fix(python): validate shared pyc header facts

src/application/python/PythonPycService.test.ts

@@ -66,6 +66,28 @@ const report: PythonPyc = {
   ],
 };
 
+it.each([
+  {
+    flags: 1,
+    hash_based: true,
+    source_hash_hex: "",
+    source_mtime: null,
+    source_size: null,
+  },
+  { flags: 1 },
+])("refuses inconsistent provider header facts %j", async (header) => {
+  const result = await new PythonPycService({
+    close: async () => {},
+    identity,
+    inspect: async () =>
+      ok({ ...report, header: { ...report.header, ...header } }),
+  }).inspect({ path: report.artifact.path });
+  expect(result).toMatchObject({
+    ok: false,
+    error: { _tag: "AnalysisOutputError" },
+  });
+});
+
 it("rejects a relative path and extra arguments before reading", async () => {
   let calls = 0;
   const service = new PythonPycService({

src/application/v8CodeCache/V8CodeCacheService.test.ts

@@ -1,68 +0,0 @@
-import { createHash } from "node:crypto";
-import { expect, it } from "vitest";
-
-import { AnalysisOutputError } from "../../domain/analysisErrorCore.js";
-import { parseEvidence } from "../../domain/evidence.js";
-import { ok } from "../../domain/result.js";
-import {
-  V8_CODE_CACHE_HEADER_BYTES,
-  V8_CODE_CACHE_MAGIC_HIGH,
-  type V8CodeCache,
-} from "../../domain/v8/v8CodeCache.js";
-import { parseV8CodeCache } from "../../v8/V8CodeCacheParser.js";
-import { V8CodeCacheService } from "./V8CodeCacheService.js";
-
-const reportFor = (payloadBytes: number): V8CodeCache => {
-  const bytes = Buffer.alloc(V8_CODE_CACHE_HEADER_BYTES + payloadBytes, 0x61);
-  bytes.fill(0, 0, V8_CODE_CACHE_HEADER_BYTES);
-  bytes.writeUInt32LE(V8_CODE_CACHE_MAGIC_HIGH * 0x10000, 0);
-  bytes.writeUInt32LE(payloadBytes, 20);
-  return parseV8CodeCache(
-    bytes,
-    {
-      path: "/artifacts/authored.jsc",
-      sha256: createHash("sha256").update(bytes).digest("hex"),
-      bytes: bytes.length,
-    },
-    { min_string_bytes: 4, max_strings: 8 },
-  );
-};
-
-const inspect = (report: V8CodeCache) =>
-  new V8CodeCacheService({
-    identity: { id: "cache-fixture", name: "Cache fixture", version: "1" },
-    close: async () => {},
-    inspect: async () => ok(report),
-  }).inspect({ path: report.artifact.path });
-
-it.each([0, 9])(
-  "rejects complete coverage claiming %s scanned bytes for an eight-byte payload",
-  async (scanned_bytes) => {
-    const report = reportFor(8);
-    const result = await inspect({
-      ...report,
-      coverage: { ...report.coverage, scanned_bytes },
-    });
-    expect(result.ok).toBe(false);
-    if (!result.ok) expect(result.error).toBeInstanceOf(AnalysisOutputError);
-  },
-);
-
-it.each([8, 5000])(
-  "retains complete scan facts for a %s-byte payload and its bounded string prefix",
-  async (payloadBytes) => {
-    const report = reportFor(payloadBytes);
-    const result = await inspect(report);
-    if (!result.ok) throw result.error;
-    expect(parseEvidence(result.value).normalized_result).toMatchObject({
-      coverage: { status: "complete", scanned_bytes: payloadBytes, strings: 1 },
-      embedded_strings: [
-        {
-          bytes: payloadBytes,
-          retained_bytes: Math.min(payloadBytes, 4096),
-          truncated: payloadBytes > 4096,
-        },
-      ],
-    });
-  },
-);

src/domain/python/pythonPyc.test.ts

@@ -3,9 +3,49 @@ import {
   inspectPythonPycInputSchema,
   pythonPycMarshalProfile,
   PYTHON_PYC_MARSHAL_PROFILES,
+  pythonPycHeaderSchema,
 } from "./pythonPyc.js";
 import { pythonBytecodeIdentity } from "./pythonBytecodeIdentity.js";
 
+const hashHeader = {
+  magic: 3531,
+  python_version: "3.12",
+  code_layout: "py311",
+  flags: 3,
+  hash_based: true,
+  check_source: true,
+  source_hash_hex: "0123456789abcdef",
+  source_mtime: null,
+  source_size: null,
+};
+
+it.each(["", "ab", "g".repeat(16)])(
+  "rejects an invalid source hash %j",
+  (source_hash_hex) => {
+    expect(
+      pythonPycHeaderSchema.safeParse({ ...hashHeader, source_hash_hex })
+        .success,
+    ).toBe(false);
+  },
+);
+
+it.each([
+  { hash_based: false },
+  { check_source: false },
+  { source_hash_hex: null },
+  { source_mtime: 0 },
+  { source_size: 0 },
+])("keeps header facts consistent with the recorded flags: %j", (changed) => {
+  expect(
+    pythonPycHeaderSchema.safeParse({ ...hashHeader, ...changed }).success,
+  ).toBe(false);
+});
+
+it("preserves unknown raw header flags and valid hash bytes", () => {
+  const header = { ...hashHeader, flags: 0x80000003 };
+  expect(pythonPycHeaderSchema.parse(header)).toEqual(header);
+});
+
 it("keeps pinned CPython identity separate from standalone marshal coverage", () => {
   expect(pythonPycMarshalProfile(3394)).toEqual({
     code_layout: "py37",

src/domain/python/pythonPyc.ts

@@ -205,17 +205,53 @@ const pythonCodeObjectSchema = z.strictObject({
 });
 export type PythonPycCodeObject = z.infer<typeof pythonCodeObjectSchema>;
 
-export const pythonPycHeaderSchema = z.strictObject({
-  magic: z.number().int().min(0).max(0xffff),
-  python_version: z.string().min(1),
-  code_layout: z.enum(PYTHON_PYC_CODE_LAYOUTS),
-  flags: uint,
-  hash_based: z.boolean(),
-  check_source: z.boolean(),
-  source_hash_hex: z.string().nullable(),
-  source_mtime: uint.nullable(),
-  source_size: uint.nullable(),
-});
+export const pythonPycHeaderSchema = z
+  .strictObject({
+    magic: z.number().int().min(0).max(0xffff),
+    python_version: z.string().min(1),
+    code_layout: z.enum(PYTHON_PYC_CODE_LAYOUTS),
+    flags: uint,
+    hash_based: z.boolean(),
+    check_source: z.boolean(),
+    source_hash_hex: z
+      .string()
+      .regex(/^[0-9a-f]{16}$/u)
+      .nullable(),
+    source_mtime: uint.nullable(),
+    source_size: uint.nullable(),
+  })
+  .superRefine((header, context) => {
+    for (const [field, flag] of [
+      ["hash_based", PYTHON_PYC_HASH_BASED],
+      ["check_source", PYTHON_PYC_CHECK_SOURCE],
+    ] as const)
+      if (header[field] !== ((header.flags & flag) !== 0))
+        context.addIssue({
+          code: "custom",
+          path: [field],
+          message: `${field} disagrees with its recorded header flag.`,
+        });
+    if (header.hash_based === (header.source_hash_hex === null))
+      context.addIssue({
+        code: "custom",
+        path: ["source_hash_hex"],
+        message:
+          "Hash-based pycs require an 8-byte source hash; others must omit it.",
+      });
+    if (header.hash_based === (header.source_mtime !== null))
+      context.addIssue({
+        code: "custom",
+        path: ["source_mtime"],
+        message: "Timestamp pycs require mtime; hash-based pycs must omit it.",
+      });
+    if (header.hash_based === (header.source_size !== null))
+      context.addIssue({
+        code: "custom",
+        path: ["source_size"],
+        message:
+          "Timestamp pycs require source_size; hash-based pycs must omit it.",
+      });
+  });
 
 /** Bounded CPython 3.7–3.13 pyc observation; marshal is walked, bytecode is not decompiled. */
 export const pythonPycSchema = z
@@ -253,29 +289,6 @@ export const pythonPycSchema = z
         message: "Marshal byte count disagrees with the header-stripped file.",
       });
     }
-    if (report.header.hash_based === (report.header.source_hash_hex === null)) {
-      context.addIssue({
-        code: "custom",
-        path: ["header", "source_hash_hex"],
-        message:
-          "Hash-based pycs require an 8-byte source hash; others must omit it.",
-      });
-    }
-    if (report.header.hash_based === (report.header.source_mtime !== null)) {
-      context.addIssue({
-        code: "custom",
-        path: ["header", "source_mtime"],
-        message: "Timestamp pycs require mtime; hash-based pycs must omit it.",
-      });
-    }
-    if (report.header.hash_based === (report.header.source_size !== null)) {
-      context.addIssue({
-        code: "custom",
-        path: ["header", "source_size"],
-        message:
-          "Timestamp pycs require source_size; hash-based pycs must omit it.",
-      });
-    }
     if (report.coverage.code_objects !== report.code_objects.length) {
       context.addIssue({
         code: "custom",

src/domain/v8/v8CodeCache.ts

@@ -196,12 +196,6 @@ export const v8CodeCacheSchema = z
         message:
           "Observation coverage disagrees with the listed embedded strings.",
       });
-    if (report.coverage.scanned_bytes !== report.payload.bytes)
-      context.addIssue({
-        code: "custom",
-        path: ["coverage", "scanned_bytes"],
-        message: "Complete coverage must examine every declared payload byte.",
-      });
     report.embedded_strings.forEach((observation, index) => {
       if (observation.retained_bytes > observation.bytes)
         context.addIssue({