Function build

graphify/build.py:635–670

Merge multiple extraction results into one graph. directed=True produces a DiGraph that preserves edge direction (source→target). directed=False (default) produces an undirected Graph for backward compatibility. dedup=True (default) runs entity deduplication before building the graph.

(
    extractions: list[dict],
    *,
    directed: bool = False,
    dedup: bool = True,
    dedup_llm_backend: str | None = None,
    root: str | Path | None = None,
)

def build(
    extractions: list[dict],
    *,
    directed: bool = False,
    dedup: bool = True,
    dedup_llm_backend: str | None = None,
    root: str | Path | None = None,
) -> nx.Graph:
    """Merge multiple extraction results into one graph.

    directed=True produces a DiGraph that preserves edge direction (source→target).
    directed=False (default) produces an undirected Graph for backward compatibility.
    dedup=True (default) runs entity deduplication before building the graph.
    dedup_llm_backend: if set (e.g. "gemini", "claude", or "kimi"), uses LLM to resolve
        ambiguous pairs in the 75–92 Jaro-Winkler score zone.
    root: if given, absolute source_file paths are made relative to root (#932).

    Extractions are merged in order. For nodes with the same ID, the last
    extraction's attributes win (NetworkX add_node overwrites). Pass AST
    results before semantic results so semantic labels take precedence, or
    reverse the order if you prefer AST source_location precision to win.
    """
    from graphify.dedup import deduplicate_entities
    combined: dict = {"nodes": [], "edges": [], "hyperedges": [], "input_tokens": 0, "output_tokens": 0}
    for ext in extractions:
        combined["nodes"].extend(ext.get("nodes", []))
        combined["edges"].extend(ext.get("edges", []))
        combined["hyperedges"].extend(ext.get("hyperedges", []))
        combined["input_tokens"] += ext.get("input_tokens", 0)
        combined["output_tokens"] += ext.get("output_tokens", 0)
    if dedup and combined["nodes"]:
        combined["nodes"], combined["edges"] = deduplicate_entities(
            combined["nodes"], combined["edges"], communities={},
            dedup_llm_backend=dedup_llm_backend,
        )
    return build_from_json(combined, directed=directed, root=root)


def _norm_label(label: str | None) -> str: