Class Categorizer

scripts/release_notes/categorize.py:22–188

class Categorizer:
    def __init__(self, path, category="Uncategorized", use_classifier: bool = False):
        self.cache = get_commit_data_cache()
        self.commits = CommitList.from_existing(path)
        if use_classifier:
            print("Using a classifier to aid with categorization.")
            device = "cuda" if torch.cuda.is_available() else "cpu"
            classifier_config = CategoryConfig(common.categories)
            author_map = get_author_map(
                Path("results/classifier"), regen_data=False, assert_stored=True
            )
            file_map = get_file_map(
                Path("results/classifier"), regen_data=False, assert_stored=True
            )
            self.classifier = CommitClassifier(
                XLMR_BASE, author_map, file_map, classifier_config
            ).to(device)
            self.classifier.load_state_dict(
                torch.load(Path("results/classifier/commit_classifier.pt"))
            )
            self.classifier.eval()
        else:
            self.classifier = None
        # Special categories: 'Uncategorized'
        # All other categories must be real
        self.category = category

    def categorize(self):
        commits = self.commits.filter(category=self.category)
        total_commits = len(self.commits.commits)
        already_done = total_commits - len(commits)
        i = 0
        while i < len(commits):
            cur_commit = commits[i]
            next_commit = commits[i + 1] if i + 1 < len(commits) else None
            jump_to = self.handle_commit(
                cur_commit, already_done + i + 1, total_commits, commits
            )

            # Increment counter
            if jump_to is not None:
                i = jump_to
            elif next_commit is None:
                i = len(commits)
            else:
                i = commits.index(next_commit)

    def features(self, commit):
        return self.cache.get(commit.commit_hash)

    def potential_reverts_of(self, commit, commits):
        submodule_update_str = [
            "Update TensorPipe submodule",
            "Updating submodules",
            "Automated submodule update",
        ]
        if any(a in commit.title for a in submodule_update_str):
            return []