chtypes 2.0.0.dev12__py3-none-any.whl → 2.0.0.dev13__py3-none-any.whl

chtypes/_ocifetch/_ensure.py

@@ -75,6 +75,7 @@ from chtypes._ocifetch._layout import (
 from chtypes._ocifetch._lock import Lock, LockPin, load_lock, new_lock, save_lock
 from chtypes._ocifetch._oci import (
     Descriptor,
+    copy_blob_verified,
     fetch_blob_bytes,
     fetch_blob_to_path,
     fetch_manifest_by_digest,
@@ -86,6 +87,8 @@ from chtypes._ocifetch._oci import (
     resolve_platform_manifest,
     spelling_components,
     validate_spelling,
+    verify_body_matches_digest,
+    verify_json_descriptor,
     version_within_request,
 )
 from chtypes._ocifetch._referrers import ReferrerListing, discover_referrers, read_listing
@@ -1346,6 +1349,7 @@ def _verify_and_install_from_local_blobs(
     trusted_keys: tuple[TrustedKey, ...],
     visible_only: bool = False,
     referrers: dict[str, list[tuple[str, dict]]] | None = None,
+    manifest_size: int | None = None,
 ) -> VerifiedRecord | None:
     """docs/guides/fetch-v1.md §1/§6: "A pre-seeded layout has entries in
     `index.json` with no corresponding `unpacked/` directory yet; the first
@@ -1357,7 +1361,14 @@ def _verify_and_install_from_local_blobs(
     `source_root/blobs/sha256/*`, never the network. Returns `None` — never
     raises — for a missing, unreadable or untrusted entry, the same
     tolerant shape `_find_verified_signature` uses for a network referrer:
-    a bad pre-seed is "nothing usable here," not a crash."""
+    a bad pre-seed is "nothing usable here," not a crash.
+
+    The manifest blob is checked against `manifest_digest` (and against
+    `manifest_size`, the index descriptor's size, when the caller has one)
+    before it is parsed, and the layer blob against the manifest's layer
+    descriptor (size and sha256) before it is decompressed: the same
+    checks a download gets (§7.3), so no byte the signature does not cover
+    reaches the install."""
     blobs_dir = source_root / "blobs" / "sha256"
     if not blobs_dir.is_dir():
         return None
@@ -1366,7 +1377,15 @@ def _verify_and_install_from_local_blobs(
     if not manifest_path.is_file():
         return None
     try:
-        manifest_doc = json.loads(manifest_path.read_bytes())
+        manifest_bytes = manifest_path.read_bytes()
+        verify_body_matches_digest(manifest_bytes, manifest_digest, what="pre-seeded manifest")
+        if manifest_size is not None:
+            verify_json_descriptor(
+                manifest_bytes,
+                Descriptor(media_type="", digest=manifest_digest, size=manifest_size),
+                what="pre-seeded manifest",
+            )
+        manifest_doc = json.loads(manifest_bytes)
         layer_desc = manifest_layer_descriptor(manifest_doc)
     except (ValueError, OSError, FetchError):
         return None
@@ -1415,9 +1434,18 @@ def _verify_and_install_from_local_blobs(
     scratch = _scratch_root(cache_write_root)
     tmp_dir = tempfile.mkdtemp(dir=str(scratch))
     try:
+        # Size and sha256 against the layer descriptor BEFORE decompressing
+        # (§7.3), on a copy in scratch, which is then the file unpacked.
+        verified_layer = os.path.join(tmp_dir, "layer")
+        copy_blob_verified(
+            str(layer_path),
+            layer_desc,
+            verified_layer,
+            what=f"pre-seeded layer {layer_desc.digest}",
+        )
         unpacked_dir = os.path.join(tmp_dir, "unpacked")
         unpack_tar_zst(
-            str(layer_path),
+            verified_layer,
             unpacked_dir,
             library_name=predicate["library"],
             library_sha256=predicate["library_sha256"],
@@ -1469,11 +1497,15 @@ def _verify_preseeded_entries(roots: Sequence[Path], trusted_keys: tuple[Trusted
         candidates: list[str] = []
         candidate_layers: set[str] = set()
         manifests_read: dict[str, bytes] = {}
+        sizes: dict[str, int] = {}
         for m in index_doc.get("manifests") or []:
             digest = m.get("digest")
             if not digest or digest in already_verified:
                 continue
             candidates.append(digest)
+            size = m.get("size")
+            if isinstance(size, int) and not isinstance(size, bool):
+                sizes[digest] = size
             try:
                 manifest_hex = parse_digest(digest)
                 manifest_bytes = (blobs_dir / manifest_hex).read_bytes()
@@ -1495,6 +1527,7 @@ def _verify_preseeded_entries(roots: Sequence[Path], trusted_keys: tuple[Trusted
                 trusted_keys=trusted_keys,
                 visible_only=True,
                 referrers=referrers,
+                manifest_size=sizes.get(digest),
             )
             if record is not None:
                 already_verified.add(digest)

chtypes/_ocifetch/_oci.py

@@ -26,6 +26,7 @@ from chtypes._ocifetch._http import (
 
 __all__ = [
     "Descriptor",
+    "copy_blob_verified",
     "fetch_blob_bytes",
     "fetch_blob_to_path",
     "fetch_manifest_by_digest",
@@ -143,6 +144,46 @@ def verify_body_matches_digest(body: bytes, digest: str, *, what: str) -> None:
         )
 
 
+def copy_blob_verified(src_path: str, descriptor: Descriptor, dest_path: str, *, what: str) -> None:
+    """Copy a LOCAL blob file (a pre-seeded layout's `blobs/sha256/<hex>`)
+    to `dest_path`, streaming, and check the copy's size and sha256 against
+    `descriptor` BEFORE the caller does anything else with it — the local
+    counterpart of `fetch_blob_to_path`, with the same §7.3 rule ("check
+    size and sha256 before decompressing"). The caller unpacks the copy it
+    verified, never the source file, so the bytes checked are the bytes
+    used. Reads at most `descriptor.size + 1` bytes; removes `dest_path` on
+    any failure."""
+    expected = parse_digest(descriptor.digest)
+    hasher = hashlib.sha256()
+    size = 0
+    try:
+        with open(src_path, "rb") as src, open(dest_path, "wb") as dst:
+            while size <= descriptor.size:
+                chunk = src.read(min(1 << 20, descriptor.size + 1 - size))
+                if not chunk:
+                    break
+                hasher.update(chunk)
+                dst.write(chunk)
+                size += len(chunk)
+        if size > descriptor.size:
+            raise ArtifactCorruptError(
+                f"{what}: larger than the descriptor's size {descriptor.size}"
+            )
+        if size != descriptor.size:
+            raise ArtifactCorruptError(
+                f"{what}: size {size} disagreed with the descriptor's {descriptor.size}"
+            )
+        actual = hasher.hexdigest()
+        if actual != expected:
+            raise ArtifactCorruptError(f"{what}: sha256 {actual} disagreed with {expected}")
+    except BaseException:
+        try:
+            os.unlink(dest_path)
+        except OSError:
+            pass
+        raise
+
+
 def resolve_platform_manifest(index_doc: dict, platform_key: str) -> Descriptor:
     """Pick the one manifest descriptor in an OCI image index matching
     `platform_key` (e.g. "linux-arm64"), refusing a duplicate or absent

{chtypes-2.0.0.dev12.dist-info → chtypes-2.0.0.dev13.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: chtypes
-Version: 2.0.0.dev12
+Version: 2.0.0.dev13
 Summary: Python binding for the chtypes artifacts: ClickHouse's own type system, per version, behind the frozen chs_* C ABI
 Project-URL: Homepage, https://github.com/wave-rf/chtypes
 Project-URL: Repository, https://github.com/wave-rf/chtypes

{chtypes-2.0.0.dev12.dist-info → chtypes-2.0.0.dev13.dist-info}/RECORD

@@ -22,7 +22,7 @@ chtypes/_ocifetch/__init__.py,sha256=Nwmn1_JqkRSy-1q8182RT5m4HJfbdJgr54it7_TKvBg
 chtypes/_ocifetch/_channel.py,sha256=1MuIj5O8N1d8OdTN4Rx4zf3kc2Nh1hjx4o5BIE6JgmA,20484
 chtypes/_ocifetch/_constants.py,sha256=Gk-h6UkPg5dvNlhkr8z-UY4Y866unUPOSpE8ZwJvoJw,6287
 chtypes/_ocifetch/_dsse.py,sha256=nYkjmRefmVTonH1KUADNNUl0J9sh2C3fg-0WEyURjcc,9838
-chtypes/_ocifetch/_ensure.py,sha256=3d00ZmJY7qjBQAsyEL6g3nikiHnaMm7BzoKTxn-EGts,79685
+chtypes/_ocifetch/_ensure.py,sha256=0v-tGzC4P0oLtADN-YBHzAMWaqOc2bYincxwUAhWtyg,81190
 chtypes/_ocifetch/_errors.py,sha256=SE0zcTUr1fSvSH7HiLGjAInYvD_im8_SyeZNf-VKzXI,7594
 chtypes/_ocifetch/_faults.py,sha256=HpoB9tJW7aLRiJdCFnspzkwhpFATZ6f6hGi49Mbfnhs,7609
 chtypes/_ocifetch/_goldens.py,sha256=8MicgeJSRd4WZkaEsdujRnOa7twQvbT-wfSMPZaD-ms,4191
@@ -30,15 +30,15 @@ chtypes/_ocifetch/_hold.py,sha256=AmXKLFWdPEmZ5fuxKbTKbzYn-Gue9AyMfYTquQ0qvOs,48
 chtypes/_ocifetch/_http.py,sha256=3D8WZXAObw-IiZqLSh9GEys2EKOEE2i9CVfUqjLh3E0,29013
 chtypes/_ocifetch/_layout.py,sha256=ezEoVrsSMhvDhQ4U04hsGdu1-_ISZQ4V6zg4gGIChTQ,21300
 chtypes/_ocifetch/_lock.py,sha256=14_uex9qXdOht-5FSrSvFvd99-gB7DmI3bDuxXs4YU8,5530
-chtypes/_ocifetch/_oci.py,sha256=UdA8QGTh1O27DBYRf5UxsevjgI52IecQlEf9IO6Y0A4,13920
+chtypes/_ocifetch/_oci.py,sha256=Xc0JEShGemdTlu2LviOBFSMTcjwCu9uP6jdG0qlKO80,15677
 chtypes/_ocifetch/_prune.py,sha256=gPToHNkOOVjCWA-lowJHnGpguqOFXklUOQs7PAGfhBQ,10203
 chtypes/_ocifetch/_referrers.py,sha256=SQ8BMVzzA1085azzfAA0Nw3GrHPjeu98smasKyS609A,10530
 chtypes/_ocifetch/_resolve.py,sha256=Qv8xFUdOR-Kb_JqmUeDiQ4_Xv2ETuNFpmoDckMp7eV8,11239
 chtypes/_ocifetch/_retired.py,sha256=mqnI9RA7UK0A5RthCeLlBeVAu5SHbtwKFwxsSZk5e6A,7018
 chtypes/_ocifetch/_revocation.py,sha256=DdKRjFp1eFAxVD7DZHK1bj3_sWVLlaEkVEVx6JyV2Zs,23710
 chtypes/_ocifetch/_unpack.py,sha256=Cn4-up63KL63L8pnkS5zwJ6VbbHPSigpCnmOdcHJ4RI,5609
-chtypes-2.0.0.dev12.dist-info/METADATA,sha256=3hHZvSuwop0hmBO2CrCF4szXxFd26F5plQWz3Jkf5oI,11365
-chtypes-2.0.0.dev12.dist-info/WHEEL,sha256=2iZrJkDIkeDh0k9SikKMqFuRP-ua6mb4Ir1AsER4BEw,87
-chtypes-2.0.0.dev12.dist-info/entry_points.txt,sha256=TdNcluWBR0VOt5mVFROsp-NiVPAdR3EHo0UsHlMC0r0,50
-chtypes-2.0.0.dev12.dist-info/licenses/LICENSE,sha256=z8d0m5b2O9McPEK1xHG_dWgUBT6EfBDz6wA0F7xSPTA,11358
-chtypes-2.0.0.dev12.dist-info/RECORD,,
+chtypes-2.0.0.dev13.dist-info/METADATA,sha256=W3xyv6dfQ9wI53cfWtQWioHnctgRuShzOSDZC5H3Jv4,11365
+chtypes-2.0.0.dev13.dist-info/WHEEL,sha256=2iZrJkDIkeDh0k9SikKMqFuRP-ua6mb4Ir1AsER4BEw,87
+chtypes-2.0.0.dev13.dist-info/entry_points.txt,sha256=TdNcluWBR0VOt5mVFROsp-NiVPAdR3EHo0UsHlMC0r0,50
+chtypes-2.0.0.dev13.dist-info/licenses/LICENSE,sha256=z8d0m5b2O9McPEK1xHG_dWgUBT6EfBDz6wA0F7xSPTA,11358
+chtypes-2.0.0.dev13.dist-info/RECORD,,