Codebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP.
Python aicode-complexitycode-healthcode-intelligencecode-qualitydead-codedeveloper-toolsdocumentationgit-analyticsmcpopen-sourcepythonrefactoringstatic-analysistechnical-debt
/**
* Pure helpers extracted from GraphFlow to remove duplication.
*/
/**
* Build the set of directed edge keys for a node trace/path, in both
* directions. Used by both the execution-flow highlighter and the path-finder
* result handler — they highlight the same way, so the key construction lives
* here once. Mirrors Sigma's `source→target` edge-key convention.
*/
/**
* Collapse an execution-flow trace onto the files it runs through.
*
* `calls` edges only ever connect symbol nodes, so a trace is a list of
* `path/to/file.py::symbol` ids — but this canvas draws files. Highlighting the
* raw trace therefore matched only the handful of symbol nodes that happened to
* survive the export's node cap, and matches nothing at all now that the export
* is file-only. Mapping each step onto its containing file highlights the route
* the reader can actually see, and collapses consecutive steps within one file
* so a 6-symbol trace through 2 files reads as 2 hops, not 6.
*/
export function traceToFileTrace(trace: readonly string[]): string[] {
const files: string[] = [];
for (const id of trace) {
const file = id.split("::")[0] ?? id;
if (files[files.length - 1] !== file) files.push(file);
}
return files;
}
export function traceToEdgeKeys(nodes: readonly string[]): Set<string> {
const edgeKeys = new Set<string>();
for (let i = 0; i < nodes.length - 1; i++) {
edgeKeys.add(`${nodes[i]}→${nodes[i + 1]}`);
edgeKeys.add(`${nodes[i + 1]}→${nodes[i]}`);
}
return edgeKeys;
}