fix(pdf): validate page selections before loading

steipete · steipete · commit f3233b1c22b9 · 2026-07-06T01:33:02.000-04:00
diff --git a/src/agents/tools/pdf-tool.helpers.ts b/src/agents/tools/pdf-tool.helpers.ts
@@ -65,8 +65,10 @@ export function parsePageRange(range: string, maxPages: number): number[] {
         pages.add(i);
       }
     } else {
-      const pageMatch = /^\d+$/.exec(part);
-      const num = pageMatch ? Number(pageMatch[0]) : Number.NaN;
+      if (!/^\d+$/.test(part)) {
+        throw new Error(`Invalid page number: "${part}"`);
+      }
+      const num = Number(part);
       if (!Number.isFinite(num) || num < 1) {
         throw new Error(`Invalid page number: "${part}"`);
       }
diff --git a/src/agents/tools/pdf-tool.test.ts b/src/agents/tools/pdf-tool.test.ts
@@ -568,11 +568,14 @@ describe("createPdfTool", () => {
     });
   });
 
-  it.each(["1.5", "1,2.5"])(
-    "rejects fractional page selection %s before fallback extraction",
-    async (pages) => {
+  it.each([
+    ["1.5", "1.5"],
+    ["1,2.5", "2.5"],
+  ])(
+    "rejects fractional page selection %s before loading or fallback extraction",
+    async (pages, invalidPage) => {
       await withTempPdfAgentDir(async (agentDir) => {
-        await stubPdfToolInfra(agentDir, {
+        const { loadSpy } = await stubPdfToolInfra(agentDir, {
           provider: "openai",
           api: "openai-responses",
           input: ["text"],
@@ -590,7 +593,8 @@ describe("createPdfTool", () => {
             pdf: "/tmp/doc.pdf",
             pages,
           }),
-        ).rejects.toThrow(`Invalid page number: "${pages.includes(",") ? "2.5" : pages}"`);
+        ).rejects.toThrow(`Invalid page number: "${invalidPage}"`);
+        expect(loadSpy).not.toHaveBeenCalled();
         expect(extractSpy).not.toHaveBeenCalled();
         expect(completeMock).not.toHaveBeenCalled();
       });
diff --git a/src/agents/tools/pdf-tool.ts b/src/agents/tools/pdf-tool.ts
@@ -374,6 +374,7 @@ export function createPdfTool(options?: {
 
       // Parse page range
       const pagesRaw = normalizeOptionalString(record.pages);
+      const pageNumbers = pagesRaw ? parsePageRange(pagesRaw, configuredMaxPages) : undefined;
       const password = typeof record.password === "string" ? record.password : undefined;
 
       const pdfModelConfig =
@@ -495,8 +496,6 @@ export function createPdfTool(options?: {
         });
       }
 
-      const pageNumbers = pagesRaw ? parsePageRange(pagesRaw, configuredMaxPages) : undefined;
-
       const getExtractions = async (): Promise<PdfExtractedContent[]> => {
         const extractedAll: PdfExtractedContent[] = [];
         for (const pdf of loadedPdfs) {

Original file line number	Diff line number	Diff line change
`@@ -65,8 +65,10 @@ export function parsePageRange(range: string, maxPages: number): number[] {`
`65`	`65`	`pages.add(i);`
`66`	`66`	`}`
`67`	`67`	`} else {`
`68`		`- const pageMatch = /^\d+$/.exec(part);`
`69`		`- const num = pageMatch ? Number(pageMatch[0]) : Number.NaN;`
	`68`	`+ if (!/^\d+$/.test(part)) {`
	`69`	+ throw new Error(`Invalid page number: "${part}"`);
	`70`	`+ }`
	`71`	`+ const num = Number(part);`
`70`	`72`	`if (!Number.isFinite(num) \|\| num < 1) {`
`71`	`73`	throw new Error(`Invalid page number: "${part}"`);
`72`	`74`	`}`