using dotenv with default values properly and

certalized default value logic

src/routes/answers.py

@@ -5,8 +5,7 @@ from utils.dataset import CommentGenSubmission
 from utils.errors import InvalidJsonFormatError
 from utils.process_data import evaluate_comments, evaluate_refinement
 from utils.observer import SocketObserver, Status, Subject
-import functools
+import functools, json, os
-import json
 
 from utils.queue_manager import QueueManager
 
@@ -56,7 +55,7 @@ def validate_json_format_for_code_refinement(data: str) -> dict[str, dict[str, s
         raise InvalidJsonFormatError()
 
 
-QUEUE_MANAGER = QueueManager(1)
+QUEUE_MANAGER = QueueManager(int(os.environ["MAX_WORKERS"]))
 
 
 def handler(type_: str, validate_json: Callable, evaluate_submission: Callable):

src/routes/datasets.py

@@ -8,7 +8,7 @@ DATASETS = {'comment_generation', 'code_refinement'}
 
 # below, the '../' + is need because the send_from send_from_directory is local
 # the file, but the DATA_DIR is local to the root of the repo
-DATA_DIR = os.path.join('..', os.getenv("DATA_PATH", "data"))
+DATA_DIR = os.path.join('..', os.environ["DATA_PATH"])
 
 
 @router.route('/download/<dataset>')

src/server.py

@@ -1,4 +1,9 @@
-# server.py
+from utils.env_defaults import set_env_defaults
+from dotenv import load_dotenv
+
+set_env_defaults()
+load_dotenv()
+
 from flask import Flask, request
 from flask_cors import CORS
 from flask_socketio import SocketIO
@@ -9,10 +14,6 @@ from routes.datasets import router as datasets_router
 from werkzeug.exceptions import HTTPException
 import os
 
-from dotenv import load_dotenv
-
-load_dotenv()
-
 app = Flask(__name__, static_folder='../public', static_url_path='/')
 
 with app.app_context():
@@ -79,7 +80,7 @@ def init_socketio(app):
 socketio = init_socketio(app)
 
 if __name__ == '__main__':
-    port = int(os.getenv('PORT', 45003))
+    port = int(os.environ['PORT'])
     socketio.run(
         app,
         use_reloader=True,

src/utils/build_handlers.py

@@ -637,7 +637,7 @@ def get_build_handler(root: str, repo: str, verbose: bool = False) -> BuildHandl
         an instance of GradleHandler or MavenHandler
     """
     path = os.path.join(root, repo)
-    if os.getenv('MOCK_BUILD_HANDLER', False):
+    if bool(os.environ["MOCK_BUILD_HANDLER"]):
         return WaitBuildHander("NO REPO PATH", "NO BUILD FILE", {})
 
     # 1) If it's a tarball, extract it

src/utils/env_defaults.py

@@ -0,0 +1,16 @@
+import os
+from typing import Any
+
+
+def set(k: str, v: Any):
+    os.environ.setdefault(k, str(v))
+
+
+def set_env_defaults():
+    set("PORT", 45003)
+    set("MAX_WORKERS", 5)
+    set("RESULTS_DIR", "submission_results")
+    set("MOCK_BUILD_HANDLER", False)
+    set("DATA_PATH", "data")
+    set("DATASET_PATH", os.path.join(os.environ["DATA_PATH"], "dataset.json"))
+    set("ARCHIVES_ROOT", os.path.join(os.environ["DATA_PATH"], "archives"))

src/utils/observer.py

@@ -4,7 +4,7 @@ from enum import Enum
 import os, tempfile, threading, json
 from typing import Callable, Optional, Set
 
-RESULTS_DIR = os.getenv("RESULTS_DIR", "submission_results")
+RESULTS_DIR = os.environ["RESULTS_DIR"]
 
 
 class Status(Enum):

src/utils/process_data.py

@@ -5,11 +5,9 @@ from utils.build_handlers import get_build_handler
 from sacrebleu import sentence_bleu as bleu
 from utils.dataset import ArchiveState, Comment, CommentGenSubmission, Dataset
 
-REFERENCE_MAP = Dataset.from_json(
+REFERENCE_MAP = Dataset.from_json(os.environ["DATASET_PATH"]).build_reference_map()
-    os.getenv("DATASET_PATH", os.path.join("data", "dataset.json")),
-).build_reference_map()
 
-ARCHIVES_ROOT = os.getenv("ARCHIVES_ROOT", "data/archives")
+ARCHIVES_ROOT = os.environ["ARCHIVES_ROOT"]
 
 
 def comment_distance(submission: CommentGenSubmission, entry: Comment):