model-research 0.1.0__py3-none-any.whl

model_research/__init__.py

@@ -0,0 +1,144 @@
+from __future__ import annotations
+
+import argparse
+import hashlib
+import importlib.resources as resources
+import json
+import os
+import tempfile
+from pathlib import Path
+
+import zstandard as zstd
+
+
+_MANIFEST = json.loads(resources.files(__package__).joinpath("asset_manifest.json").read_text(encoding="utf-8"))
+_SHARDS = [
+    ("model_research_part01", "part-0001.bin"),
+    ("model_research_part02", "part-0002.bin"),
+    ("model_research_part03", "part-0003.bin"),
+    ("model_research_part04", "part-0004.bin"),
+    ("model_research_part05", "part-0005.bin"),
+    ("model_research_part06", "part-0006.bin"),
+    ("model_research_part07", "part-0007.bin")
+]
+
+
+def _default_target_dir() -> Path:
+    if os.name == "nt":
+        return Path.cwd()
+    return Path("/tmp")
+
+
+def _file_digest(path: Path, chunk_size: int = 1024 * 1024) -> tuple[str, int]:
+    digest = hashlib.sha256()
+    size = 0
+    with path.open("rb") as handle:
+        while True:
+            chunk = handle.read(chunk_size)
+            if not chunk:
+                break
+            digest.update(chunk)
+            size += len(chunk)
+    return digest.hexdigest(), size
+
+
+def _target_path(target_dir: str | os.PathLike[str] | None) -> Path:
+    base_dir = _default_target_dir() if target_dir is None else Path(target_dir)
+    base_dir.mkdir(parents=True, exist_ok=True)
+    return base_dir.resolve() / _MANIFEST["original_filename"]
+
+
+def ensure_asset(target_dir: str | os.PathLike[str] | None = None, force: bool = False) -> str:
+    destination = _target_path(target_dir)
+    if destination.exists() and not force:
+        current_hash, current_size = _file_digest(destination)
+        if current_hash == _MANIFEST["original_sha256"] and current_size == _MANIFEST["original_size"]:
+            return str(destination)
+
+    temp_handle = tempfile.NamedTemporaryFile(
+        mode="wb",
+        prefix=f"{destination.name}.",
+        suffix=".tmp",
+        delete=False,
+        dir=str(destination.parent),
+    )
+    tmp_path = Path(temp_handle.name)
+    original_hash = hashlib.sha256()
+    compressed_hash = hashlib.sha256()
+    written_size = 0
+    try:
+        with temp_handle:
+            decompressor = zstd.ZstdDecompressor().decompressobj()
+            for module_name, part_filename in _SHARDS:
+                resource = resources.files(f"{module_name}._data").joinpath(part_filename)
+                with resource.open("rb") as shard_handle:
+                    while True:
+                        chunk = shard_handle.read(1024 * 1024)
+                        if not chunk:
+                            break
+                        compressed_hash.update(chunk)
+                        output = decompressor.decompress(chunk)
+                        if output:
+                            original_hash.update(output)
+                            written_size += len(output)
+                            temp_handle.write(output)
+            tail = decompressor.flush()
+            if tail:
+                original_hash.update(tail)
+                written_size += len(tail)
+                temp_handle.write(tail)
+
+        if compressed_hash.hexdigest() != _MANIFEST["compressed_sha256"]:
+            raise RuntimeError("Compressed shard digest mismatch.")
+        if original_hash.hexdigest() != _MANIFEST["original_sha256"]:
+            raise RuntimeError("Restored asset digest mismatch.")
+        if written_size != _MANIFEST["original_size"]:
+            raise RuntimeError("Restored asset size mismatch.")
+
+        os.replace(tmp_path, destination)
+        return str(destination)
+    except Exception:
+        if tmp_path.exists():
+            tmp_path.unlink()
+        raise
+
+
+
+def unzip_asset(
+    asset_path: str | os.PathLike[str] | None = None,
+    extract_path: str | os.PathLike[str] = "./",
+    force: bool = False,
+) -> str:
+    archive_path = Path(ensure_asset(".", force=force)) if asset_path is None else Path(asset_path).expanduser()
+    if not archive_path.exists():
+        raise FileNotFoundError(f"Archive asset does not exist: {archive_path}")
+
+    destination = Path(extract_path).expanduser()
+    destination.mkdir(parents=True, exist_ok=True)
+    archive_format = _MANIFEST.get("archive_format")
+    if archive_format == "zip":
+        import zipfile
+
+        with zipfile.ZipFile(archive_path, "r") as archive:
+            archive.extractall(destination)
+    elif archive_format == "7z":
+        try:
+            import py7zr
+        except ModuleNotFoundError as exc:
+            raise RuntimeError("py7zr is required to extract .7z assets.") from exc
+
+        with py7zr.SevenZipFile(archive_path, mode="r") as archive:
+            archive.extractall(path=str(destination))
+    else:
+        raise RuntimeError(f"Unsupported archive format: {archive_format}")
+    return str(destination.resolve())
+
+
+
+def main(argv: list[str] | None = None) -> int:
+    parser = argparse.ArgumentParser(description="Restore the packaged asset into a target directory.")
+    parser.add_argument("--target-dir", default=None, help="Directory where the restored asset will be written.")
+    parser.add_argument("--force", action="store_true", help="Force regeneration even if the target already matches.")
+    args = parser.parse_args(argv)
+    print(ensure_asset(args.target_dir, force=args.force))
+    return 0

model_research/__main__.py

@@ -0,0 +1,5 @@
+from . import main
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

model_research/asset_manifest.json

@@ -0,0 +1,12 @@
+{
+  "archive_format": "zip",
+  "chunk_bytes": 100663296,
+  "compressed_sha256": "ee2a577841c84a04e884aea1d4c325598888e6ea9eb0c01b82e1bf909b5e52b5",
+  "compression_level": 22,
+  "is_zip_asset": true,
+  "original_filename": "cancer_generation.zip",
+  "original_sha256": "64758df881102dfb6f1dc4d1a3f6fe5c7c9cd946ce0057a926d9d8dd1a18d262",
+  "original_size": 633003159,
+  "package_version": "0.1.0",
+  "shard_count": 7
+}

model_research-0.1.0.dist-info/METADATA

@@ -0,0 +1,12 @@
+Metadata-Version: 2.4
+Name: model-research
+Version: 0.1.0
+Summary: Loader package for the model-research asset.
+Requires-Dist: model-research-part01==0.1.0
+Requires-Dist: model-research-part02==0.1.0
+Requires-Dist: model-research-part03==0.1.0
+Requires-Dist: model-research-part04==0.1.0
+Requires-Dist: model-research-part05==0.1.0
+Requires-Dist: model-research-part06==0.1.0
+Requires-Dist: model-research-part07==0.1.0
+Requires-Dist: zstandard>=0.23.0

model_research-0.1.0.dist-info/RECORD

@@ -0,0 +1,7 @@
+model_research/__init__.py,sha256=BnP7D9CUdJnsriF1FWV8_rCAyAj6az3Q9Xw39PJdQPg,5456
+model_research/__main__.py,sha256=7L9asSJNpo-bjIqFDRk3qkXtybFElw3oEE1WPxUuzSg,82
+model_research/asset_manifest.json,sha256=PYgMI533r68g_3-DeCERDpU6oFKqEJZeTLbMfr5qM6Q,429
+model_research-0.1.0.dist-info/METADATA,sha256=odLeTpY6okUlj3OlHsmfpJg-1i6mhNfs8Km2vd6DDYs,465
+model_research-0.1.0.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
+model_research-0.1.0.dist-info/top_level.txt,sha256=WZEX-4jMFPPBtgJmR1Jeb-1nojcmqwQ7B9lOe9csC_M,15
+model_research-0.1.0.dist-info/RECORD,,

model_research-0.1.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (82.0.1)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

model_research-0.1.0.dist-info/top_level.txt

@@ -0,0 +1 @@
+model_research