all2md.search.service

High-level orchestration for indexing and querying documents.

all2md.search.service.compute_corpus_fingerprint(documents: Sequence[SearchDocumentInput], options: SearchOptions) → str: Fingerprint a document set + index-relevant options for cache validation.

class all2md.search.service.SearchDocumentInput

Bases: object

Represents a document scheduled for indexing.

source: str | Path | bytes

document_id: str | None = None

source_format: Literal['auto', 'archive', 'asciidoc', 'ast', 'bbcode', 'chm', 'csv', 'docx', 'dokuwiki', 'eml', 'enex', 'epub', 'fb2', 'html', 'ini', 'ipynb', 'jinja', 'json', 'latex', 'markdown', 'mbox', 'mediawiki', 'mhtml', 'odp', 'ods', 'odt', 'openapi', 'org', 'outlook', 'pdf', 'plaintext', 'pptx', 'rst', 'rtf', 'sourcecode', 'textile', 'toml', 'webarchive', 'xlsx', 'yaml', 'zip'] | str | None = 'auto'

metadata: Mapping[str, object] | None = None

__init__(source: str | Path | bytes, document_id: str | None = None, source_format: Literal['auto', 'archive', 'asciidoc', 'ast', 'bbcode', 'chm', 'csv', 'docx', 'dokuwiki', 'eml', 'enex', 'epub', 'fb2', 'html', 'ini', 'ipynb', 'jinja', 'json', 'latex', 'markdown', 'mbox', 'mediawiki', 'mhtml', 'odp', 'ods', 'odt', 'openapi', 'org', 'outlook', 'pdf', 'plaintext', 'pptx', 'rst', 'rtf', 'sourcecode', 'textile', 'toml', 'webarchive', 'xlsx', 'yaml', 'zip'] | str | None = 'auto', metadata: Mapping[str, object] | None = None) → None

class all2md.search.service.SearchIndexState

Bases: object

Container for the active index backends.

chunks: list[Chunk]

documents: list[tuple[Document, SearchDocumentInput]] | None = None

keyword_index: BM25Index | None = None

vector_index: VectorIndex | None = None

available_modes() → set[SearchMode]: Get available modes of search.

property chunk_count: int: Return number of cached chunks.

__init__(chunks: list[Chunk], documents: list[tuple[Document, SearchDocumentInput]] | None = None, keyword_index: BM25Index | None = None, vector_index: VectorIndex | None = None) → None

class all2md.search.service.SearchService

Bases: object

Service object coordinating indexing and search execution.

Initialise the service with optional search configuration overrides.

__init__(options: SearchOptions | None = None) → None: Initialise the service with optional search configuration overrides.

property state: SearchIndexState: Return the current in-memory search index state.

build_indexes(documents: Sequence[SearchDocumentInput], *, modes: Iterable[SearchMode] | None = None, progress_callback: Callable[[ProgressEvent], None] | None = None) → SearchIndexState: Convert sources into chunks and materialise requested indexes.

save(directory: Path) → None: Persist all active indexes to disk.

static persisted_index_matches(directory: Path, documents: Sequence[SearchDocumentInput], options: SearchOptions) → bool

Return True if a persisted index in directory is still current.

Compares the fingerprint recorded at save time (corpus files + index-relevant options) against the fingerprint of the documents/options supplied now. A missing or unreadable manifest, or any mismatch, returns False — the caller should rebuild rather than serve a stale index.

classmethod load(directory: Path, options: SearchOptions | None = None) → SearchService: Rehydrate indexes from disk.

search(query: str, *, mode: SearchMode | str | None = None, top_k: int = 10, progress_callback: Callable[[ProgressEvent], None] | None = None) → list[SearchResult]: Execute a query using the configured index backends.