#4236: Fix file touch after resolving path

[nikdenalb]

Summary

• Enabled tests-resolves-and-touches for fs.file.
• Fixed resolved file path handling so touch works after resolved.
• Removed the obsolete @todo #4096 marker.

Test plan

• mvn -pl eo-runtime test -Dtest="EOfileTest#tests_resolves_and_touches"
• mvn clean install -PskipITs
• mvn clean install -Pqulice on Ubuntu 24.04 with OpenJDK 21

Closes #4236.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[github-actions]

🚀 Performance Analysis

All benchmarks are within the acceptable range. No critical degradation detected (threshold is 100%). Please refer to the detailed report for more information.

Click to see the detailed report

Test	Base Score	PR Score	Change	% Change	Unit	Mode
benchmarks.XmirBench.xmirToEO	12772.743	12682.603	-90.140	-0.71%	ms/op	Average Time

✅ Performance gain: benchmarks.XmirBench.xmirToEO is faster by 90.140 ms/op (0.71%)

[akmhatey-ai]

Reviewed current head 0d66148.

The patch is limited to eo-runtime/src/main/eo/fs/file.eo and does the two things needed for #4236: it changes tests-resolves-and-touches from a disabled test to an enabled +> test, and it makes the resolved nested directory before creating the temp file there. The assertion now checks that the touched file exists and that its dirname equals the resolved directory path, so the test covers touch after path resolution without relying on substring matching.

Validation:

• Issue and same-scope PR readback: #5168 is the only open PR I found for #4236/resolves-and-touches.
• GitHub Actions are green for mvn on Ubuntu, macOS, and Windows, plus qulice, integration, snippets, reuse, xcop, and the rest of the runtime checks.
• The only red check, pr-size, fails before running because GitHub cannot resolve maidsafe/pr_size_checker; that is an infrastructure/action-reference failure, not a patch failure.
• git diff --check origin/master...HEAD passed.
• Diff-only Gitleaks over origin/master...HEAD found no leaks.
• Local source review confirmed the new test uses existing fs.path dirname and fs.dir tmpfile patterns.

Local limitation: I did not run Maven or EO tests on this host because Java, Maven, and eo are not installed here.