docker-assemble 0.2.2__tar.gz → 0.3.0__tar.gz

{docker_assemble-0.2.2 → docker_assemble-0.3.0}/.gitignore

@@ -10,4 +10,5 @@ dist/
 build/
 venv/
 .env
+.idea
 

{docker_assemble-0.2.2/docker_assemble.egg-info → docker_assemble-0.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docker-assemble
-Version: 0.2.2
+Version: 0.3.0
 Summary: A CLI tool to extract and analyze Docker images
 Author: Sina
 License: Apache-2.0

docker_assemble-0.3.0/docker_assemble/image_exporter.py

@@ -0,0 +1,169 @@
+import docker
+import tarfile
+import tempfile
+import os
+from pathlib import Path
+import logging
+import io
+
+def extract_image(image_name: str, output_dir: str):
+    client = docker.from_env()
+
+    try:
+        image = client.images.get(image_name)
+        logging.info(f"Image '{image_name}' found locally.")
+    except docker.errors.ImageNotFound:
+        logging.info(f"Image '{image_name}' not found locally. Pulling...")
+        image = client.images.pull(image_name)
+
+    container = client.containers.run(image=image_name, command="sleep infinity", detach=True)
+    logging.debug(f"Created temporary container: {container.id[:12]}")
+
+    try:
+        stream, _ = container.get_archive("/")
+        tmp_tar_path = tempfile.mktemp(suffix=".tar")
+        with open(tmp_tar_path, "wb") as f:
+            for chunk in stream:
+                f.write(chunk)
+
+        logging.debug(f"Filesystem archive saved to: {tmp_tar_path}")
+
+        output_path = Path(output_dir).resolve()
+        output_path.mkdir(parents=True, exist_ok=True)
+
+        extract_tar_safely(tmp_tar_path, output_path)
+
+        logging.info(f"Image filesystem extracted to: {output_path}")
+
+    finally:
+        container.remove(force=True)
+        if os.path.exists(tmp_tar_path):
+            os.remove(tmp_tar_path)
+        logging.debug("Cleaned up temporary container and tar file.")
+
+
+def extract_tar_safely(tar_path: str, output_path: Path):
+    # def is_safe_path(base: Path, target: Path) -> bool:
+    #     try:
+    #         return target.resolve().is_relative_to(base.resolve())
+    #     except AttributeError:
+    #         # For Python < 3.9 fallback
+    #         return str(target.resolve()).startswith(str(base.resolve()))
+
+    with tarfile.open(tar_path, "r") as tar:
+        for member in tar.getmembers():
+            member.name = member.name.lstrip("/")
+            member_path = output_path / member.name
+
+            # if not is_safe_path(output_path, member_path):
+            #     logging.warning(f"Blocked unsafe path: {member.name}, output_path: {output_path}, member_path: {member_path}")
+            #     continue
+
+            tar.extract(member, path=output_path)
+            logging.debug(f"Extracted: {member.name}")
+
+    logging.info(f"Extraction completed to: {output_path}")
+
+def check_large_files(output_dir, max_size_bytes):
+    logging.info(f"Checking for files larger than {max_size_bytes} bytes.")
+    large_files = []
+    for root, _, files in os.walk(output_dir):
+        for file in files:
+            file_path = Path(root) / file
+            try:
+                file_size = os.path.getsize(file_path)
+                if file_size > max_size_bytes:
+                    large_files.append((file_path, file_size))
+            except FileNotFoundError:
+                logging.error(f"File not found: {file_path}")
+            except OSError as e:
+                logging.error(f"OS error while getting size of {file_path}: {e}")
+
+    if large_files:
+        logging.warning("The following files exceed the maximum file size:")
+        for path, size in large_files:
+            logging.warning(f"- {path}: {size} bytes")
+    else:
+        logging.info("No files exceed the maximum file size.")
+
+    return large_files
+
+def remove_files(large_files):
+    while True:
+        indices_str = input("Enter the indices of files to remove (comma-separated, or 'no' to skip): ")
+        if indices_str.lower() == 'no':
+            logging.info("No files will be removed.")
+            break
+
+        try:
+            indices = [int(i) for i in indices_str.split(',')]
+            files_to_remove = [large_files[i][0] for i in indices]
+
+            print("Files to be removed:")
+            for file in files_to_remove:
+                print(file)
+
+            confirmation = input("Are you sure you want to delete these files? (yes/no): ")
+            if confirmation.lower() == 'yes':
+                for file in files_to_remove:
+                    os.remove(file)
+                    logging.info(f"Removed file: {file}")
+                break
+            else:
+                print("Removal cancelled.")
+        except (ValueError, IndexError) as e:
+            print(f"Invalid input: {e}")
+
+def create_new_image(output_dir, new_image_name):
+    client = docker.from_env()
+    logging.info(f"Creating new image {new_image_name} from directory: {output_dir}")
+
+    # Create a temporary Dockerfile
+    dockerfile_content = f"""
+    FROM scratch
+    COPY . /
+    """
+    with tempfile.TemporaryDirectory() as build_context:
+        dockerfile_path = Path(build_context) / 'Dockerfile'
+        with open(dockerfile_path, 'w') as f:
+            f.write(dockerfile_content)
+
+        # Create a tar archive of the output directory
+        def generate_tar(directory):
+            tar_buffer = io.BytesIO()
+            with tarfile.open(fileobj=tar_buffer, mode='w:gz') as tar:
+                tar.add(dockerfile_path, arcname='Dockerfile')
+
+                for root, _, files in os.walk(directory):
+                    for file in files:
+                        file_path = os.path.join(root, file)
+                        rel_path = os.path.relpath(file_path, directory)
+                        try:
+                            tar.add(file_path, arcname=rel_path)
+                        except FileNotFoundError:
+                            logging.error(f"File not found while creating tar: {file_path}")
+                            continue
+                        except Exception as e:
+                            logging.error(f"Error adding {file_path} to tar: {e}")
+                            continue
+
+            tar_buffer.seek(0)
+            return tar_buffer.getvalue()
+
+        tar_stream = generate_tar(output_dir)
+
+        try:
+            with open(dockerfile_path, 'rb') as df:
+                response = client.images.build(
+                    fileobj=io.BytesIO(tar_stream),
+                    tag=new_image_name,
+                    custom_context=True,
+                    rm=True
+                )
+                for line in response:
+                    logging.info(line)
+                logging.info(f"New image created: {new_image_name}")
+
+        except docker.errors.BuildError as e:
+            logging.error(f"Failed to build image: {e}")
+            raise

docker_assemble-0.3.0/docker_assemble/main.py

@@ -0,0 +1,40 @@
+import argparse
+import logging
+import docker_assemble.image_exporter as image_exporter
+
+
+def parse_size(size_str):
+    suffixes = {'K': 1024, 'M': 1024**2, 'G': 1024**3}
+    size_str = size_str.upper()
+    if size_str[-1] in suffixes:
+        num = size_str[:-1]
+        suffix = size_str[-1]
+        return int(float(num) * suffixes[suffix])
+    else:
+        return int(size_str)
+
+def run():
+    parser = argparse.ArgumentParser(description="Docker Assemble CLI")
+    parser.add_argument("-d", action="store_true", help="Disassemble an image")
+    parser.add_argument("--debug", action="store_true", help="Enable debug mode")
+    parser.add_argument("--maximum-file-size", type=str, help="Maximum file size (e.g., 1G, 100M, 10K). Files larger than this size will be listed.")
+    parser.add_argument("--new-image-name", type=str, help="Name for the new Docker image after removing files.")
+    parser.add_argument("image", help="Docker image name")
+    parser.add_argument("output_dir", nargs="?", default=".", help="Optional output directory")
+
+    args = parser.parse_args()
+
+    logging.basicConfig(level=logging.DEBUG if args.debug else logging.INFO)
+
+    logging.debug(f"Extracting image: {args.image} to directory: {args.output_dir}")
+    image_exporter.extract_image(image_name=args.image, output_dir=args.output_dir)
+
+    if args.maximum_file_size:
+        max_size_bytes = parse_size(args.maximum_file_size)
+        large_files = image_exporter.check_large_files(args.output_dir, max_size_bytes)
+
+        if large_files:
+            image_exporter.remove_files(large_files)
+
+        if args.new_image_name:
+            image_exporter.create_new_image(args.output_dir, args.new_image_name)

{docker_assemble-0.2.2 → docker_assemble-0.3.0/docker_assemble.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docker-assemble
-Version: 0.2.2
+Version: 0.3.0
 Summary: A CLI tool to extract and analyze Docker images
 Author: Sina
 License: Apache-2.0

docker_assemble-0.2.2/docker_assemble/image_exporter.py

@@ -1,68 +0,0 @@
-import docker
-import tarfile
-import tempfile
-import os
-import shutil
-from pathlib import Path
-import logging
-
-def extract_image(image_name: str, output_dir: str):
-    client = docker.from_env()
-
-    try:
-        image = client.images.get(image_name)
-        logging.info(f"Image '{image_name}' found locally.")
-    except docker.errors.ImageNotFound:
-        logging.info(f"Image '{image_name}' not found locally. Pulling...")
-        image = client.images.pull(image_name)
-
-    container = client.containers.run(image=image_name, command="sleep infinity", detach=True)
-    logging.debug(f"Created temporary container: {container.id[:12]}")
-
-    try:
-        stream, _ = container.get_archive("/")
-        tmp_tar_path = tempfile.mktemp(suffix=".tar")
-        with open(tmp_tar_path, "wb") as f:
-            for chunk in stream:
-                f.write(chunk)
-
-        logging.debug(f"Filesystem archive saved to: {tmp_tar_path}")
-
-        output_path = Path(output_dir).resolve()
-        output_path.mkdir(parents=True, exist_ok=True)
-
-        extract_tar_safely(tmp_tar_path, output_path)
-
-        logging.info(f"Image filesystem extracted to: {output_path}")
-
-    finally:
-        container.remove(force=True)
-        if os.path.exists(tmp_tar_path):
-            os.remove(tmp_tar_path)
-        logging.debug("Cleaned up temporary container and tar file.")
-
-
-def extract_tar_safely(tar_path: str, output_path: Path):
-    # def is_safe_path(base: Path, target: Path) -> bool:
-    #     try:
-    #         return target.resolve().is_relative_to(base.resolve())
-    #     except AttributeError:
-    #         # For Python < 3.9 fallback
-    #         return str(target.resolve()).startswith(str(base.resolve()))
-
-    with tarfile.open(tar_path, "r") as tar:
-        for member in tar.getmembers():
-            member.name = member.name.lstrip("/")
-            member_path = output_path / member.name
-
-            # if not is_safe_path(output_path, member_path):
-            #     logging.warning(f"Blocked unsafe path: {member.name}, output_path: {output_path}, member_path: {member_path}")
-            #     continue
-
-            tar.extract(member, path=output_path)
-            logging.debug(f"Extracted: {member.name}")
-
-    logging.info(f"Extraction completed to: {output_path}")
-
-
-

docker_assemble-0.2.2/docker_assemble/main.py

@@ -1,18 +0,0 @@
-import argparse
-import logging
-import os
-from docker_assemble.image_exporter import extract_image
-
-def run():
-    parser = argparse.ArgumentParser(description="Docker Assemble CLI")
-    parser.add_argument("-d", action="store_true", help="Disassemble an image")
-    parser.add_argument("--debug", action="store_true", help="Enable debug mode")
-    parser.add_argument("image", help="Docker image name")
-    parser.add_argument("output_dir", nargs="?", default=".", help="Optional output directory")
-
-    args = parser.parse_args()
-
-    logging.basicConfig(level=logging.DEBUG if args.debug else logging.INFO)
-
-    logging.debug(f"Extracting image: {args.image} to directory: {args.output_dir}")
-    extract_image(image_name=args.image, output_dir=args.output_dir)