skip file from being processed based on their magic

unblob/cli.py

@@ -16,6 +16,7 @@
 from .processing import (
     DEFAULT_DEPTH,
     DEFAULT_PROCESS_NUM,
+    DEFAULT_SKIP_MAGIC,
     ExtractionConfig,
     process_file,
 )
@@ -116,6 +117,16 @@ def __init__(
     help="Load plugins from the provided path.",
     show_default=True,
 )
+@click.option(
+    "-S",
+    "--skip-magic",
+    "skip_magic",
+    type=click.STRING,
+    default=DEFAULT_SKIP_MAGIC,
+    help="Skip processing files with given magic prefix",
+    show_default=True,
+    multiple=True,
+)
 @click.option(
     "-p",
     "--process-num",
@@ -146,6 +157,7 @@ def cli(
     extract_root: Path,
     depth: int,
     entropy_depth: int,
+    skip_magic: List[str],
     process_num: int,
     keep_extracted_chunks: bool,
     verbose: int,
@@ -164,6 +176,7 @@ def cli(
         max_depth=depth,
         entropy_depth=entropy_depth,
         entropy_plot=bool(verbose >= 3),
+        skip_magic=skip_magic,
         process_num=process_num,
         handlers=handlers,
         keep_extracted_chunks=keep_extracted_chunks,

unblob/processing.py

@@ -6,6 +6,7 @@
 from typing import List
 
 import attr
+import magic
 import plotext as plt
 from structlog import get_logger
 
@@ -32,6 +33,7 @@
 
 DEFAULT_DEPTH = 10
 DEFAULT_PROCESS_NUM = multiprocessing.cpu_count()
+DEFAULT_SKIP_MAGIC = ["ELF", "JPEG", "GIF", "PNG"]
 
 
 @attr.define(kw_only=True)
@@ -40,6 +42,7 @@ class ExtractionConfig:
     entropy_depth: int
     entropy_plot: bool = False
     max_depth: int = DEFAULT_DEPTH
+    skip_magic: List[str] = DEFAULT_SKIP_MAGIC
     process_num: int = DEFAULT_PROCESS_NUM
     keep_extracted_chunks: bool = False
     handlers: Handlers = BUILTIN_HANDLERS
@@ -146,6 +149,18 @@ def __init__(
     def process(self):
         logger.debug("Processing file", path=self.task.path, size=self.size)
 
+        detect = magic.detect_from_filename(self.task.path)
+        skip = any(
+            [detect.name.startswith(pattern) for pattern in self.config.skip_magic]
+        )
+        if skip:
+            logger.debug(
+                "Skip file from processing based on file-magic",
+                magic=detect.name,
+                path=self.task.path,
+            )
+            return
+
         with self.task.path.open("rb") as file:
             all_chunks = search_chunks_by_priority(
                 self.task.path, file, self.size, self.config.handlers, self.result