From 302e9946087e1bbd6251c94e19fe3286c01d8982 Mon Sep 17 00:00:00 2001
From: joaovictor91123 <joaovictor91123@gmail.com>
Date: Sun, 26 Jul 2026 06:42:48 -0700
Subject: [PATCH] fix(content-lane): harden duplicates block-scalar header
 detection

Mirror source-evidence.ts block-scalar grammar so chomping/indent permutations and trailing inline comments are recognized in parseSimpleFrontmatter and findDuplicateFrontmatterKeys.

Closes #9290
---
 src/review/content-lane/duplicates.ts     | 17 ++++++++-
 test/unit/content-lane-duplicates.test.ts | 46 +++++++++++++++++++++++
 2 files changed, 61 insertions(+), 2 deletions(-)

diff --git a/src/review/content-lane/duplicates.ts b/src/review/content-lane/duplicates.ts
index 64fb1cdcbd..b23100aa07 100644
--- a/src/review/content-lane/duplicates.ts
+++ b/src/review/content-lane/duplicates.ts
@@ -40,6 +40,10 @@ export type ContentDuplicateReview = {
   relatedCandidates: ContentDuplicateMatch[];
 };
 
+function stripYamlComment(value: string): string {
+  return value.replace(/\s+#.*$/, "").trim();
+}
+
 function unquoteYamlScalar(value: string): string {
   const trimmed = value.trim();
   if ((trimmed.startsWith('"') && trimmed.endsWith('"')) || (trimmed.startsWith("'") && trimmed.endsWith("'"))) {
@@ -48,6 +52,15 @@ function unquoteYamlScalar(value: string): string {
   return trimmed.replace(/\s+#.*$/, "").trim();
 }
 
+// A YAML block-scalar header indicator: `|` or `>` with an optional chomping (`+`/`-`) and/or a single indentation
+// digit, in EITHER order — `|`, `>`, `|-`, `>+`, `|2`, `|2-`, `|-2`, `>2+`. Kept in one place so parseSimpleFrontmatter
+// and findDuplicateFrontmatterKeys agree on what is a block-scalar header (mirrors source-evidence.ts).
+const BLOCK_SCALAR_INDICATOR = /^[|>](?:[+-]?\d?|\d[+-]?)$/;
+
+function isBlockScalarHeader(raw: string): boolean {
+  return BLOCK_SCALAR_INDICATOR.test(stripYamlComment(raw));
+}
+
 /**
  * Parse YAML frontmatter into a flat key→string map, capturing each top-level field's value
  * REGARDLESS of scalar style — inline, quoted, block literal (`|`), folded (`>`), or a block/flow
@@ -74,7 +87,7 @@ export function parseSimpleFrontmatter(source: string): Record<string, string> {
     /* v8 ignore next -- noUncheckedIndexedAccess fallback: capture group 2 (.*) always participates when head matches, so head[2] is never undefined */
     const inline = (head[2] ?? "").trim();
     i += 1;
-    if (/^[|>][+-]?\d*$/.test(inline)) {
+    if (isBlockScalarHeader(inline)) {
       // Block literal (`|`) / folded (`>`) scalar: gather the indented block that follows.
       const block: string[] = [];
       /* v8 ignore next -- noUncheckedIndexedAccess fallback: i < lines.length guards the index; split() elements are always strings */
@@ -129,7 +142,7 @@ export function findDuplicateFrontmatterKeys(source: string): string[] {
     /* v8 ignore next -- noUncheckedIndexedAccess fallback: capture group 2 (.*) always participates when head matches, so head[2] is never undefined */
     const inline = (head[2] ?? "").trim();
     i += 1;
-    if (/^[|>][+-]?\d*$/.test(inline)) {
+    if (isBlockScalarHeader(inline)) {
       /* v8 ignore next -- noUncheckedIndexedAccess fallback: i < lines.length guards the index; split() elements are always strings */
       while (i < lines.length && ((lines[i] ?? "").trim() === "" || /^\s/.test(lines[i] ?? ""))) i += 1;
     } else if (inline === "") {
diff --git a/test/unit/content-lane-duplicates.test.ts b/test/unit/content-lane-duplicates.test.ts
index 9d63f1846b..0069b9736f 100644
--- a/test/unit/content-lane-duplicates.test.ts
+++ b/test/unit/content-lane-duplicates.test.ts
@@ -58,6 +58,22 @@ describe("parseSimpleFrontmatter", () => {
     expect(f.title).toBe("Real");
     expect(Object.keys(f)).toEqual(["title"]);
   });
+
+  it("recognizes block-scalar headers with chomping/indent in either order and trailing comments (#9290)", () => {
+    const indicators = ["|", ">", "|-", ">-", "|+", ">+", "|2", ">2", "|-2", "|2-", ">2+", ">2-"];
+    for (const indicator of indicators) {
+      const src = ["---", `description: ${indicator}`, "  line one", "  line two", "title: Real", "---", "", "body"].join("\n");
+      const f = parseSimpleFrontmatter(src);
+      expect(f.description).toBe("line one\nline two");
+      expect(f.title).toBe("Real");
+    }
+    for (const indicator of ["|-", ">", "|2-"]) {
+      const src = ["---", `description: ${indicator} # sources below`, "  only line", "title: Real", "---", "", "body"].join("\n");
+      const f = parseSimpleFrontmatter(src);
+      expect(f.description).toBe("only line");
+      expect(f.title).toBe("Real");
+    }
+  });
 });
 
 describe("findDuplicateFrontmatterKeys", () => {
@@ -283,6 +299,36 @@ describe("findDuplicateFrontmatterKeys — block-scalar + sequence skipping", ()
     // The blank + comment lines fail the key regex → the `if (!head) continue` branch runs; no dupes.
     expect(findDuplicateFrontmatterKeys(src)).toEqual([]);
   });
+
+  it("skips block-scalar headers with chomping/indent in either order and trailing comments (#9290)", () => {
+    const indicators = ["|", ">", "|-", ">-", "|+", ">+", "|2", ">2", "|-2", "|2-", ">2+", ">2-"];
+    for (const indicator of indicators) {
+      const src = [
+        "---",
+        `description: ${indicator}`,
+        "  title: not-a-real-key",
+        "title: Real",
+        "title: DupReal",
+        "---",
+        "",
+        "body",
+      ].join("\n");
+      expect(findDuplicateFrontmatterKeys(src)).toEqual(["title"]);
+    }
+    for (const indicator of ["|-", ">", "|2-"]) {
+      const src = [
+        "---",
+        `description: ${indicator} # sources below`,
+        "  title: not-a-real-key",
+        "title: Real",
+        "title: DupReal",
+        "---",
+        "",
+        "body",
+      ].join("\n");
+      expect(findDuplicateFrontmatterKeys(src)).toEqual(["title"]);
+    }
+  });
 });
 
 describe("normalizeUrl edge cases (via extractContentDuplicateSignals)", () => {