Split the system tests log file into stages

.github/workflows/generate_reference_results_workflow.yml

@@ -92,14 +92,17 @@ jobs:
         fi
         git commit -m "${{inputs.commit_msg}}"
         git push
-    - name: Archive main log files
+    - name: Archive system test logs (combined + per-stage)
       if: ${{ always() }}
       uses: actions/upload-artifact@v7
       with:
         name: system_tests_run_${{ github.run_id }}_${{ github.run_attempt }}_reference_logs
         path: |
           runs/*/system-tests-stdout.log
           runs/*/system-tests-stderr.log
+          runs/*/system-tests-build.log
+          runs/*/system-tests-solver.log
+          runs/*/system-tests-fieldcompare.log
           runs/*/*/system-tests_*.log
     - name: Archive run files
       if: failure()

.github/workflows/run_testsuite_workflow.yml

@@ -78,14 +78,17 @@ jobs:
         cd tools/tests
         python systemtests.py --build_args=${{ inputs.build_args}} --suites=${{ inputs.suites}} --log-level=${{ inputs.log_level}}
         cd ../../
-    - name: Archive main log files
+    - name: Archive system test logs (combined + per-stage)
       if: ${{ always() }}
       uses: actions/upload-artifact@v7
       with:
         name: system_tests_run_${{ github.run_id }}_${{ github.run_attempt }}_logs
         path: |
           runs/*/system-tests-stdout.log
           runs/*/system-tests-stderr.log
+          runs/*/system-tests-build.log
+          runs/*/system-tests-solver.log
+          runs/*/system-tests-fieldcompare.log
           runs/*/*/system-tests_*.log
     - name: Archive run files
       if: ${{ failure() || inputs.upload_artifacts == 'TRUE' }}

changelog-entries/801.md

@@ -0,0 +1 @@
+- System tests now write logs during each run (build, solver, fieldcompare), including `system-tests-build.log`, `system-tests-solver.log`, and `system-tests-fieldcompare.log`, so failures can be debugged before a test finishes. Console output is easier to scan in local runs and GitHub Actions (fixes [#790](https://github.com/precice/tutorials/issues/790)).

tools/tests/systemtests.py

@@ -7,9 +7,20 @@
 from metadata_parser.metdata import Tutorials, Case
 import logging
 import time
+import os
+import sys
 from paths import PRECICE_TUTORIAL_DIR, PRECICE_TESTS_RUN_DIR, PRECICE_TESTS_DIR
 
 
+class _ConsoleLogFormatter(logging.Formatter):
+    """Omit level prefix for INFO/DEBUG; keep it for warnings and errors."""
+
+    def format(self, record: logging.LogRecord) -> str:
+        if record.levelno >= logging.WARNING:
+            return f"{record.levelname}: {record.getMessage()}"
+        return record.getMessage()
+
+
 def main():
     parser = argparse.ArgumentParser(description='systemtest')
 
@@ -29,11 +40,30 @@ def main():
     # Parse the command-line arguments
     args = parser.parse_args()
 
-    # Configure logging based on the provided log level
-    logging.basicConfig(level=args.log_level, format='%(levelname)s: %(message)s')
+    # Configure logging: no "INFO:" prefix on routine messages (#790)
+    handler = logging.StreamHandler()
+    handler.setFormatter(_ConsoleLogFormatter())
+    logging.basicConfig(level=args.log_level, handlers=[handler])
 
     print(f"Using log-level: {args.log_level}")
 
+    gh_actions = os.environ.get("GITHUB_ACTIONS", "").lower() == "true"
+    ansi_colors = sys.stdout.isatty() and os.environ.get("TERM", "") not in {"", "dumb"}
+
+    def _style(text: str, color_code: int | None) -> str:
+        if not ansi_colors or color_code is None:
+            return text
+        return f"\x1b[{color_code}m{text}\x1b[0m"
+
+    def _group_start(title: str) -> None:
+        # Only apply folding markers on GitHub actions.
+        if gh_actions:
+            print(f"::group::{title}")
+
+    def _group_end() -> None:
+        if gh_actions:
+            print("::endgroup::")
+
     systemtests_to_run = []
     available_tutorials = Tutorials.from_path(PRECICE_TUTORIAL_DIR)
 
@@ -72,17 +102,34 @@ def main():
     if not systemtests_to_run:
         raise RuntimeError("Did not find any Systemtests to execute.")
 
-    logging.info(f"About to run the following systemtest in the directory {run_directory}:\n {systemtests_to_run}")
+    total = len(systemtests_to_run)
+    logging.info(
+        f"About to run {total} systemtest(s) in the directory {run_directory}:\n {systemtests_to_run}")
 
     results = []
     for number, systemtest in enumerate(systemtests_to_run, start=1):
-        logging.info(f"Started running {systemtest},  {number}/{len(systemtests_to_run)}")
+        print()
+        test_header = f"[{number}/{total}] {systemtest}"
+        _group_start(test_header)
+        print("=" * 80)
+        logging.info(f"[{number}/{total}] Started {systemtest}")
         t = time.perf_counter()
         result = systemtest.run(run_directory)
         elapsed_time = time.perf_counter() - t
-        logging.info(f"Running {systemtest} took {elapsed_time:^.1f} seconds")
+
+        if result.success:
+            status_label = _style("✅ PASS", 32)
+        else:
+            status_label = _style("❌ FAIL", 31)
+
+        logging.info(
+            f"[{number}/{total}] Finished {systemtest} in {elapsed_time:.1f}s [{status_label}]"
+        )
+        _group_end()
+        print("=" * 80)
         results.append(result)
 
+    print()
     system_test_success = True
     for result in results:
         if not result.success: