fix: show partial-credit answers as yellow ~ in progress bar, not red incorrect

src/benchmark_runner.py

@@ -429,7 +429,7 @@ def process_question(
 def log_question_result(result_data: Dict[str, Any], prefix: str = "") -> str:
     """Logs a color-coded result for a single question.
 
-    Returns a category string: 'correct', 'incorrect', 'skipped', 'parse_fail', or 'api_fail'.
+    Returns a category string: 'correct', 'partial', 'incorrect', 'skipped', 'parse_fail', or 'api_fail'.
     """
     question_id = result_data["question_id"]
     log_message_prefix = f"{prefix}Question {question_id}:"
@@ -451,6 +451,7 @@ def log_question_result(result_data: Dict[str, Any], prefix: str = "") -> str:
     evaluation_status_value = result_data.get("evaluation_status")
 
     is_considered_correct = False
+    is_partial = False
     log_display_status = "N/A"
     status_check_string = ""
 
@@ -461,8 +462,10 @@ def log_question_result(result_data: Dict[str, Any], prefix: str = "") -> str:
     elif isinstance(evaluation_status_value, str):
         log_display_status = evaluation_status_value
         status_check_string = evaluation_status_value.strip().upper()
-        if "CORRECT" in status_check_string:
+        if status_check_string.startswith("CORRECT"):
             is_considered_correct = True
+        elif status_check_string.startswith("PARTIAL_"):
+            is_partial = True
     elif evaluation_status_value is None:
         log_display_status = "None"
         status_check_string = "NONE_STATUS"
@@ -475,6 +478,9 @@ def log_question_result(result_data: Dict[str, Any], prefix: str = "") -> str:
     if is_considered_correct:
         print(f"{GREEN}{log_message_prefix} Correct - Marks: {marks_awarded}, Status: {log_display_status} {log_message_suffix}{RESET}")
         return "correct"
+    elif is_partial:
+        print(f"{YELLOW}{log_message_prefix} Partial - Marks: {marks_awarded}, Status: {log_display_status} {log_message_suffix}{RESET}")
+        return "partial"
     elif status_check_string in known_eval_skip_statuses:
         print(f"{YELLOW}{log_message_prefix} Skipped by Eval - Marks: {marks_awarded}, Status: {log_display_status} {log_message_suffix}{RESET}")
         return "skipped"
@@ -852,6 +858,7 @@ def run_benchmark(
 
         # Counters for tqdm postfix
         initial_correct_count = 0
+        initial_partial_count = 0
         initial_incorrect_count = 0
         initial_skipped_count = 0
         initial_parse_fail_count = 0
@@ -876,6 +883,8 @@ def run_benchmark(
 
             if category == "correct":
                 initial_correct_count += 1
+            elif category == "partial":
+                initial_partial_count += 1
             elif category == "incorrect":
                 initial_incorrect_count += 1
             elif category == "skipped":
@@ -884,7 +893,7 @@ def run_benchmark(
                 initial_parse_fail_count += 1
             elif category == "api_fail":
                 initial_api_fail_count += 1
-            pbar_initial.set_postfix_str(f"✓:{initial_correct_count} ✗:{initial_incorrect_count} S:{initial_skipped_count} P!:{initial_parse_fail_count} A!:{initial_api_fail_count}")
+            pbar_initial.set_postfix_str(f"✓:{initial_correct_count} ~:{initial_partial_count} ✗:{initial_incorrect_count} S:{initial_skipped_count} P!:{initial_parse_fail_count} A!:{initial_api_fail_count}")
 
         pbar_initial.close()
 
@@ -893,6 +902,7 @@ def run_benchmark(
             logging.info(f"--- Retrying {len(failed_questions_data)} questions with initial API failures for model: {model_id} ---")
 
             retry_correct_count = 0
+            retry_partial_count = 0
             retry_incorrect_count = 0
             retry_skipped_count = 0
             retry_parse_fail_count = 0
@@ -938,6 +948,8 @@ def run_benchmark(
 
                 if category == "correct":
                     retry_correct_count += 1
+                elif category == "partial":
+                    retry_partial_count += 1
                 elif category == "incorrect":
                     retry_incorrect_count += 1
                 elif category == "skipped":
@@ -946,7 +958,7 @@ def run_benchmark(
                     retry_parse_fail_count += 1
                 elif category == "api_fail":
                     retry_api_fail_count += 1
-                pbar_retry.set_postfix_str(f"✓:{retry_correct_count} ✗:{retry_incorrect_count} S:{retry_skipped_count} P!:{retry_parse_fail_count} A!:{retry_api_fail_count}")
+                pbar_retry.set_postfix_str(f"✓:{retry_correct_count} ~:{retry_partial_count} ✗:{retry_incorrect_count} S:{retry_skipped_count} P!:{retry_parse_fail_count} A!:{retry_api_fail_count}")
             pbar_retry.close()
 
         # --- Final Evaluation for the current model ---

tests/test_benchmark_runner.py

@@ -0,0 +1,52 @@
+"""Tests for src/benchmark_runner.py — pure functions only."""
+
+import sys, os
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'src'))
+
+import pytest
+from benchmark_runner import log_question_result
+
+
+def _result(evaluation_status, marks_awarded, api_ok=True, parse_ok=True, pred=["A"]):
+    return {
+        "question_id": "TEST001",
+        "api_call_successful": api_ok,
+        "parse_successful": parse_ok,
+        "predicted_answer": pred,
+        "marks_awarded": marks_awarded,
+        "evaluation_status": evaluation_status,
+        "attempt": 1,
+    }
+
+
+def test_correct_returns_correct():
+    assert log_question_result(_result("correct", 4)) == "correct"
+
+
+def test_incorrect_returns_incorrect():
+    assert log_question_result(_result("incorrect", -1)) == "incorrect"
+
+
+def test_partial_3_of_4_returns_partial():
+    assert log_question_result(_result("partial_3_of_4", 3)) == "partial"
+
+
+def test_partial_2_of_3_plus_returns_partial():
+    assert log_question_result(_result("partial_2_of_3_plus", 2)) == "partial"
+
+
+def test_partial_1_of_2_plus_returns_partial():
+    assert log_question_result(_result("partial_1_of_2_plus", 1)) == "partial"
+
+
+def test_skip_returns_skipped():
+    assert log_question_result(_result("skipped", 0, pred="SKIP")) == "skipped"
+
+
+def test_api_fail_returns_api_fail():
+    assert log_question_result(_result("failure_api_or_parse", 0, api_ok=False, parse_ok=False, pred=None)) == "api_fail"
+
+
+def test_parse_fail_returns_parse_fail():
+    r = _result("failure_api_or_parse", 0, api_ok=True, parse_ok=False, pred=None)
+    assert log_question_result(r) == "parse_fail"