Commit 1c5e761b authored by Jan Reimes's avatar Jan Reimes
Browse files

feat(cli): enhance CLI with hybrid server checks and async handling 

* Add ensure_hybrid_server_for_profile function to validate hybrid server availability.
* Update handle_clear_options calls in meetings, specs, and tdocs to run asynchronously.
* Replace resolve_spec_release with resolve_spec_release_from_db in members.py.
* Introduce normalize_spec_number_compact function for compact spec number formatting.
parent bcbaa943

src/tdoc_crawler/cli/_shared.py

+18 −0
Original line number Diff line number Diff line
@@ -7,6 +7,8 @@ from pathlib import Path 
from rich.progress import BarColumn, MofNCompleteColumn, Progress, SpinnerColumn, TaskID, TextColumn



from tdoc_crawler.database.base import DocDatabase

from tdoc_crawler.extraction.hybrid_server import ensure_hybrid_server

from tdoc_crawler.extraction.profiles import ExtractionProfile

from tdoc_crawler.logging import get_console

from tdoc_crawler.specs.operations.checkout import clear_checkout_specs

from tdoc_crawler.tdocs.operations.checkout import clear_checkout_tdocs
@@ -82,8 +84,24 @@ def create_progress_bar(description: str, total: float = 100) -> tuple[Progress, 
    return progress, task





def ensure_hybrid_server_for_profile(profile: ExtractionProfile) -> bool:

    """Ensure hybrid server is available for extraction profiles that require it."""

    if profile == ExtractionProfile.PDF_ONLY:

        return True



    console.print(f"[dim]Ensuring hybrid server for profile '{profile.value}'...[/dim]")

    _, status = ensure_hybrid_server(auto_start=True, progress_callback=lambda message: console.print(f"[dim]{message}[/dim]"))

    if status.running:

        return True



    error = status.error or "unknown startup failure"

    console.print(f"[red]Hybrid server is not available: {error}[/red]")

    return False





__all__ = [

    "console",

    "create_progress_bar",

    "ensure_hybrid_server_for_profile",

    "handle_clear_options",

]

src/tdoc_crawler/cli/crawl/meetings.py

+9 −7
Original line number Diff line number Diff line
@@ -96,6 +96,7 @@ def crawl_meetings( 
        scope_parts.append(f"working groups: {', '.join(wg.value for wg in working_groups)}")

    console.print(f"[cyan]Crawling meetings ({', '.join(scope_parts)})[/cyan]")



    asyncio.run(

        handle_clear_options(

            db_file,

            crawler_config.path.checkout_dir,
@@ -103,6 +104,7 @@ def crawl_meetings( 
            clear_tdocs=clear_tdocs,

            clear_specs=clear_specs,

            clear_db=clear_db,

        ),

    )



    async def run_meeting_crawl() -> object:

src/tdoc_crawler/cli/crawl/specs.py

+8 −6
Original line number Diff line number Diff line
@@ -57,12 +57,14 @@ def crawl_specs( 


    sources = build_default_spec_sources()



    asyncio.run(

        handle_clear_options(

            crawler_config.path.db_file,

            crawler_config.path.checkout_dir,

            SpecDatabase,

            clear_tdocs=clear_tdocs,

            clear_specs=clear_specs,

        ),

    )



    async def crawl_specs_db() -> list:

src/tdoc_crawler/cli/crawl/tdocs.py

+8 −6
Original line number Diff line number Diff line
@@ -128,12 +128,14 @@ def crawl_tdocs( 


    console.print(f"[cyan]Crawling TDocs ({', '.join(scope_parts)})[/cyan]")



    asyncio.run(

        handle_clear_options(

            db_file,

            crawler_config.path.checkout_dir,

            TDocDatabase,

            clear_tdocs=clear_tdocs,

            clear_specs=clear_specs,

        ),

    )



    async def run_tdoc_crawl() -> tuple:

src/tdoc_crawler/cli/workspace/members.py

+2 −2
Original line number Diff line number Diff line
@@ -31,7 +31,6 @@ from tdoc_crawler.config import PathConfig 
from tdoc_crawler.database.tdocs import TDocDatabase

from tdoc_crawler.models.base import OutputFormat, SortOrder

from tdoc_crawler.models.workspaces import SourceKind

from tdoc_crawler.specs.operations.checkout import resolve_spec_release

from tdoc_crawler.tdocs.models import TDocQueryConfig

from tdoc_crawler.utils.date_parser import parse_partial_date

from tdoc_crawler.workspaces import (
@@ -42,6 +41,7 @@ from tdoc_crawler.workspaces import ( 
    normalize_workspace_name,

    remove_workspace_member,

)

from tdoc_crawler.workspaces.utils import resolve_spec_release_from_db





def workspace_members(
@@ -78,7 +78,7 @@ def _resolve_spec_release_for_add( 
    if source_kind != SourceKind.SPEC:

        return release

    try:

        resolved = asyncio.run(resolve_spec_release(item, release, auto_crawl=auto_crawl))

        resolved, _ = asyncio.run(resolve_spec_release_from_db(item, release))

        console.print(f"[dim]  Resolved {item} release '{release}' → {resolved}[/dim]")

        return resolved

    except ValueError as exc: