HBASE-24845 Git/Jira Release Audit: limit branches when building audit db (#2238)

Populating the audit database with release tag information from git is
time consuming. Until that's sorted out, give the user a flag for
limiting which branches they want to be reviewed.

Signed-off-by:  Andrew Purtell <apurtell@apache.org >

dev-support/git-jira-release-audit/README.md

@@ -62,6 +62,7 @@ usage: git_jira_release_audit.py [-h] [--populate-from-git POPULATE_FROM_GIT]
                                  [--release-line-regexp RELEASE_LINE_REGEXP]
                                  [--parse-release-tags PARSE_RELEASE_TAGS]
                                  [--fallback-actions-path FALLBACK_ACTIONS_PATH]
+                                 [--branch-filter-regexp BRANCH_FILTER_REGEXP]
                                  [--jira-url JIRA_URL] --branch-1-fix-version
                                  BRANCH_1_FIX_VERSION --branch-2-fix-version
                                  BRANCH_2_FIX_VERSION
@@ -119,6 +120,9 @@ Interactions with the Git repo:
   --fallback-actions-path FALLBACK_ACTIONS_PATH
                         Path to a file containing _DB.Actions applicable to
                         specific git shas. (default: fallback_actions.csv)
+  --branch-filter-regexp BRANCH_FILTER_REGEXP
+                        Limit repo parsing to branch names that match this
+                        filter expression. (default: .*)
   --branch-1-fix-version BRANCH_1_FIX_VERSION
                         The Jira fixVersion used to indicate an issue is
                         committed to the specified release line branch
@@ -175,8 +179,9 @@ fetch from Jira 100%|███████████████████
 
 Optionally, the database can be build to include release tags, by specifying
 `--parse-release-tags=true`. This is more time-consuming, but is necessary for
-auditing discrepancies between git and Jira. Running the same command but
+auditing discrepancies between git and Jira. Optionally, limit the branches
-including this flag looks like this:
+under consideration by specifying a regex filter with `--branch-filter-regexp`.
+Running the same command but including this flag looks like this:
 
 ```shell script
 origin/branch-1 100%|███████████████████████████████████████| 4084/4084 [08:58<00:00, 7.59 commit/s]

dev-support/git-jira-release-audit/git_jira_release_audit.py

@@ -199,13 +199,14 @@ class _RepoReader:
     _identify_amend_jira_id_pattern = re.compile(r'^amend (.+)', re.IGNORECASE)
 
     def __init__(self, db, fallback_actions_path, remote_name, development_branch,
-                 release_line_regexp, parse_release_tags, **_kwargs):
+                 release_line_regexp, branch_filter_regexp, parse_release_tags, **_kwargs):
         self._db = db
         self._repo = _RepoReader._open_repo()
         self._fallback_actions = _RepoReader._load_fallback_actions(fallback_actions_path)
         self._remote_name = remote_name
         self._development_branch = development_branch
         self._release_line_regexp = release_line_regexp
+        self._branch_filter_regexp = branch_filter_regexp
         self._parse_release_tags = parse_release_tags
 
     @property
@@ -364,6 +365,10 @@ class _RepoReader:
             release_branch (str): The name of the ref whose history is to be parsed.
         """
         global MANAGER
+        branch_filter_pattern = re.compile('%s/%s' % (self._remote_name, self._branch_filter_regexp))
+        if not branch_filter_pattern.match(release_branch):
+            return
+
         commits = list(self._repo.iter_commits(
             "%s...%s" % (origin_commit.hexsha, release_branch), reverse=True))
         LOG.info("%s has %d commits since its origin at %s.", release_branch, len(commits),
@@ -638,6 +643,10 @@ class Auditor:
             '--fallback-actions-path',
             help='Path to a file containing _DB.Actions applicable to specific git shas.',
             default='fallback_actions.csv')
+        git_repo_group.add_argument(
+            '--branch-filter-regexp',
+            help='Limit repo parsing to branch names that match this filter expression.',
+            default=r'.*')
         jira_group = parser.add_argument_group('Interactions with Jira')
         jira_group.add_argument(
             '--jira-url',