logger = logging.getLogger(__name__)


def clean_external_prefix(path: Path) -> Path:
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Wonder if this can be used in the testlink cleaning too, or if they are too different.

Comment on lines +74 to +78
metadata: MetaData = {
"module_name": "",
"hash": "",
"url": "",
}
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think it makes sense to make a 'default metadata' function or something.

This setup seems to be called a lot

Comment on lines +81 to +83
if "known_good.json" not in str(file_path) and not metadata_set:
metadata["module_name"] = parse_module_name_from_path(file_path)
metadata_set = True
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
if "known_good.json" not in str(file_path) and not metadata_set:
metadata["module_name"] = parse_module_name_from_path(file_path)
metadata_set = True
if "known_good.json" not in str(file_path) and not metadata_set:
# Doing this so we only have to parse the module name once, not every file.
metadata["module_name"] = parse_module_name_from_path(file_path)
metadata_set = True

Comment on lines +30 to +35
"""
if bazel-out/k8-fastbuild/bin/external/ in file_path => module is external
otherwise it's local
if local => module_name & hash == empty
if external => parse thing for module_name => look up known_good json for hash & url
"""
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Remove debug strings here.

)

args = parser.parse_args()
all_files = [x for x in args.files if "known_good.json" not in str(x)]
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Add a comment:
#Known_good.json is getting passed in 'args.files' as well.
#Therefore we filter it out here in order to make following logic cleaner

# If the file is empty e.g. '[]' there is nothing to parse, we continue
if not data:
continue
metadata = data[0]
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we need a worldview assertion here?
e.g. check if data is a list?

Comment on lines +72 to +76
logger.warning(
f"Unexpected schema in sourcelinks file '{json_file}': "
"expected first element to be a metadata dict "
"with a 'module_name' key. "
)
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should this be an error instead?

for d in data[1:]:
d.update(metadata)
assert isinstance(data, list), repr(data)
merged.extend(data[1:])
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Add comment:
# Taking only the needlinks. As data[0] = metadata we start from 1


# COMBO BUILD

if str(path).startswith("external/"):
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Redudant?



@dataclass
class ModuleInfo:
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Module => Repo

Comment on lines +153 to +158
# # Pouplate Metadata
# # Since all metadata inside the Codelinks is the same
# # we can just arbitrarily grab the first one
# module_name=need_links.CodeLinks[0].module_name,
# hash=need_links.CodeLinks[0].hash,
# url=need_links.CodeLinks[0].url,
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Delete

Copy link
Contributor

@a-zw a-zw left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good enough to merge to me.

Documentation is still missing but we can address that in a separate PR.

@MaximilianSoerenPollak
Copy link
Contributor Author

Looks good enough to merge to me.

Documentation is still missing but we can address that in a separate PR.

Let me just make sure all the consumer tests & ref-integration pass the test now.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: Backlog

Development

Successfully merging this pull request may close these issues.

3 participants