chore: partial release script to handle sdk-platform-java

[suztomo]

No description provided.

[gemini-code-assist]

Code Review

This pull request introduces a flag to automate versioning for Java artifacts in partial_release.py and refactors apply_versions.sh to support XML, YAML, and Java files with a progress bar. Feedback identifies potential crashes from missing default values, a division-by-zero risk in the progress calculation, and opportunities to improve the accuracy and performance of the version replacement logic.

[gemini-code-assist]

The --artifact-ids option is now optional, but it lacks a default value. If this option is omitted and --sdk-platform-java is not set, artifact_ids will be None, which will cause an AttributeError when bump_version attempts to call artifact_ids.split(","). Adding default="" ensures the variable is always a string and avoids the crash.

Suggested change

	required=False,
	required=False,
	default="",

[gemini-code-assist]

Similar to the bump-snapshot-version command, the --artifact-ids option here should have a default value to prevent a NoneType error in bump_version when the flag is omitted.

Suggested change

	required=False,
	required=False,
	default="",

[gemini-code-assist]

The sed replacement patterns for XML, YAML, and Java are quite generic and rely on the marker being the only relevant content on the line (or the next line for Java). For example, the XML pattern s|>\([^<]*\)<|>$V<| will replace the first tag's content on the line, which might not be the version tag if multiple tags exist (e.g., <groupId>foo</groupId><version>1.0</version>). \n\nConsider making these patterns more specific by anchoring them to the marker or the expected tag name (e.g., <version>) to avoid accidental replacements in complex files.

[gemini-code-assist]

If no files are found matching the search criteria, TOTAL_FILES will be 0, resulting in a division by zero error in the progress calculation. It is safer to wrap the progress output in a conditional check.

Suggested change

	echo -ne "Progress: $CURRENT_FILE/$TOTAL_FILES ($((CURRENT_FILE * 100 / TOTAL_FILES))%)\r"
	if [ "$TOTAL_FILES" -gt 0 ]; then
	echo -ne "Progress: $CURRENT_FILE/$TOTAL_FILES ($((CURRENT_FILE * 100 / TOTAL_FILES))%)\r"
	fi

[gemini-code-assist]

Spawning a new sed process for every file in a loop is significantly slower than processing all files in a single sed invocation (e.g., using xargs sed -i -f "$SED_SCRIPT"). While the progress bar provides feedback, it introduces a performance bottleneck that may become substantial in large repositories. Consider batching the files or using xargs if execution time becomes a concern.