Spaces:
Sleeping
Sleeping
| """Pure offline catalog loader. No network.""" | |
| from __future__ import annotations | |
| import json | |
| import re | |
| from dataclasses import dataclass, field | |
| from pathlib import Path | |
| from scripts.lib.term_extraction import build_regex | |
| class Catalog: | |
| records: list[dict] | |
| term_to_docs: dict[str, list[str]] | |
| concept_graph: dict[str, dict] | |
| embedding_model: str | |
| embedding_dim: int | |
| term_regex: re.Pattern | |
| registry: dict = field(default_factory=dict) | |
| chunk_to_referencing_cards: dict[str, list[dict]] = field(default_factory=dict) | |
| def load_catalog(derived_dir: Path, *, repo_root: Path | None = None) -> Catalog: | |
| """Load student-chat catalog and supporting layers from derived/ artifacts. | |
| repo_root defaults to derived_dir.parent.parent (assumes <repo>/derived/student-chat/ | |
| layout); pass explicitly for non-standard paths. | |
| """ | |
| if repo_root is None: | |
| repo_root = derived_dir.parent.parent | |
| catalog_path = derived_dir / "catalog.jsonl" | |
| if not catalog_path.is_file(): | |
| raise FileNotFoundError(f"catalog.jsonl not found at {catalog_path}") | |
| records = [json.loads(line) for line in catalog_path.read_text(encoding="utf-8").splitlines() if line.strip()] | |
| if not records: | |
| raise ValueError(f"catalog.jsonl at {catalog_path} is empty") | |
| term_to_docs: dict[str, list[str]] = {} | |
| if (derived_dir / "term-to-docs.json").is_file(): | |
| term_to_docs = json.loads((derived_dir / "term-to-docs.json").read_text(encoding="utf-8")) | |
| concept_graph: dict[str, dict] = {} | |
| if (derived_dir / "concept-graph.json").is_file(): | |
| concept_graph = json.loads((derived_dir / "concept-graph.json").read_text(encoding="utf-8")) | |
| chunk_to_referencing_cards: dict[str, list[dict]] = {} | |
| rev_path = derived_dir / "chunk_to_referencing_cards.json" | |
| if rev_path.is_file(): | |
| chunk_to_referencing_cards = json.loads(rev_path.read_text(encoding="utf-8")) | |
| cfg_path = derived_dir / "build-config.json" | |
| if not cfg_path.is_file(): | |
| raise FileNotFoundError(f"build-config.json not found at {cfg_path}") | |
| cfg = json.loads(cfg_path.read_text(encoding="utf-8")) | |
| embedding_model = cfg["embedding_model"] | |
| embedding_dim = cfg["embedding_dimensions"] | |
| reg_path = repo_root / "content" / "syntopicon-mgrecon780" / "_registry.json" | |
| registry: dict = {} | |
| if reg_path.is_file(): | |
| registry = json.loads(reg_path.read_text(encoding="utf-8")) | |
| term_regex = build_regex(registry) if registry else re.compile(r"$^") | |
| return Catalog( | |
| records=records, term_to_docs=term_to_docs, concept_graph=concept_graph, | |
| embedding_model=embedding_model, embedding_dim=embedding_dim, | |
| term_regex=term_regex, registry=registry, | |
| chunk_to_referencing_cards=chunk_to_referencing_cards, | |
| ) | |