using dotenv with default values properly and

certalized default value logic

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(
-    os.getenv("DATASET_PATH", os.path.join("data", "dataset.json")),
-).build_reference_map()
+REFERENCE_MAP = Dataset.from_json(os.environ["DATASET_PATH"]).build_reference_map()
 
-ARCHIVES_ROOT = os.getenv("ARCHIVES_ROOT", "data/archives")
+ARCHIVES_ROOT = os.environ["ARCHIVES_ROOT"]
 
 
 def comment_distance(submission: CommentGenSubmission, entry: Comment):