Feature/secureli 436 language detection (#481)

secureli-436

<!-- Include general description here -->
closes #436 

Adds focused language detection during scan for specified or committed
files


## Changes
<!-- A detailed list of changes -->
* Limiting language detection to a specified set of files during scan
(unless all files is provided)
* Adding the usage of git.Repo to check the commit file diff

## Testing
<!--
Mention updated tests and any manual testing performed.
Are aspects not yet tested or not easily testable?
Feel free to include screenshots if appropriate.
 -->
* Verify that language detection can be run on either specified or
committed files during scan

## Clean Code Checklist
<!-- This is here to support you. Some/most checkboxes may not apply to
your change -->
- [x] Meets acceptance criteria for issue
- [x] New logic is covered with automated tests
- [x] Appropriate exception handling added
- [x] Thoughtful logging included
- [x] Documentation is updated
- [x] Follow-up work is documented in TODOs
- [x] TODOs have a ticket associated with them
- [x] No commented-out code included


<!--
Github-flavored markdown reference:
https://docs.github.com/en/get-started/writing-on-github
-->

---------

Co-authored-by: Jordan Heffernan <[email protected]>

.secureli/.pre-commit-config.yaml

@@ -9,7 +9,7 @@ repos:
     -   id: check-yaml
     -   id: check-added-large-files
 -   repo: https://github.com/psf/black-pre-commit-mirror
-    rev: 24.2.0
+    rev: 24.3.0
     hooks:
     -   id: black
 -   repo: https://github.com/yelp/detect-secrets

secureli/actions/action.py

@@ -47,16 +47,24 @@ class Action(ABC):
     def __init__(self, action_deps: ActionDependencies):
         self.action_deps = action_deps
 
+    def get_secureli_config(self, reset: bool) -> secureli_config.SecureliConfig:
+        return (
+            secureli_config.SecureliConfig()
+            if reset
+            else self.action_deps.secureli_config.load()
+        )
+
     def verify_install(
-        self, folder_path: Path, reset: bool, always_yes: bool
+        self, folder_path: Path, reset: bool, always_yes: bool, files: list[Path]
     ) -> VerifyResult:
         """
         Installs, upgrades or verifies the current seCureLI installation
         :param folder_path: The folder path to initialize the repo for
         :param reset: If true, disregard existing configuration and start fresh
         :param always_yes: Assume "Yes" to all prompts
+        :param files: A List of files to scope the install to. This allows language
+        detection to run on only a selected list of files when scanning the repo.
         """
-
         if (
             self.action_deps.secureli_config.verify()
             == secureli_config.VerifyConfigOutcome.OUT_OF_DATE
@@ -86,15 +94,11 @@ def verify_install(
                 )
                 return update_result
 
-        config = (
-            secureli_config.SecureliConfig()
-            if reset
-            else self.action_deps.secureli_config.load()
-        )
+        config = self.get_secureli_config(reset=reset)
         languages = []
 
         try:
-            languages = self._detect_languages(folder_path)
+            languages = self._detect_languages(folder_path, files)
         except (ValueError, language.LanguageNotSupportedError) as e:
             if config.languages and config.version_installed:
                 self.action_deps.echo.warning(
@@ -298,14 +302,14 @@ def _run_post_install_scan(
                 f"{format_sentence_list(config.languages)} does not support secrets detection, skipping"
             )
 
-    def _detect_languages(self, folder_path: Path) -> list[str]:
+    def _detect_languages(self, folder_path: Path, files: list[Path]) -> list[str]:
         """
         Detects programming languages present in the repository
         :param folder_path: The folder path to initialize the repo for
         :return: A list of all languages found in the repository
         """
 
-        analyze_result = self.action_deps.language_analyzer.analyze(folder_path)
+        analyze_result = self.action_deps.language_analyzer.analyze(folder_path, files)
 
         if analyze_result.skipped_files:
             self.action_deps.echo.warning(

secureli/actions/initializer.py

@@ -27,7 +27,7 @@ def initialize_repo(
         :param reset: If true, disregard existing configuration and start fresh
         :param always_yes: Assume "Yes" to all prompts
         """
-        verify_result = self.verify_install(folder_path, reset, always_yes)
+        verify_result = self.verify_install(folder_path, reset, always_yes, files=None)
         if verify_result.outcome in ScanAction.halting_outcomes:
             self.logging.failure(LogAction.init, verify_result.outcome)
         else:

secureli/actions/scan.py

@@ -3,9 +3,11 @@
 from pathlib import Path
 from time import time
 from typing import Optional
+from git import Repo
 
 from secureli.modules.shared.abstractions.echo import EchoAbstraction
 from secureli.actions import action
+from secureli.modules.shared.abstractions.repo import GitRepo
 from secureli.modules.shared.models.exit_codes import ExitCode
 from secureli.modules.shared.models.install import VerifyOutcome, VerifyResult
 from secureli.modules.shared.models.logging import LogAction
@@ -35,11 +37,13 @@ def __init__(
         echo: EchoAbstraction,
         logging: LoggingService,
         scanner: ScannerService,
+        git_repo: GitRepo,
     ):
         super().__init__(action_deps)
         self.scanner = scanner
         self.echo = echo
         self.logging = logging
+        self.git_repo = git_repo
 
     def _check_secureli_hook_updates(self, folder_path: Path) -> VerifyResult:
         """
@@ -93,6 +97,24 @@ def publish_results(
             else:
                 self.logging.failure(LogAction.publish, result.result_message)
 
+    def get_commited_files(self, scan_mode: ScanMode) -> list[Path]:
+        """
+        Attempts to build a list of commited files for use in language detection if
+        the user is scanning staged files for an existing installation
+        :param scan_mode: Determines which files are scanned in the repo (i.e. staged only or all)
+        :returns: a list of Path objects for the commited files
+        """
+        config = self.get_secureli_config(reset=False)
+        installed = bool(config.languages and config.version_installed)
+
+        if not installed or scan_mode != ScanMode.STAGED_ONLY:
+            return None
+        try:
+            committed_files = self.git_repo.get_commit_diff()
+            return [Path(file) for file in committed_files]
+        except:
+            return None
+
     def scan_repo(
         self,
         folder_path: Path,
@@ -112,7 +134,16 @@ def scan_repo(
         :param specific_test: If set, limits scanning to the single pre-commit hook.
         Otherwise, scans with all hooks.
         """
-        verify_result = self.verify_install(folder_path, False, always_yes)
+
+        scan_files = [Path(file) for file in files or []] or self.get_commited_files(
+            scan_mode
+        )
+        verify_result = self.verify_install(
+            folder_path,
+            False,
+            always_yes,
+            scan_files,
+        )
 
         # Check if pre-commit hooks are up-to-date
         secureli_config = self.action_deps.secureli_config.load()

secureli/container.py

@@ -8,6 +8,7 @@
 from secureli.actions.scan import ScanAction
 from secureli.actions.build import BuildAction
 from secureli.actions.update import UpdateAction
+from secureli.modules.shared.abstractions.repo import GitRepo
 from secureli.repositories.repo_files import RepoFilesRepository
 from secureli.repositories.secureli_config import SecureliConfigRepository
 from secureli.repositories.repo_settings import SecureliRepository
@@ -82,6 +83,9 @@ class Container(containers.DeclarativeContainer):
         echo=echo,
     )
 
+    """Wraps the execution and management of git commands"""
+    git_repo = providers.Factory(GitRepo)
+
     # Services
 
     """Analyzes a set of files to try to determine the most common languages"""
@@ -172,7 +176,7 @@ class Container(containers.DeclarativeContainer):
         echo=echo,
         logging=logging_service,
         scanner=scanner_service,
-        # settings_repository=settings_repository,
+        git_repo=git_repo,
     )
 
     """Update Action, representing what happens when the update command is invoked"""

secureli/modules/language_analyzer/language_analyzer.py

@@ -20,7 +20,7 @@ def __init__(
         self.repo_files = repo_files
         self.lexer_guesser = lexer_guesser
 
-    def analyze(self, folder_path: Path) -> AnalyzeResult:
+    def analyze(self, folder_path: Path, files: list[Path]) -> AnalyzeResult:
         """
         Analyzes the folder structure and lists languages found
         :param folder_path: The path to the repository to analyze
@@ -29,7 +29,7 @@ def analyze(self, folder_path: Path) -> AnalyzeResult:
         40% of the repo is JavaScript, the result will be a dictionary containing keys
         "Python" and "JavaScript" with values 0.6 and 0.4 respectively
         """
-        file_paths = self.repo_files.list_repo_files(folder_path)
+        file_paths = files if files else self.repo_files.list_repo_files(folder_path)
         results = defaultdict(int)
 
         skipped_files = []

secureli/modules/shared/abstractions/repo.py

@@ -0,0 +1,21 @@
+from abc import ABC, abstractmethod
+import git
+
+
+class RepoAbstraction(ABC):
+    """
+    Abstracts the configuring and execution of git repo features.
+    """
+
+    @abstractmethod
+    def get_commit_diff(self) -> list[str]:
+        pass
+
+
+class GitRepo(RepoAbstraction):
+    """
+    Implementation and wrapper around git repo features
+    """
+
+    def get_commit_diff(self) -> list[str]:
+        return git.Repo().head.commit.diff()