fix(reference): preserve distinct parse failures

src/application/ReferenceSourceImport.ts

@@ -1,6 +1,7 @@
 import { isSecretLikePath } from "../domain/referenceSourceClassification.js";
 import {
   createHistoricalSourceGraph,
+  historicalSourceParseFailureKey,
   type HistoricalSourceGraph,
   type HistoricalSourceGraphInput,
 } from "../domain/referenceSourceGraph.js";
@@ -66,6 +67,27 @@ const deduplicateRelationships = (
   return result;
 };
 
+/** Sort parse failures deterministically while retaining distinct reasons. */
+export const normalizeHistoricalSourceParseFailures = (
+  failures: HistoricalSourceGraphInput["parse_failures"],
+): HistoricalSourceGraphInput["parse_failures"] =>
+  [...failures]
+    .sort((left, right) =>
+      byCodePoint(
+        historicalSourceParseFailureKey(left),
+        historicalSourceParseFailureKey(right),
+      ),
+    )
+    .filter((value, index, array) => {
+      if (index === 0) return true;
+      const previous = array[index - 1];
+      return (
+        previous === undefined ||
+        historicalSourceParseFailureKey(value) !==
+          historicalSourceParseFailureKey(previous)
+      );
+    });
+
 const buildProvenance = (
   options: ReferenceSourceImportOptions,
 ): HistoricalSourceGraphInput["provenance"] => ({
@@ -204,18 +226,7 @@ export const importReferenceSource = async (
   if (isAborted(options.signal)) return err(cancelled());
 
   const uniqueRelationships = deduplicateRelationships(relationships);
-  const uniqueFailures = [...parseFailures]
-    .sort((left, right) => {
-      const byPath = byCodePoint(left.path, right.path);
-      if (byPath !== 0) return byPath;
-      return byCodePoint(left.parser, right.parser);
-    })
-    .filter((value, index, array) => {
-      if (index === 0) return true;
-      const previous = array[index - 1];
-      if (previous === undefined) return true;
-      return value.path !== previous.path || value.parser !== previous.parser;
-    });
+  const uniqueFailures = normalizeHistoricalSourceParseFailures(parseFailures);
 
   const sortedEntries = [...entries].sort((left, right) =>
     byCodePoint(left.path, right.path),

src/domain/referenceSourceGraph.ts

@@ -225,6 +225,11 @@ export type HistoricalSourceManifest = z.infer<
   typeof historicalSourceManifestBaseSchema
 >;
 
+/** Stable identity used to sort and deduplicate source parse failures. */
+export const historicalSourceParseFailureKey = (
+  failure: HistoricalSourceGraphInput["parse_failures"][number],
+): string => `${failure.path}\u0000${failure.parser}\u0000${failure.reason}`;
+
 const canonicalJson = (value: unknown): string => {
   const serialized = canonicalize(value);
   if (serialized === undefined)
@@ -400,7 +405,7 @@ function checkGraphInvariants(
     context,
   );
   checkSortedUnique(
-    graph.parse_failures.map(({ path, parser }) => `${path}\u0000${parser}`),
+    graph.parse_failures.map(historicalSourceParseFailureKey),
     "parse_failures",
     context,
   );

tests/referenceSourceGraph.test.ts

@@ -5,6 +5,7 @@ import {
   createHistoricalSourceGraph,
   createHistoricalSourceManifest,
   historicalSourceGraphSchema,
+  historicalSourceParseFailureKey,
   parseHistoricalSourceGraph,
   parseHistoricalSourceManifest,
   type HistoricalSourceGraphInput,
@@ -192,6 +193,29 @@ describe("historical source graph", () => {
       ).toThrow(/Complete inventory/u);
   });
 
+  it("keeps distinct parse failure reasons for the same path and parser", () => {
+    const input = graphInput();
+    const malformed = {
+      path: "src/main.ts",
+      parser: "typescript",
+      reason: "Malformed input",
+    };
+    const unexpected = {
+      path: "src/main.ts",
+      parser: "typescript",
+      reason: "Unexpected token",
+    };
+    input.inventory_state = "partial";
+    input.parse_failures = [malformed, unexpected];
+
+    expect(createHistoricalSourceGraph(input).parse_failures).toEqual(
+      input.parse_failures,
+    );
+    expect(historicalSourceParseFailureKey(malformed)).not.toBe(
+      historicalSourceParseFailureKey(unexpected),
+    );
+  });
+
   it("rejects strict unknown fields, invalid VCS, dangling internal edges, and false manifests", () => {
     expect(() =>
       createHistoricalSourceGraph({ ...graphInput(), surprise: true }),

tests/referenceSourceImport.test.ts

@@ -6,7 +6,10 @@ import { join } from "node:path";
 import { add, commit, init } from "isomorphic-git";
 import { describe, expect, it } from "vitest";
 
-import { importReferenceSource } from "../src/application/ReferenceSourceImport.js";
+import {
+  importReferenceSource,
+  normalizeHistoricalSourceParseFailures,
+} from "../src/application/ReferenceSourceImport.js";
 import { projectReferenceSourceEntryFailure } from "../src/application/ReferenceSourceImportEntries.js";
 import {
   projectReferenceSourceImportError,
@@ -52,6 +55,27 @@ const importTree = (
   });
 
 describe("reference source import", () => {
+  it("deduplicates exact parse failures without collapsing distinct reasons", () => {
+    const malformed = {
+      path: "src/main.ts",
+      parser: "babel",
+      reason: "Malformed input",
+    };
+    const unexpected = {
+      path: "src/main.ts",
+      parser: "babel",
+      reason: "Unexpected token",
+    };
+
+    expect(
+      normalizeHistoricalSourceParseFailures([
+        unexpected,
+        malformed,
+        unexpected,
+      ]),
+    ).toEqual([malformed, unexpected]);
+  });
+
   it("projects entry failures without low-level reader diagnostics", () => {
     for (const [kind, code] of [
       ["directory", "io"],