Bound adapter assemblies and extraction caches

tests/test_incremental_cache.py

@@ -0,0 +1,102 @@
+from __future__ import annotations
+
+import hashlib
+import os
+import tempfile
+import unittest
+from pathlib import Path
+
+from docforge.errors import DocForgeError
+from docforge.incremental import (
+    CachedSource,
+    ExtractionCache,
+    load_extraction_cache,
+    write_extraction_cache,
+)
+
+
+class ExtractionCacheBoundsTests(unittest.TestCase):
+    @staticmethod
+    def cache(*, payload: str = "ok", count: int = 1) -> ExtractionCache:
+        sources = tuple(
+            CachedSource(
+                source_id=f"source-{index}",
+                source_path=f"src/{index}.txt",
+                fingerprint=hashlib.sha256(str(index).encode()).hexdigest(),
+                extractor_version="fixture@1",
+                dependencies=(),
+                payload={"content": payload},
+            )
+            for index in range(count)
+        )
+        return ExtractionCache("fixture", "fixture-adapter", "1", sources)
+
+    def test_bounded_round_trip_and_oversized_read_is_a_cache_miss(self) -> None:
+        with tempfile.TemporaryDirectory() as directory:
+            path = Path(directory) / "extractions.json"
+            write_extraction_cache(path, self.cache(), max_bytes=1_000, max_sources=2)
+            loaded = load_extraction_cache(
+                path,
+                project_id="fixture",
+                adapter_id="fixture-adapter",
+                adapter_version="1",
+                max_bytes=1_000,
+                max_sources=2,
+            )
+            self.assertEqual(self.cache(), loaded)
+
+            with path.open("ab") as handle:
+                handle.write(b" " * 1_000)
+            self.assertIsNone(
+                load_extraction_cache(
+                    path,
+                    project_id="fixture",
+                    adapter_id="fixture-adapter",
+                    adapter_version="1",
+                    max_bytes=1_000,
+                    max_sources=2,
+                )
+            )
+
+    def test_write_rejects_byte_and_source_limits_without_replacing_target(self) -> None:
+        with tempfile.TemporaryDirectory() as directory:
+            path = Path(directory) / "extractions.json"
+            path.write_bytes(b"preserved\n")
+            for cache, maximum_bytes, maximum_sources in (
+                (self.cache(payload="x" * 500), 100, 2),
+                (self.cache(count=2), 1_000, 1),
+            ):
+                with self.assertRaises(DocForgeError) as captured:
+                    write_extraction_cache(
+                        path,
+                        cache,
+                        max_bytes=maximum_bytes,
+                        max_sources=maximum_sources,
+                    )
+                self.assertEqual("cache_limit", captured.exception.code)
+                self.assertEqual(b"preserved\n", path.read_bytes())
+
+    def test_symlink_and_non_regular_targets_are_cache_misses(self) -> None:
+        with tempfile.TemporaryDirectory() as directory:
+            root = Path(directory)
+            real = root / "real.json"
+            write_extraction_cache(real, self.cache(), max_bytes=1_000, max_sources=2)
+            linked = root / "linked.json"
+            linked.symlink_to(real)
+            fifo = root / "fifo"
+            os.mkfifo(fifo)
+            for path in (linked, fifo):
+                self.assertIsNone(
+                    load_extraction_cache(
+                        path,
+                        project_id="fixture",
+                        adapter_id="fixture-adapter",
+                        adapter_version="1",
+                        max_bytes=1_000,
+                        max_sources=2,
+                    )
+                )
+
+
+if __name__ == "__main__":
+    unittest.main()