Merge pull request #897 from ATheorell/main

Updates of build system with related changes to CI

.pre-commit-config.yaml

@@ -11,11 +11,11 @@ repos:
         args: [--config, pyproject.toml]
         types: [python]
 
-#  - repo: https://github.com/charliermarsh/ruff-pre-commit
-#    rev: "v0.0.272"
-#    hooks:
-#      - id: ruff
-#        args: [--fix, --exit-non-zero-on-fix]
+  - repo: https://github.com/charliermarsh/ruff-pre-commit
+    rev: "v0.0.272"
+    hooks:
+      - id: ruff
+        args: [--fix, --exit-non-zero-on-fix]
 
   - repo: https://github.com/pre-commit/pre-commit-hooks
     rev: v4.4.0

gpt_engineer/__init__.py

@@ -1,8 +1,4 @@
 # Adding convenience imports to the package
-from gpt_engineer.core import (
-    ai,
-    chat_to_files,
-)
 
 # from gpt_engineer.tools import code_vector_repository
 # from gpt_engineer.core.default import on_disk_repository

gpt_engineer/applications/cli/cli_agent.py

@@ -1,22 +1,21 @@
-from gpt_engineer.core.files_dict import FilesDict
+from typing import Callable, TypeVar
 
 # from gpt_engineer.core.default.git_version_manager import GitVersionManager
 from gpt_engineer.core.ai import AI
+from gpt_engineer.core.base_agent import BaseAgent
+from gpt_engineer.core.base_execution_env import BaseExecutionEnv
+from gpt_engineer.core.base_memory import BaseMemory
+from gpt_engineer.core.default.disk_execution_env import DiskExecutionEnv
+from gpt_engineer.core.default.disk_memory import DiskMemory
+from gpt_engineer.core.default.paths import ENTRYPOINT_FILE, PREPROMPTS_PATH
 from gpt_engineer.core.default.steps import (
+    execute_entrypoint,
     gen_code,
     gen_entrypoint,
-    execute_entrypoint,
     improve,
 )
-from gpt_engineer.core.base_memory import BaseMemory
-from gpt_engineer.core.default.disk_memory import DiskMemory
-from gpt_engineer.core.base_execution_env import BaseExecutionEnv
-from gpt_engineer.core.default.disk_execution_env import DiskExecutionEnv
-from gpt_engineer.core.default.paths import memory_path, ENTRYPOINT_FILE, PREPROMPTS_PATH
-from gpt_engineer.core.base_agent import BaseAgent
+from gpt_engineer.core.files_dict import FilesDict
 from gpt_engineer.core.preprompts_holder import PrepromptsHolder
-from typing import TypeVar, Callable, Union
-from pathlib import Path
 
 CodeGenType = TypeVar("CodeGenType", bound=Callable[[AI, str, BaseMemory], FilesDict])
 CodeProcessor = TypeVar(

gpt_engineer/applications/cli/collect.py

@@ -23,17 +23,16 @@
     Consent logic is in gpt_engineer/learning.py.
 
 """
-import hashlib
 
-from typing import List, Tuple
+from typing import Tuple
 
-from gpt_engineer.core.default.disk_memory import DiskMemory
 from gpt_engineer.applications.cli.learning import (
     Learning,
+    Review,
     extract_learning,
     human_review_input,
-    Review,
 )
+from gpt_engineer.core.default.disk_memory import DiskMemory
 
 
 def send_learning(learning: Learning):
@@ -88,7 +87,7 @@ def collect_learnings(
     learnings = extract_learning(prompt, model, temperature, config, memory, review)
     try:
         send_learning(learnings)
-    except RuntimeError as e:
+    except RuntimeError:
         # try to remove some parts of learning that might be too big
         # rudderstack max event size is 32kb
         max_size = 32 << 10  # 32KB in bytes
@@ -100,7 +99,8 @@ def collect_learnings(
         remove_length = overflow + len(f"[REMOVED {overflow} CHARACTERS]") + 100
 
         learnings.logs = (
-            learnings.logs[:-remove_length] + f"\n\n[REMOVED {remove_length} CHARACTERS]"
+            learnings.logs[:-remove_length]
+            + f"\n\n[REMOVED {remove_length} CHARACTERS]"
         )
 
         print(
@@ -109,7 +109,7 @@ def collect_learnings(
         )
         try:
             send_learning(learnings)
-        except RuntimeError as e:
+        except RuntimeError:
             print(
                 "Sending learnings crashed despite truncation. Progressing without saving learnings."
             )

gpt_engineer/applications/cli/file_selector.py

@@ -119,7 +119,9 @@ def display_name(self) -> str:
         return self.path.name
 
     @classmethod
-    def make_tree(cls, root: Union[str, Path], parent=None, is_last=False, criteria=None):
+    def make_tree(
+        cls, root: Union[str, Path], parent=None, is_last=False, criteria=None
+    ):
         """
         Generate a tree of DisplayablePath objects.
 

gpt_engineer/applications/cli/learning.py

@@ -47,7 +47,7 @@
 from dataclasses import dataclass, field
 from datetime import datetime
 from pathlib import Path
-from typing import List, Optional, Tuple
+from typing import Optional, Tuple
 
 from dataclasses_json import dataclass_json
 from termcolor import colored

gpt_engineer/applications/cli/main.py

@@ -25,32 +25,39 @@
 
 """
 
+import logging
 import os
+
+from importlib.util import find_spec
 from pathlib import Path
 
 import openai
+import toml
 import typer
+
 from dotenv import load_dotenv
 
-from gpt_engineer.core.default.file_store import FileStore
-from gpt_engineer.core.default.disk_memory import DiskMemory
-from gpt_engineer.core.ai import AI
-from gpt_engineer.core.default.paths import PREPROMPTS_PATH, memory_path
-from gpt_engineer.applications.cli.file_selector import ask_for_files, get_all_code
-from gpt_engineer.tools.custom_steps import (
-    lite_gen,
-    clarified_gen,
-    self_heal,
-)
-from gpt_engineer.tools.experimental.experimental_steps import (
-    improve_automatic_file_selection,
-)
-from gpt_engineer.core.default.steps import gen_code, execute_entrypoint, improve
 from gpt_engineer.applications.cli.cli_agent import CliAgent
 from gpt_engineer.applications.cli.collect import collect_and_send_human_review
-from gpt_engineer.core.preprompts_holder import PrepromptsHolder
+from gpt_engineer.applications.cli.file_selector import ask_for_files
+from gpt_engineer.core.ai import AI
 from gpt_engineer.core.default.disk_execution_env import DiskExecutionEnv
-import logging
+from gpt_engineer.core.default.disk_memory import DiskMemory
+from gpt_engineer.core.default.file_store import FileStore
+from gpt_engineer.core.default.paths import PREPROMPTS_PATH, memory_path
+from gpt_engineer.core.default.steps import execute_entrypoint, gen_code, improve
+from gpt_engineer.core.preprompts_holder import PrepromptsHolder
+from gpt_engineer.tools.custom_steps import clarified_gen, lite_gen, self_heal
+
+# Load the names of the optional dependencies from the pyprojecct file and determine whether
+# they can be imported
+with open("pyproject.toml", "r") as file:
+    pyproject = toml.load(file)
+
+dependency_group = pyproject["tool"]["poetry"]["group"]["experimental"]["dependencies"]
+optional_deps_importable = all(
+    [find_spec(dep_name.replace("-", "_")) is not None for dep_name in dependency_group]
+)
 
 app = typer.Typer()  # creates a CLI app
 
@@ -105,7 +112,7 @@ def main(
     ),
     improve_all_mode: bool = typer.Option(
         False,
-        "--improve_all_experimental",
+        "--improve-all-experimental",
         help="Improve files_dict from existing project, without manually choosing which files to improve, using vector store (EXPERIMENTAL).",
     ),
     lite_mode: bool = typer.Option(
@@ -185,8 +192,16 @@ def main(
     else:
         execution_fn = execute_entrypoint
 
-    if improve_all_mode:
+    if improve_all_mode and optional_deps_importable:
+        from gpt_engineer.tools.experimental.experimental_steps import (
+            improve_automatic_file_selection,
+        )
+
         improve_fn = improve_automatic_file_selection
+    elif improve_all_mode:
+        raise ImportError(
+            "The experimental improve_all_mode is selected, but the optional dependencies to use it are not installed. Please run 'poetry install --with experimental'"
+        )
     else:
         improve_fn = improve
 

gpt_engineer/benchmark/__main__.py

@@ -1,5 +1,7 @@
 import importlib
-from typing import Optional, Annotated
+
+from typing import Annotated, Optional
+
 import typer
 
 from langchain.cache import SQLiteCache
@@ -22,7 +24,9 @@ def main(
             help="python file that contains a function called 'default_config_agent'"
         ),
     ],
-    benchmarks: Annotated[str, typer.Argument(help="benchmark name(s) separated by ','")],
+    benchmarks: Annotated[
+        str, typer.Argument(help="benchmark name(s) separated by ','")
+    ],
     task_name: Annotated[
         Optional[str], typer.Argument(help="optional task name in benchmark")
     ] = None,
@@ -38,7 +42,9 @@ def main(
         agent = get_agent(path_to_agent)
 
         results = run(agent, benchmark, task_name, verbose=verbose)
-        print(f"\n--- Results for agent {path_to_agent}, benchmark: {benchmark_name} ---")
+        print(
+            f"\n--- Results for agent {path_to_agent}, benchmark: {benchmark_name} ---"
+        )
         print_results(results)
         print()
 

gpt_engineer/benchmark/benchmarks/gpteng/eval_tools.py

@@ -9,13 +9,6 @@
 be expanded.
 """
 
-import subprocess
-
-from datetime import datetime
-
-import yaml
-
-from tabulate import tabulate
 
 from gpt_engineer.core.files_dict import FilesDict
 

gpt_engineer/benchmark/benchmarks/gpteng/load.py

@@ -1,10 +1,10 @@
 from pathlib import Path
 
-from gpt_engineer.benchmark.benchmarks.gpteng.eval_tools import check_evaluation_component
-from gpt_engineer.benchmark.types import Benchmark, Task, Assertable
-from gpt_engineer.core import chat_to_files
+from gpt_engineer.benchmark.benchmarks.gpteng.eval_tools import (
+    check_evaluation_component,
+)
+from gpt_engineer.benchmark.types import Assertable, Benchmark, Task
 from gpt_engineer.core.chat_to_files import chat_to_files_dict
-from gpt_engineer.core.files_dict import FilesDict
 
 evaluations = [
     {
@@ -156,4 +156,6 @@ def eval_to_task(case):
 
 
 def load_gpteng():
-    return Benchmark(name="gpte_eval", tasks=[eval_to_task(case) for case in evaluations])
+    return Benchmark(
+        name="gpte_eval", tasks=[eval_to_task(case) for case in evaluations]
+    )

gpt_engineer/benchmark/benchmarks/gptme/load.py

@@ -1,6 +1,5 @@
 from gpt_engineer.benchmark.types import Benchmark, Task
 from gpt_engineer.core.files_dict import FilesDict
-from gpt_engineer.core.base_execution_env import BaseExecutionEnv
 
 
 def load_gptme():

gpt_engineer/benchmark/run.py

@@ -1,7 +1,5 @@
 import time
 
-import typer
-
 from gpt_engineer.benchmark.types import Assertable, Benchmark, TaskResult
 from gpt_engineer.core.base_agent import BaseAgent
 from gpt_engineer.core.default.disk_execution_env import DiskExecutionEnv
@@ -67,7 +65,9 @@ def print_results(results: list[TaskResult]):
         )
         for task_result in results
     )
-    total_assertions = sum(len(task_result.assertion_results) for task_result in results)
+    total_assertions = sum(
+        len(task_result.assertion_results) for task_result in results
+    )
     print(f"Total correct assertions: {correct_assertions}/{total_assertions}")
 
     correct_tasks = sum(

gpt_engineer/benchmark/types.py

@@ -2,8 +2,8 @@
 from subprocess import Popen
 from typing import Callable
 
-from gpt_engineer.core.files_dict import FilesDict
 from gpt_engineer.core.base_execution_env import BaseExecutionEnv
+from gpt_engineer.core.files_dict import FilesDict
 
 
 @dataclass

gpt_engineer/core/ai.py

@@ -9,8 +9,6 @@
 import backoff
 import openai
 
-from gpt_engineer.core.token_usage import TokenUsageLog
-
 from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
 from langchain.chat_models import AzureChatOpenAI, ChatOpenAI
 from langchain.chat_models.base import BaseChatModel
@@ -22,6 +20,8 @@
     messages_to_dict,
 )
 
+from gpt_engineer.core.token_usage import TokenUsageLog
+
 # Type hint for a chat message
 Message = Union[AIMessage, HumanMessage, SystemMessage]
 

gpt_engineer/core/base_agent.py

@@ -1,8 +1,7 @@
-from gpt_engineer.core.files_dict import FilesDict
-from gpt_engineer.core.version_manager import BaseVersionManager
-from gpt_engineer.core.ai import AI
 from abc import ABC, abstractmethod
 
+from gpt_engineer.core.files_dict import FilesDict
+
 
 class BaseAgent(ABC):
     """

gpt_engineer/core/base_execution_env.py

@@ -1,7 +1,8 @@
 from abc import ABC, abstractmethod
-from gpt_engineer.core.files_dict import FilesDict
 from subprocess import Popen
 
+from gpt_engineer.core.files_dict import FilesDict
+
 
 class BaseExecutionEnv(ABC):
     """

gpt_engineer/core/base_memory.py

@@ -1,5 +1,4 @@
 from pathlib import Path
-from typing import TypeVar, MutableMapping
-from abc import ABC
+from typing import MutableMapping
 
 BaseMemory = MutableMapping[str | Path, str]

gpt_engineer/core/chat_to_files.py

@@ -26,11 +26,11 @@
 - apply_edits: Applies file edits to a workspace.
 """
 
-import re
 import logging
+import re
 
 from dataclasses import dataclass
-from typing import List, Tuple
+from typing import List
 
 from gpt_engineer.core.files_dict import FilesDict
 

gpt_engineer/core/default/disk_execution_env.py

@@ -2,8 +2,8 @@
 import time
 
 from gpt_engineer.core.base_execution_env import BaseExecutionEnv
-from gpt_engineer.core.files_dict import FilesDict
 from gpt_engineer.core.default.file_store import FileStore
+from gpt_engineer.core.files_dict import FilesDict
 
 
 class DiskExecutionEnv(BaseExecutionEnv):

gpt_engineer/core/default/disk_memory.py

@@ -30,9 +30,10 @@
 import json
 import shutil
 
-from gpt_engineer.core.base_memory import BaseMemory
 from pathlib import Path
-from typing import Any, Optional, Union, Iterator, Dict
+from typing import Any, Dict, Iterator, Optional, Union
+
+from gpt_engineer.core.base_memory import BaseMemory
 from gpt_engineer.tools.experimental.supported_languages import SUPPORTED_LANGUAGES