CM-30526 - Make Git executable optional

cycode/cli/commands/scan/code_scanner.py

@@ -7,7 +7,6 @@
 from uuid import UUID, uuid4
 
 import click
-from git import NULL_TREE, Repo
 
 from cycode.cli import consts
 from cycode.cli.config import configuration_manager
@@ -28,9 +27,8 @@
 from cycode.cli.models import CliError, Document, DocumentDetections, LocalScanResult, Severity
 from cycode.cli.printers import ConsolePrinter
 from cycode.cli.utils import scan_utils
-from cycode.cli.utils.path_utils import (
-    get_path_by_os,
-)
+from cycode.cli.utils.git_proxy import git_proxy
+from cycode.cli.utils.path_utils import get_path_by_os
 from cycode.cli.utils.progress_bar import ScanProgressBarSection
 from cycode.cli.utils.scan_batch import run_parallel_batched_scan
 from cycode.cli.utils.scan_utils import set_issue_detected
@@ -244,7 +242,7 @@ def scan_commit_range(
     documents_to_scan = []
     commit_ids_to_scan = []
 
-    repo = Repo(path)
+    repo = git_proxy.get_repo(path)
     total_commits_count = int(repo.git.rev_list('--count', commit_range))
     logger.debug(f'Calculating diffs for {total_commits_count} commits in the commit range {commit_range}')
 
@@ -261,7 +259,7 @@ def scan_commit_range(
 
         commit_id = commit.hexsha
         commit_ids_to_scan.append(commit_id)
-        parent = commit.parents[0] if commit.parents else NULL_TREE
+        parent = commit.parents[0] if commit.parents else git_proxy.get_null_tree()
         diff = commit.diff(parent, create_patch=True, R=True)
         commit_documents_to_scan = []
         for blob in diff:
@@ -658,7 +656,7 @@ def get_scan_parameters(context: click.Context, paths: Tuple[str]) -> dict:
 
 def try_get_git_remote_url(path: str) -> Optional[str]:
     try:
-        remote_url = Repo(path).remotes[0].config_reader.get('url')
+        remote_url = git_proxy.get_repo(path).remotes[0].config_reader.get('url')
         logger.debug(f'Found Git remote URL "{remote_url}" in path "{path}"')
         return remote_url
     except Exception as e:

cycode/cli/commands/scan/pre_commit/pre_commit_command.py

@@ -2,7 +2,6 @@
 from typing import List
 
 import click
-from git import Repo
 
 from cycode.cli import consts
 from cycode.cli.commands.scan.code_scanner import scan_documents, scan_sca_pre_commit
@@ -12,6 +11,7 @@
     get_diff_file_path,
 )
 from cycode.cli.models import Document
+from cycode.cli.utils.git_proxy import git_proxy
 from cycode.cli.utils.path_utils import (
     get_path_by_os,
 )
@@ -31,7 +31,7 @@ def pre_commit_command(context: click.Context, ignored_args: List[str]) -> None:
         scan_sca_pre_commit(context)
         return
 
-    diff_files = Repo(os.getcwd()).index.diff('HEAD', create_patch=True, R=True)
+    diff_files = git_proxy.get_repo(os.getcwd()).index.diff('HEAD', create_patch=True, R=True)
 
     progress_bar.set_section_length(ScanProgressBarSection.PREPARE_LOCAL_FILES, len(diff_files))
 

cycode/cli/exceptions/handle_scan_errors.py

@@ -1,19 +1,19 @@
 from typing import Optional
 
 import click
-from git import InvalidGitRepositoryError
 
 from cycode.cli.exceptions import custom_exceptions
 from cycode.cli.models import CliError, CliErrors
 from cycode.cli.printers import ConsolePrinter
+from cycode.cli.utils.git_proxy import git_proxy
 
 
 def handle_scan_exception(
     context: click.Context, e: Exception, *, return_exception: bool = False
 ) -> Optional[CliError]:
     context.obj['did_fail'] = True
 
-    ConsolePrinter(context).print_exception()
+    ConsolePrinter(context).print_exception(e)
 
     errors: CliErrors = {
         custom_exceptions.NetworkError: CliError(
@@ -49,7 +49,7 @@ def handle_scan_exception(
             'Please make sure that your file is well formed '
             'and execute the scan again',
         ),
-        InvalidGitRepositoryError: CliError(
+        git_proxy.get_invalid_git_repository_error(): CliError(
             soft_fail=False,
             code='invalid_git_error',
             message='The path you supplied does not correlate to a git repository. '
@@ -69,10 +69,13 @@ def handle_scan_exception(
         ConsolePrinter(context).print_error(error)
         return None
 
+    unknown_error = CliError(code='unknown_error', message=str(e))
+
     if return_exception:
-        return CliError(code='unknown_error', message=str(e))
+        return unknown_error
 
     if isinstance(e, click.ClickException):
         raise e
 
-    raise click.ClickException(str(e))
+    ConsolePrinter(context).print_error(unknown_error)
+    exit(1)

cycode/cli/files_collector/repository_documents.py

@@ -4,6 +4,7 @@
 from cycode.cli import consts
 from cycode.cli.files_collector.sca import sca_code_scanner
 from cycode.cli.models import Document
+from cycode.cli.utils.git_proxy import git_proxy
 from cycode.cli.utils.path_utils import get_file_content, get_path_by_os
 
 if TYPE_CHECKING:
@@ -13,8 +14,6 @@
 
     from cycode.cli.utils.progress_bar import BaseProgressBar, ProgressBarSection
 
-from git import Repo
-
 
 def should_process_git_object(obj: 'Blob', _: int) -> bool:
     return obj.type == 'blob' and obj.size > 0
@@ -23,14 +22,14 @@ def should_process_git_object(obj: 'Blob', _: int) -> bool:
 def get_git_repository_tree_file_entries(
     path: str, branch: str
 ) -> Union[Iterator['IndexObjUnion'], Iterator['TraversedTreeTup']]:
-    return Repo(path).tree(branch).traverse(predicate=should_process_git_object)
+    return git_proxy.get_repo(path).tree(branch).traverse(predicate=should_process_git_object)
 
 
 def parse_commit_range(commit_range: str, path: str) -> Tuple[str, str]:
     from_commit_rev = None
     to_commit_rev = None
 
-    for commit in Repo(path).iter_commits(rev=commit_range):
+    for commit in git_proxy.get_repo(path).iter_commits(rev=commit_range):
         if not to_commit_rev:
             to_commit_rev = commit.hexsha
         from_commit_rev = commit.hexsha
@@ -52,7 +51,7 @@ def get_pre_commit_modified_documents(
     git_head_documents = []
     pre_committed_documents = []
 
-    repo = Repo(os.getcwd())
+    repo = git_proxy.get_repo(os.getcwd())
     diff_files = repo.index.diff(consts.GIT_HEAD_COMMIT_REV, create_patch=True, R=True)
     progress_bar.set_section_length(progress_bar_section, len(diff_files))
     for file in diff_files:
@@ -82,7 +81,7 @@ def get_commit_range_modified_documents(
     from_commit_documents = []
     to_commit_documents = []
 
-    repo = Repo(path)
+    repo = git_proxy.get_repo(path)
     diff = repo.commit(from_commit_rev).diff(to_commit_rev)
 
     modified_files_diff = [
@@ -131,7 +130,8 @@ def _get_end_commit_from_branch_update_details(update_details: str) -> str:
 def _get_oldest_unupdated_commit_for_branch(commit: str) -> Optional[str]:
     # get a list of commits by chronological order that are not in the remote repository yet
     # more info about rev-list command: https://git-scm.com/docs/git-rev-list
-    not_updated_commits = Repo(os.getcwd()).git.rev_list(commit, '--topo-order', '--reverse', '--not', '--all')
+    repo = git_proxy.get_repo(os.getcwd())
+    not_updated_commits = repo.git.rev_list(commit, '--topo-order', '--reverse', '--not', '--all')
 
     commits = not_updated_commits.splitlines()
     if not commits:

cycode/cli/files_collector/sca/sca_code_scanner.py

@@ -2,16 +2,18 @@
 from typing import TYPE_CHECKING, Dict, List, Optional
 
 import click
-from git import GitCommandError, Repo
 
 from cycode.cli import consts
 from cycode.cli.files_collector.sca.maven.restore_gradle_dependencies import RestoreGradleDependencies
 from cycode.cli.files_collector.sca.maven.restore_maven_dependencies import RestoreMavenDependencies
 from cycode.cli.models import Document
+from cycode.cli.utils.git_proxy import git_proxy
 from cycode.cli.utils.path_utils import get_file_content, get_file_dir, join_paths
 from cycode.cyclient import logger
 
 if TYPE_CHECKING:
+    from git import Repo
+
     from cycode.cli.files_collector.sca.maven.base_restore_maven_dependencies import BaseRestoreMavenDependencies
 
 BUILD_GRADLE_FILE_NAME = 'build.gradle'
@@ -27,21 +29,21 @@ def perform_pre_commit_range_scan_actions(
     to_commit_documents: List[Document],
     to_commit_rev: str,
 ) -> None:
-    repo = Repo(path)
+    repo = git_proxy.get_repo(path)
     add_ecosystem_related_files_if_exists(from_commit_documents, repo, from_commit_rev)
     add_ecosystem_related_files_if_exists(to_commit_documents, repo, to_commit_rev)
 
 
 def perform_pre_hook_range_scan_actions(
     git_head_documents: List[Document], pre_committed_documents: List[Document]
 ) -> None:
-    repo = Repo(os.getcwd())
+    repo = git_proxy.get_repo(os.getcwd())
     add_ecosystem_related_files_if_exists(git_head_documents, repo, consts.GIT_HEAD_COMMIT_REV)
     add_ecosystem_related_files_if_exists(pre_committed_documents)
 
 
 def add_ecosystem_related_files_if_exists(
-    documents: List[Document], repo: Optional[Repo] = None, commit_rev: Optional[str] = None
+    documents: List[Document], repo: Optional['Repo'] = None, commit_rev: Optional[str] = None
 ) -> None:
     documents_to_add: List[Document] = []
     for doc in documents:
@@ -56,7 +58,7 @@ def add_ecosystem_related_files_if_exists(
 
 
 def get_doc_ecosystem_related_project_files(
-    doc: Document, documents: List[Document], ecosystem: str, commit_rev: Optional[str], repo: Optional[Repo]
+    doc: Document, documents: List[Document], ecosystem: str, commit_rev: Optional[str], repo: Optional['Repo']
 ) -> List[Document]:
     documents_to_add: List[Document] = []
     for ecosystem_project_file in consts.PROJECT_FILES_BY_ECOSYSTEM_MAP.get(ecosystem):
@@ -136,10 +138,10 @@ def get_manifest_file_path(document: Document, is_monitor_action: bool, project_
     return join_paths(project_path, document.path) if is_monitor_action else document.path
 
 
-def get_file_content_from_commit(repo: Repo, commit: str, file_path: str) -> Optional[str]:
+def get_file_content_from_commit(repo: 'Repo', commit: str, file_path: str) -> Optional[str]:
     try:
         return repo.git.show(f'{commit}:{file_path}')
-    except GitCommandError:
+    except git_proxy.get_git_command_error():
         return None
 
 

cycode/cli/printers/printer_base.py

@@ -43,7 +43,7 @@ def print_exception(self, e: Optional[BaseException] = None) -> None:
             # gets the most recent exception caught by an except clause
             message = f'Error: {traceback.format_exc()}'
         else:
-            traceback_message = ''.join(traceback.format_exception(e))
+            traceback_message = ''.join(traceback.format_exception(None, e, e.__traceback__))
             message = f'Error: {traceback_message}'
 
         click.secho(message, err=True, fg=self.RED_COLOR_NAME)

cycode/cli/utils/git_proxy.py

@@ -0,0 +1,82 @@
+from abc import ABC, abstractmethod
+from functools import lru_cache
+from typing import TYPE_CHECKING, Optional, Type
+
+_GIT_ERROR_MESSAGE = """
+Cycode CLI needs the git executable to be installed on the system.
+Git executable must be available in the PATH.
+Git 1.7.x or newer is required.
+You can help Cycode CLI to locate the Git executable
+by setting GIT_PYTHON_GIT_EXECUTABLE=<path/to/git> environment variable.
+""".strip().replace('\n', ' ')
+
+try:
+    import git
+except ImportError:
+    git = None
+
+if TYPE_CHECKING:
+    from git import PathLike, Repo
+
+
+class GitProxyError(Exception):
+    pass
+
+
+class _AbstractGitProxy(ABC):
+    @abstractmethod
+    def get_repo(self, path: Optional['PathLike'] = None) -> 'Repo':
+        ...
+
+    @abstractmethod
+    def get_null_tree(self) -> object:
+        ...
+
+    @abstractmethod
+    def get_invalid_git_repository_error(self) -> Type[GitProxyError]:
+        ...
+
+    @abstractmethod
+    def get_git_command_error(self) -> Type[GitProxyError]:
+        ...
+
+
+class _DummyGitProxy(_AbstractGitProxy):
+    def get_repo(self, path: Optional['PathLike'] = None) -> 'Repo':
+        raise RuntimeError(_GIT_ERROR_MESSAGE)
+
+    def get_null_tree(self) -> object:
+        return object()
+
+    def get_invalid_git_repository_error(self) -> Type[GitProxyError]:
+        return GitProxyError
+
+    def get_git_command_error(self) -> Type[GitProxyError]:
+        return GitProxyError
+
+
+class _GitProxy(_AbstractGitProxy):
+    def get_repo(self, path: Optional['PathLike'] = None) -> 'Repo':
+        return git.Repo(path)
+
+    def get_null_tree(self) -> object:
+        return git.NULL_TREE
+
+    @lru_cache(maxsize=None)  # noqa: B019
+    def get_invalid_git_repository_error(self) -> Type[GitProxyError]:
+        # we must cache it because we want to return the same class every time
+        class InvalidGitRepositoryError(GitProxyError, git.InvalidGitRepositoryError):
+            ...
+
+        return InvalidGitRepositoryError
+
+    @lru_cache(maxsize=None)  # noqa: B019
+    def get_git_command_error(self) -> Type[GitProxyError]:
+        # we must cache it because we want to return the same class every time
+        class GitCommandError(GitProxyError, git.GitCommandError):
+            ...
+
+        return GitCommandError
+
+
+git_proxy = _GitProxy() if git else _DummyGitProxy()

tests/cli/exceptions/test_handle_scan_errors.py

@@ -3,11 +3,11 @@
 import click
 import pytest
 from click import ClickException
-from git import InvalidGitRepositoryError
 from requests import Response
 
 from cycode.cli.exceptions import custom_exceptions
 from cycode.cli.exceptions.handle_scan_errors import handle_scan_exception
+from cycode.cli.utils.git_proxy import git_proxy
 
 if TYPE_CHECKING:
     from _pytest.monkeypatch import MonkeyPatch
@@ -26,7 +26,7 @@ def ctx() -> click.Context:
         (custom_exceptions.HttpUnauthorizedError('msg', Response()), True),
         (custom_exceptions.ZipTooLargeError(1000), True),
         (custom_exceptions.TfplanKeyError('msg'), True),
-        (InvalidGitRepositoryError(), None),
+        (git_proxy.get_invalid_git_repository_error()(), None),
     ],
 )
 def test_handle_exception_soft_fail(
@@ -40,7 +40,7 @@ def test_handle_exception_soft_fail(
 
 
 def test_handle_exception_unhandled_error(ctx: click.Context) -> None:
-    with ctx, pytest.raises(ClickException):
+    with ctx, pytest.raises(SystemExit):
         handle_scan_exception(ctx, ValueError('test'))
 
         assert ctx.obj.get('did_fail') is True
@@ -58,10 +58,12 @@ def test_handle_exception_click_error(ctx: click.Context) -> None:
 def test_handle_exception_verbose(monkeypatch: 'MonkeyPatch') -> None:
     ctx = click.Context(click.Command('path'), obj={'verbose': True, 'output': 'text'})
 
+    error_text = 'test'
+
     def mock_secho(msg: str, *_, **__) -> None:
-        assert 'Error:' in msg or 'Correlation ID:' in msg
+        assert error_text in msg or 'Correlation ID:' in msg
 
     monkeypatch.setattr(click, 'secho', mock_secho)
 
-    with ctx, pytest.raises(ClickException):
-        handle_scan_exception(ctx, ValueError('test'))
+    with pytest.raises(SystemExit):
+        handle_scan_exception(ctx, ValueError(error_text))