Merge pull request #68 from FalkorDB/fix-staging

move to pygit2

Author: AviAvni

api/git_utils/git_graph.py

@@ -1,9 +1,10 @@
 import os
 import logging
-from git import Commit
 from falkordb import FalkorDB, Node
 from typing import List, Optional
 
+from pygit2 import Commit
+
 # Configure logging
 logging.basicConfig(level=logging.DEBUG, format='%(filename)s - %(asctime)s - %(levelname)s - %(message)s')
 
@@ -112,7 +113,7 @@ def connect_commits(self, child: str, parent: str) -> None:
         self.g.query(q, params)
 
 
-    def set_parent_transition(self, child: str, parent: str, queries: [str], params: [str]) -> None:
+    def set_parent_transition(self, child: str, parent: str, queries: list[str], params: list[str]) -> None:
         """
             Sets the queries and parameters needed to transition the code-graph
             from the child commit to the parent commit
@@ -126,7 +127,7 @@ def set_parent_transition(self, child: str, parent: str, queries: [str], params:
         self.g.query(q, _params)
 
 
-    def set_child_transition(self, child: str, parent: str, queries: [str], params: [str]) -> None:
+    def set_child_transition(self, child: str, parent: str, queries: list[str], params: list[str]) -> None:
         """
             Sets the queries and parameters needed to transition the code-graph
             from the parent commit to the child commit

api/git_utils/git_utils.py

@@ -1,8 +1,10 @@
 import os
 import json
 import logging
+
+from pygit2 import Commit
 from ..info import *
-from git import Repo
+from pygit2.repository import Repository
 from pathlib import Path
 from ..graph import Graph
 from .git_graph import GitGraph
@@ -85,9 +87,9 @@ def build_commit_graph(path: str, repo_name: str, ignore_list: Optional[List[str
 
     # Initialize with the current commit
     # Save current git for later restoration
-    repo = Repo('.')
-    current_commit = repo.head.commit
-    current_commit_hexsha = current_commit.hexsha
+    repo = Repository('.')
+    current_commit = repo.walk(repo.head.target).__next__()
+    current_commit_hexsha = current_commit.hex
 
     # Add commit to the git graph
     git_graph.add_commit(current_commit)
@@ -106,7 +108,7 @@ def build_commit_graph(path: str, repo_name: str, ignore_list: Optional[List[str
         git_graph.add_commit(parent_commit)
 
         # connect child parent commits relation
-        git_graph.connect_commits(child_commit.hexsha, parent_commit.hexsha)
+        git_graph.connect_commits(child_commit.hex, parent_commit.hex)
 
         # Represents the changes going backward!
         # e.g. which files need to be deleted when moving back one commit
@@ -126,7 +128,7 @@ def build_commit_graph(path: str, repo_name: str, ignore_list: Optional[List[str
 
         # Checkout prev commit
         logging.info(f"Checking out commit: {parent_commit.hexsha}")
-        repo.git.checkout(parent_commit.hexsha)
+        repo.checkout(parent_commit.hex)
 
         #-----------------------------------------------------------------------
         # Apply changes going backwards
@@ -165,15 +167,15 @@ def build_commit_graph(path: str, repo_name: str, ignore_list: Optional[List[str
 
             # Log transitions
             logging.debug(f"""Save graph transition from
-                             commit: {child_commit.hexsha}
+                             commit: {child_commit.hex}
                              to
-                             commit: {parent_commit.hexsha}
+                             commit: {parent_commit.hex}
                              Queries: {queries}
                              Parameters: {params}
                           """)
 
-            git_graph.set_parent_transition(child_commit.hexsha,
-                                            parent_commit.hexsha, queries, params)
+            git_graph.set_parent_transition(child_commit.hex,
+                                            parent_commit.hex, queries, params)
         # advance to the next commit
         child_commit = parent_commit
 
@@ -183,24 +185,24 @@ def build_commit_graph(path: str, repo_name: str, ignore_list: Optional[List[str
 
     logging.info("Computing transition queries moving forward")
     parent_commit = child_commit
-    while parent_commit.hexsha != current_commit_hexsha:
-        child_commit = git_graph.get_child_commit(parent_commit.hexsha)
-        child_commit = repo.commit(child_commit['hash'])
+    while parent_commit.hex != current_commit_hexsha:
+        child_commit = git_graph.get_child_commit(parent_commit.hex)
+        child_commit = repo.walk(child_commit['hash']).__next__()
 
         # Represents the changes going forward
         # e.g. which files need to be deleted when moving forward one commit
 
         # Process file changes in this commit
         logging.info(f"""Computing diff between
-            child {parent_commit.hexsha}: {parent_commit.message}
-            and {child_commit.hexsha}: {child_commit.message}""")
+            child {parent_commit.hex}: {parent_commit.message}
+            and {child_commit.hex}: {child_commit.message}""")
 
-        diff = parent_commit.diff(child_commit)
+        diff = repo.diff(parent_commit, child_commit)
         added, deleted, modified = classify_changes(diff, ignore_list)
 
         # Checkout child commit
-        logging.info(f"Checking out commit: {child_commit.hexsha}")
-        repo.git.checkout(child_commit.hexsha)
+        logging.info(f"Checking out commit: {child_commit.hex}")
+        repo.checkout(child_commit.hex)
 
         #-----------------------------------------------------------------------
         # Apply changes going forward
@@ -239,15 +241,15 @@ def build_commit_graph(path: str, repo_name: str, ignore_list: Optional[List[str
 
             # Log transitions
             logging.debug(f"""Save graph transition from
-                             commit: {parent_commit.hexsha}
+                             commit: {parent_commit.hex}
                              to
-                             commit: {child_commit.hexsha}
+                             commit: {child_commit.hex}
                              Queries: {queries}
                              Parameters: {params}
                           """)
 
-            git_graph.set_child_transition(child_commit.hexsha,
-                                            parent_commit.hexsha, queries, params)
+            git_graph.set_child_transition(child_commit.hex,
+                                            parent_commit.hex, queries, params)
         # advance to the child_commit
         parent_commit = child_commit
 

api/project.py

@@ -3,7 +3,7 @@
 import logging
 import validators
 import subprocess
-from git import Repo
+from pygit2.repository import Repository
 from .info import *
 from shlex import quote
 from pathlib import Path
@@ -70,7 +70,7 @@ def from_local_repository(cls, path: Path|str):
 
         # adjust url
         # '[email protected]:FalkorDB/code_graph.git'
-        url  = Repo(path).remotes[0].url
+        url  = Repository(path).remotes[0].url
         url = url.replace("git@", "https://").replace(":", "/").replace(".git", "")
 
         name = path.name
@@ -85,9 +85,9 @@ def analyze_sources(self, ignore: Optional[List[str]] = None) -> Graph:
 
         try:
             # Save processed commit hash to the DB
-            repo = Repo(self.path)
-            current_commit = repo.head.commit
-            set_repo_commit(self.name, current_commit.hexsha)
+            repo = Repository(self.path)
+            current_commit = repo.walk(repo.head.target).__next__()
+            set_repo_commit(self.name, current_commit.hex)
         except Exception:
             # Probably not .git folder is missing
             pass

poetry.lock

@@ -10,7 +10,6 @@ python = "^3.10"
 graphrag-sdk = { version = "^0.5.0", extras = ["litellm"] }
 tree-sitter = "^0.24.0"
 validators = "^0.34.0"
-GitPython = "^3.1.44"
 falkordb = "^1.0.10"
 tree-sitter-c = "^0.23.4"
 tree-sitter-python = "^0.23.6"
@@ -19,6 +18,7 @@ flask = "^3.1.0"
 python-dotenv = "^1.0.1"
 multilspy = {git = "https://github.com/AviAvni/multilspy.git", rev = "update-pydantic"}
 javatools = "^1.6.0"
+pygit2 = "^1.17.0"
 
 [tool.poetry.group.test.dependencies]
 pytest = "^8.2.0"

pyproject.toml

@@ -13,7 +13,7 @@ blinker==1.9.0 ; python_version >= "3.10" and python_version < "4.0"
 bs4==0.0.2 ; python_version >= "3.10" and python_version < "4.0"
 cattrs==24.1.2 ; python_version >= "3.10" and python_version < "4.0"
 certifi==2024.12.14 ; python_version >= "3.10" and python_version < "4.0"
-cffi==1.17.1 ; python_version >= "3.10" and python_version < "4.0" and implementation_name == "pypy"
+cffi==1.17.1 ; python_version >= "3.10" and python_version < "4.0"
 charset-normalizer==3.4.1 ; python_version >= "3.10" and python_version < "4.0"
 click==8.1.8 ; python_version >= "3.10" and python_version < "4.0"
 colorama==0.4.6 ; python_version >= "3.10" and python_version < "4.0" and sys_platform == "win32" or python_version >= "3.10" and python_version < "4.0" and platform_system == "Windows"
@@ -31,8 +31,6 @@ fix-busted-json==0.0.18 ; python_version >= "3.10" and python_version < "4.0"
 flask==3.1.0 ; python_version >= "3.10" and python_version < "4.0"
 frozenlist==1.5.0 ; python_version >= "3.10" and python_version < "4.0"
 fsspec==2024.12.0 ; python_version >= "3.10" and python_version < "4.0"
-gitdb==4.0.12 ; python_version >= "3.10" and python_version < "4.0"
-gitpython==3.1.44 ; python_version >= "3.10" and python_version < "4.0"
 graphrag-sdk==0.5.0 ; python_version >= "3.10" and python_version < "4.0"
 h11==0.14.0 ; python_version >= "3.10" and python_version < "4.0"
 httpcore==1.0.7 ; python_version >= "3.10" and python_version < "4.0"
@@ -70,9 +68,10 @@ propcache==0.2.1 ; python_version >= "3.10" and python_version < "4.0"
 psutil==6.1.1 ; python_version >= "3.10" and python_version < "4.0"
 ptyprocess==0.7.0 ; python_version >= "3.10" and python_version < "4.0" and (sys_platform != "win32" and sys_platform != "emscripten")
 pure-eval==0.2.3 ; python_version >= "3.10" and python_version < "4.0"
-pycparser==2.22 ; python_version >= "3.10" and python_version < "4.0" and implementation_name == "pypy"
+pycparser==2.22 ; python_version >= "3.10" and python_version < "4.0"
 pydantic-core==2.27.2 ; python_version >= "3.10" and python_version < "4.0"
 pydantic==2.10.6 ; python_version >= "3.10" and python_version < "4.0"
+pygit2==1.17.0 ; python_version >= "3.10" and python_version < "4.0"
 pygls==1.3.1 ; python_version >= "3.10" and python_version < "4.0"
 pygments==2.19.1 ; python_version >= "3.10" and python_version < "4.0"
 pypdf==4.3.1 ; python_version >= "3.10" and python_version < "4.0"
@@ -89,7 +88,6 @@ regex==2024.11.6 ; python_version >= "3.10" and python_version < "4.0"
 requests==2.32.3 ; python_version >= "3.10" and python_version < "4.0"
 rpds-py==0.22.3 ; python_version >= "3.10" and python_version < "4.0"
 six==1.17.0 ; python_version >= "3.10" and python_version < "4.0"
-smmap==5.0.2 ; python_version >= "3.10" and python_version < "4.0"
 sniffio==1.3.1 ; python_version >= "3.10" and python_version < "4.0"
 soupsieve==2.6 ; python_version >= "3.10" and python_version < "4.0"
 stack-data==0.6.3 ; python_version >= "3.10" and python_version < "4.0"