pydriller

.mypy.ini

@@ -49,5 +49,8 @@ ignore_missing_imports = True
 [mypy-docx.*]
 ignore_missing_imports = True
 
+[mypy-pydriller.*]
+ignore_missing_imports = True
+
 [mypy-base62.*]
 ignore_missing_imports = True

credsweeper/__main__.py

@@ -1,10 +1,15 @@
+import base64
 import binascii
+import hashlib
+import io
 import logging
 import os
 import sys
 import time
 from argparse import ArgumentParser, ArgumentTypeError, Namespace
-from typing import Any, Union, Optional, Dict
+from typing import Any, Union, Optional, Dict, List, Tuple
+
+from pydriller import Repository
 
 from credsweeper import __version__
 from credsweeper.app import APP_PATH, CredSweeper
@@ -116,6 +121,17 @@ def get_arguments() -> Namespace:
                        const="log.yaml",
                        dest="export_log_config",
                        metavar="PATH")
+    group.add_argument("--git", nargs="+", help="git repo to scan", dest="git", metavar="PATH")
+    parser.add_argument("--commits",
+                        help="scan git repo for N commits only",
+                        type=positive_int,
+                        dest="commits",
+                        default=0,
+                        metavar="POSITIVE_INT")
+    parser.add_argument("--branch",
+                        help="scan git repo for single branch, otherwise - all branches were scanned (slow)",
+                        dest="branch",
+                        type=str)
     parser.add_argument("--rules",
                         help="path of rule config file (default: credsweeper/rules/config.yaml). "
                         f"severity:{[i.value for i in Severity]} "
@@ -320,9 +336,84 @@ def scan(args: Namespace, content_provider: AbstractProvider, json_filename: Opt
         return credsweeper.run(content_provider=content_provider)
     except Exception as exc:
         logger.critical(exc, exc_info=True)
+        logger.exception(exc)
     return -1
 
 
+def scan_git(args: Namespace) -> Tuple[int, int, int]:
+    """Scan repository for branches and commits
+    Returns:
+        total credentials found
+        total scanned branches
+        total scanned commits
+    """
+    total_credentials = 0
+    total_branches = 0
+    total_commits = 0
+    try:
+        sha1git = hashlib.sha1(str(args.git).encode()).digest()
+        repo_hash = base64.b32encode(sha1git).decode("ascii")
+        journal_filename = f"{repo_hash}.json"
+        logger.info(f"{args.git} sha1 in base32 {repo_hash}")
+        repo_journal = Util.json_load(journal_filename)
+        if not isinstance(repo_journal, dict):
+            with open(journal_filename, "w") as f:
+                f.write("{}")
+            repo_journal = {"repo": args.git}
+        credsweeper = CredSweeper(rule_path=args.rule_path,
+                                  config_path=args.config_path,
+                                  api_validation=args.api_validation,
+                                  sort_output=args.sort_output,
+                                  use_filters=args.no_filters,
+                                  pool_count=args.jobs,
+                                  ml_batch_size=args.ml_batch_size,
+                                  ml_threshold=args.ml_threshold,
+                                  ml_providers=args.ml_providers,
+                                  find_by_ext=args.find_by_ext,
+                                  depth=args.depth,
+                                  doc=args.doc,
+                                  severity=args.severity,
+                                  size_limit=args.size_limit,
+                                  log_level=args.log)
+        repository = Repository(args.git, only_in_branch=args.branch)
+        for commit in repository.traverse_commits():
+            if commit.hash in repo_journal:
+                logger.debug(f"Skip already scanned commit: {commit.hash}")
+                continue
+            logger.info(f"Scan commit: {commit.hash}")
+            paths: List[Tuple[str, io.BytesIO]] = []
+            for file in commit.modified_files:
+                logger.info(f"FILE: {file.old_path} -> {file.new_path}")
+                try:
+                    if file.new_path is not None:
+                        _io = io.BytesIO(file.content)
+                        paths.append((file.filename, _io))
+                except ValueError as exc:
+                    logger.error("Possible missed submodule:%s", str(exc))
+            provider = FilesProvider(paths)
+            if args.json_filename:
+                ext = Util.get_extension(args.json_filename, False)
+                credsweeper.json_filename = f"{args.json_filename[:-len(ext)]}.{commit.hash}{ext}"
+            if args.xlsx_filename:
+                ext = Util.get_extension(args.xlsx_filename, False)
+                credsweeper.xlsx_filename = f"{args.xlsx_filename[:-len(ext)]}.{commit.hash}{ext}"
+
+            commit_cred_number = credsweeper.run(provider)
+            if credsweeper.is_ml_validator_inited:
+                # reset not-pickled object for multiprocess
+                credsweeper.ml_validator = None
+            credsweeper.credential_manager.candidates.clear()
+            total_credentials += commit_cred_number
+            total_commits += 1
+            repo_journal[commit.hash] = commit_cred_number
+            Util.json_dump(repo_journal, journal_filename)
+        total_branches += 1
+    except Exception as exc:
+        logger.critical(exc, exc_info=True)
+        return -1, total_branches, total_commits
+    return total_credentials, total_branches, total_commits
+
+
 def main() -> int:
     """Main function"""
     result = EXIT_FAILURE
@@ -331,7 +422,7 @@ def main() -> int:
     if args.banner:
         print(f"CredSweeper {__version__} crc32:{check_integrity():08x}")
     Logger.init_logging(args.log, args.log_config_path)
-    logger.info(f"Init CredSweeper object with arguments: {args}")
+    logger.info(f"Init CredSweeper object with arguments: {args} CWD: {os.getcwd()}")
     summary: Dict[str, int] = {}
     if args.path:
         logger.info(f"Run analyzer on path: {args.path}")
@@ -354,6 +445,13 @@ def main() -> int:
         summary["Deleted File Credentials"] = del_credentials_number
         if 0 <= add_credentials_number and 0 <= del_credentials_number:
             result = EXIT_SUCCESS
+    elif args.git:
+        logger.info(f"Run analyzer on GIT: {args.git}")
+        credentials_number, branches_number, commits_number = scan_git(args)
+        summary[
+            f"Detected Credentials in {branches_number} branches and {commits_number} commits "] = credentials_number
+        if 0 <= credentials_number:
+            result = EXIT_SUCCESS
     elif args.export_config:
         logging.info(f"Exporting default config to file: {args.export_config}")
         config_dict = Util.json_load(APP_PATH / "secret" / "config.json")

credsweeper/app.py

@@ -47,7 +47,7 @@ def __init__(self,
                  sort_output: bool = False,
                  use_filters: bool = True,
                  pool_count: int = 1,
-                 ml_batch_size: Optional[int] = None,
+                 ml_batch_size: Optional[int] = 16,
                  ml_threshold: Union[float, ThresholdPreset] = ThresholdPreset.medium,
                  ml_config: Union[None, str, Path] = None,
                  ml_model: Union[None, str, Path] = None,
@@ -187,11 +187,18 @@ def _use_ml_validation(self) -> bool:
 
     # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
 
+    @property
+    def is_ml_validator_inited(self) -> bool:
+        """method to check whether ml_validator was inited without creation"""
+        return bool(self.__ml_validator)
+
+    # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
+
     @property
     def ml_validator(self) -> MlValidator:
         """ml_validator getter"""
         from credsweeper.ml_model import MlValidator
-        if not self.__ml_validator:
+        if not self.is_ml_validator_inited:
             self.__ml_validator: MlValidator = MlValidator(
                 threshold=self.ml_threshold,  #
                 ml_config=self.ml_config,  #
@@ -381,6 +388,7 @@ def post_processing(self) -> None:
             if ml_cred_groups:
                 logger.info(f"Run ML Validation for {len(ml_cred_groups)} groups")
                 is_cred, probability = self.ml_validator.validate_groups(ml_cred_groups, self.ml_batch_size)
+                logger.info(f"DONE ML Validation for {len(is_cred)} results")
                 for i, (_, group_candidates) in enumerate(ml_cred_groups):
                     for candidate in group_candidates:
                         if candidate.use_ml:
@@ -404,6 +412,12 @@ def export_results(self) -> None:
 
         credentials = self.credential_manager.get_credentials()
 
+        if credentials:
+            logger.info(f"Exporting {len(credentials)} credentials")
+        else:
+            logger.info("No credentials were found")
+            return
+
         if self.sort_output:
             credentials.sort(key=lambda x: (  #
                 x.line_data_list[0].path,  #

credsweeper/ml_model/ml_validator.py

@@ -11,6 +11,7 @@
 from credsweeper.credentials import Candidate, CandidateKey
 import credsweeper.ml_model.features as features
 from credsweeper.utils import Util
+import psutil
 
 logger = logging.getLogger(__name__)
 
@@ -238,6 +239,9 @@ def validate_groups(self, group_list: List[Tuple[CandidateKey, List[Candidate]]]
                 variable_input_list.clear()
                 value_input_list.clear()
                 features_list.clear()
+            elif 0 == tail % 4391:
+                process = psutil.Process()
+                logger.info(f"ML Validation tail {tail} meminfo:{process.memory_info()}")
         if head != tail:
             probability[head:tail] = self._batch_call_model(line_input_list, variable_input_list, value_input_list,
                                                             features_list)

docs/source/guide.rst

@@ -13,7 +13,9 @@ Get all argument list:
 
 .. code-block:: text
 
-    usage: python -m credsweeper [-h] (--path PATH [PATH ...] | --diff_path PATH [PATH ...] | --export_config [PATH] | --export_log_config [PATH])
+    usage: python -m credsweeper [-h]
+                             (--path PATH [PATH ...] | --diff_path PATH [PATH ...] | --export_config [PATH] | --export_log_config [PATH] | --git PATH [PATH ...])
+                             [--commits POSITIVE_INT] [--branch BRANCH]
                              [--rules PATH] [--severity SEVERITY] [--config PATH] [--log_config PATH] [--denylist PATH]
                              [--find-by-ext] [--depth POSITIVE_INT] [--no-filters] [--doc] [--ml_threshold FLOAT_OR_STR]
                              [--ml_batch_size POSITIVE_INT] [--ml_config PATH] [--ml_model PATH] [--ml_providers STR]
@@ -31,6 +33,11 @@ Get all argument list:
                             exporting default config to file (default: config.json)
       --export_log_config [PATH]
                             exporting default logger config to file (default: log.yaml)
+      --git PATH [PATH ...]
+                            git repo to scan
+      --commits POSITIVE_INT
+                            scan git repo for N commits only
+      --branch BRANCH       scan git repo for single branch, otherwise - all branches were scanned (slow)
       --rules PATH          path of rule config file (default: credsweeper/rules/config.yaml). severity:['critical', 'high', 'medium', 'low', 'info'] type:['keyword', 'pattern', 'pem_key', 'multi']
       --severity SEVERITY   set minimum level for rules to apply ['critical', 'high', 'medium', 'low', 'info'](default: 'Severity.INFO', case insensitive)
       --config PATH         use custom config (default: built-in)

requirements.txt

@@ -18,6 +18,7 @@ pdfminer.six==20231228
 password-strength==0.0.3.post2
 python-dateutil==2.8.2
 pyjks==20.0.0
+PyDriller==2.6
 pybase62==1.0.0
 base58==2.1.1