Commit a30c2c71 authored by Jan Reimes's avatar Jan Reimes
Browse files

refactor(meetings/utils): update normalize_subgroup_alias return type and logic 

- Change return type documentation to reflect a single enum value.
- Adjust logic to append the short code of the normalized subgroup.
- Improve clarity in the function's purpose and output.
parent d115aeef

src/tdoc_crawler/meetings/utils.py

+2 −2
Original line number Diff line number Diff line
@@ -47,7 +47,7 @@ def normalize_working_group_alias(alias: str) -> WorkingGroup: 
def normalize_subgroup_alias(alias: str) -> SubWorkingGroup:

    """Normalize subgroup aliases to canonical subgroup enums.



    Returns a list of possible matching subgroup enum values (e.g., SubWorkingGroup.S4).

    Returns a SubWorkingGroup enum value (e.g., SubWorkingGroup.S4).

    Supports:

    - Short codes: S4, R1, C3

    - Full names: SA4, RAN1, CT3
@@ -58,7 +58,7 @@ def normalize_subgroup_alias(alias: str) -> SubWorkingGroup: 
        alias: Subgroup alias or code



    Returns:

        List of possible matching canonical subgroup enum values

        Canonical subgroup enum value (e.g., SubWorkingGroup.S4)

    """

    input_str = alias.strip().upper()

    cleaned = re.sub(r"[#\-_\s]", "", input_str)

src/tdoc_crawler/tdocs/operations/crawl.py

+6 −10
Original line number Diff line number Diff line
@@ -238,12 +238,10 @@ class TDocCrawler: 
                    meeting.meeting_id,

                    base_url,

                    meeting.short_name or "",

                    config.timeout,

                    config.max_retries,

                    target_tuple,

                    str(config.cache_dir),

                    config.http_cache.ttl,

                    config.http_cache.refresh_ttl_on_access,

                    config.timeout,

                    config.cache_manager_name,

                    config.http_cache,

                )

                futures[future] = meeting
@@ -573,12 +571,10 @@ class HybridTDocCrawler: 
                        meeting.meeting_id,

                        base_url,

                        meeting.short_name or "",

                        config.timeout,

                        config.max_retries,

                        target_tuple,

                        str(config.cache_dir),

                        config.http_cache.ttl,

                        config.http_cache.refresh_ttl_on_access,

                        config.timeout,

                        config.cache_manager_name,

                        config.http_cache,

                    )

                )

                tasks[task] = meeting

src/tdoc_crawler/utils/parse.py

+2 −1
Original line number Diff line number Diff line
@@ -97,7 +97,8 @@ def parse_subgroups(values: list[str] | None) -> list[str] | None: 
            _logger.warning(f"Unknown subgroup: {item}")

            raise typer.Exit(code=2)



        resolved.extend(normalized)

        # Use .name to get the short code (e.g., "S4" not "SA4")

        resolved.append(normalized.name)



    return resolved