[mohgupta-ship-it]

Summary

• Convert editable-install file URLs with the standard platform path adapter.
• Preserve the empty-authority delimiter so valid POSIX double-slash paths remain local on Python 3.14 while Windows drive URLs lose their extra leading slash.
• Add portable regression coverage for both URL shapes.

This builds on #486 and preserves the original author attribution while closing the compatibility gap found during validation.

Fixes #485

Validation

• make lint
• make format-check
• ruff check scripts/compare_scan_accuracy.py
• Targeted path-conversion tests: 3 passed
• Full Python 3.12 suite: 3,980 passed, 14 skipped, 4 xfailed
• Python 3.14 standard-library conversion probe for empty-authority double-slash paths

Submitted by Codex on behalf of Mohit Gupta.

[rng1995]

[SkillSpector Review]

Changes requested at head ea42d8e0e0ca2fabf049345b3856fa166a3792d5.

• scripts/compare_scan_accuracy.py:187: for an empty-authority URL whose parsed path begins with exactly //, the new branch prepends two more slashes. On supported pre-3.14 POSIX Python, url2pathname leaves those four slashes unchanged and Path.resolve() collapses them to a single root slash, losing the implementation-defined exactly-two-leading-slash path that was parsed. The test mocks url2pathname and asserts only its input, so it cannot catch the resulting path corruption. Use a version-independent conversion that preserves the parsed local path semantics on both POSIX and Windows, and test the actual converted/resolved result across supported Python versions rather than only a mocked argument.

No checks are reported and GitHub reports merge state BLOCKED; the code issue and missing required CI both block merge. This PR appears to supersede #486, so avoid landing both versions of the same fix.

[rng1995]

[SkillSpector Review]

Re-reviewed current head e3b5b16fa8b3bcc876770e911b42e39ee2f642da. I am withdrawing the previous path-corruption blocker: it relied on an incorrect assumption about urllib.request.url2pathname. CPython 3.12 and 3.13 deliberately remove the two empty-authority delimiter slashes from a ////path input, preserving the parsed //path; CPython 3.14 parses the same input as a local file: URL with empty authority and also preserves //path. The /C:/... form is likewise handed to the platform adapter so Windows can remove the drive-letter prefix slash.

The checked-in test pins both converter inputs and would fail if the standard-library adapter were bypassed, while the existing runtime probe continues to exercise editable dependency traversal and hashing. I found no remaining required change in the current two-file diff.

No hosted checks are reported and GitHub currently reports mergeStateStatus=BLOCKED, so this approval does not make the PR merge-ready.

[rng1995]

[SkillSpector Review]

Re-reviewed current head ce56485a4c4a243f36c4fa5f8f3ad54c25e7d585. The feature blobs remain identical to the reviewed implementation, and the previous empty-authority concern remains correctly withdrawn. A distinct supported-runtime regression remains for accepted file://localhost//… inputs on Python 3.14: the current delimiter repair covers only an empty authority, allowing the adapter to reinterpret the first path component as authority. Preserve the localhost double-leading path and add a real-adapter regression.

No required checks are attached, GitHub reports mergeStateStatus=BLOCKED, and this PR must also be coordinated with superseded #486.

[rng1995]

[P2] Preserve localhost double-leading paths

The accepted URL file://localhost//path reaches this branch with parsed.netloc == "localhost" and parsed.path == "//path", so the delimiter is not restored. On Python 3.14, url2pathname reparses that input as file://path (authority path); POSIX then raises URLError instead of resolving the local //path. The repository supports Python 3.14. Restore the delimiter for this accepted localhost form as well and add a regression that exercises the real adapter rather than only mocking its argument.

[yashrajp22]

I checked this on the latest commit and can still reproduce the issue on Python 3.14: file://localhost//tmp/dep raises URLError instead of resolving the local path. Could we preserve // for localhost URLs too and add a regression test using the real converter?

[yashrajp22]

Please address the remaining localhost path issue discussed in the inline thread.

[rng1995]

[SkillSpector Review]

Re-reviewed current head 94a45065ef3eea389a07429e0bb6dcf664260df4. The head change only synchronizes unrelated main work; neither PR-owned file changed, so the current non-outdated blocker remains unchanged. scripts/compare_scan_accuracy.py:186 repairs the delimiter only for an empty authority; accepted file://localhost//path still passes //path to Python 3.14 url2pathname, which reparses the first path component as authority and raises URLError. Preserve the localhost double-leading local path and add a real-adapter regression as requested in the existing thread.

No exact-head checks are attached, GitHub reports BLOCKED, the current thread remains unresolved, and this must be coordinated with superseded #486.

[rng1995]

[SkillSpector Review]

Value and readiness

The Windows platform-adapter correction remains useful, but the previously documented localhost conversion blocker is still present at this head. Changes remain required.

Material findings

Blocker still open — scripts/compare_scan_accuracy.py:186-188. Accepted authorities include localhost, but delimiter restoration remains gated by not parsed.netloc. Consequently file://localhost//path passes //path to Python 3.14 url2pathname; its file: prefix and reparse treat path as a nonlocal authority and POSIX raises URLError. Apply authority-safe delimiter preservation to both accepted local authorities, and add the localhost case through the real platform converter. tests/unit/test_compare_scan_accuracy.py:1120-1173 still covers only drive/empty-authority forms with a mocked adapter, so it cannot detect this case. The existing non-outdated thread remains unresolved and includes another reviewer's reproduction; no duplicate inline finding is needed.

Prior issue status: the original empty-authority path-corruption concern was explicitly withdrawn in the September 12 marked review and remains withdrawn. All subsequent marked localhost change requests remain applicable to the current condition and tests. The other human reviewer's change request remains unaddressed.

PIC tradeoffs

None identified. The earlier alternative #486 is already CLOSED and unmerged; no duplicate landing is needed.

Verification and gaps

Read the full two-file diff, exact-head script/tests and supported Python metadata, every marked review and all discussion/thread evidence. Re-read the official CPython 3.14 adapter source, including prefix/reparse and nonlocal-authority rejection; Python 3.14 execution was not performed. No exact-head checks are attached. Exact-head CI run 41513876346 has conclusion action_required, and GitHub reports MERGEABLE/BLOCKED. Workflow authorization is a separate validation gap; the code finding independently requires correction. Static review only; no contributor code, tests, snippets, imports, builds, setup or scanner invocations were executed.

Reviewed head: 807b4c09f740d941852a798041c76575144a615f.

[rng1995]

[SkillSpector Review]

Decision: Changes Requested.

Value and readiness

The standard platform adapter remains the right Windows-drive correction for #485, but the current head still retains the separately documented localhost double-leading-path failure.

Material findings

Prior blocker still open — scripts/compare_scan_accuracy.py:180-188. The authority check accepts localhost, while delimiter restoration is still gated on not parsed.netloc. Therefore file://localhost//path reaches url2pathname as //path. The current official CPython 3.14 converter prefixes file:, reparses path as the authority, and rejects a nonlocal authority on POSIX with URLError. Apply authority-safe delimiter preservation to both accepted local authorities and cover the localhost form with the real adapter. tests/unit/test_compare_scan_accuracy.py:1120-1173 still supplies only the drive/empty-authority cases and replaces the adapter with a mock, so it cannot detect this failure. The existing unresolved, non-outdated thread already contains the issue and another reviewer's reproduction; no duplicate inline finding is needed.

Prior issue status: I checked all 16 completed marked reviews by rng1995. The original empty-authority corruption concern was explicitly withdrawn in the September 12 approval and stays withdrawn. The later localhost correction and real-adapter regression request remain unaddressed; no exonerated issue is being reopened.

PIC tradeoffs

None identified. This remains within development-tool portability. The earlier variant #486 is closed; it does not need to land alongside this PR.

Verification and gaps

Read the complete two-file diff, pinned script and regression assertions, supported-Python metadata (including 3.14), linked #485/#486, every prior marked review and complete thread/comment evidence. Re-read official CPython 3.14 source as data; Python 3.14 execution was not performed. No checks attach to the exact head. Its CI run 42309939806 reports action_required, which is a workflow-authorization/validation gap rather than evidence of a failing test. GitHub reports MERGEABLE/BLOCKED, and the other reviewer's Changes Requested remains active.
Static source review only. No contributor code, tests, snippets, imports, builds, setup, hooks, scanner invocations or model calls were executed.

Reviewed head: 9362bc8c149093486f9afc7f3526d8182b16c060.