work with Upstream-testsuite logs; skip passed beaker logs

The testout.log format used by Upstream-testsuite is slightly different.
Improve performance by only parsing ansible logs from failed
test runs in beaker.

check_logs.py

@@ -234,6 +234,8 @@ def get_statuses(gh, org, repo, pr_num):
         statuses.append(status)
     return statuses
 
+SYSTEM_ROLE_LOG_RE = re.compile(r"/SYSTEM-ROLE-(?P<role>[a-z0-9_]+)_(?P<test_name>tests_[a-z0-9_]+[.]yml)-.*-ANSIBLE-(?P<ansible_ver>[0-9.]+).*[.]log$")
+
 
 def get_errors_from_ansible_log(args, log_url):
     errors = []
@@ -258,11 +260,8 @@ def get_errors_from_ansible_log(args, log_url):
         )
     else:
         # https://....//SYSTEM-ROLE-$ROLENAME_$TEST_NAME.yml-legacy-ANSIBLE-2.log
-        match = re.search(
-            r"/SYSTEM-ROLE-([a-z0-9_]+)_(tests_[a-z0-9_]+[.]yml)-.*-ANSIBLE-([0-9.]+)[.]log$",
-            log_url,
-        )
-        role = match.group(1)
+        match = SYSTEM_ROLE_LOG_RE.search(log_url)
+        role = match.group("role")
         if args.role != ["ALL"] and role not in args.role:
             logging.info(
                 "Skipping log - role [%s] not in args.role [%s]: [%s]",
@@ -272,7 +271,7 @@ def get_errors_from_ansible_log(args, log_url):
             )
             return []
         # test = match.group(2)  # unused for now
-        ansible_version = match.group(3)
+        ansible_version = match.group("ansible_ver")
 
     for line in get_file_data(args, log_url):
         if (
@@ -466,9 +465,11 @@ def parse_beaker_job_log(args, start_dt_utc, taskout_url):
     result_re = re.compile(r"^::   OVERALL RESULT: ([A-Z]+)")
     test_re_str = (
         r"^:: \[ (?P<hour>[0-9]{2}):(?P<min>[0-9]{2}):(?P<sec>[0-9]{2}) \] "
-        r":: \[ +(?P<status>[A-Z]+) +\] :: Test (?P<role>[a-z0-9_]+)/(?P<test_name>[^ ]+)"
+        r":: \[ +(?P<status>[A-Z]+) +\] :: Test ((?P<role>[a-z0-9_]+)/)?(?P<test_name>tests_[^ ]+)"
     )
     test_re = re.compile(test_re_str)
+    test_role_re_str = r"^::\s+Test role: (?P<role>[a-z0-9_]+)"
+    test_role_re = re.compile(test_role_re_str)
     duration_re = re.compile(r"Duration: ([0-9a-zA-Z_]+)")
     start_dt = None
     start_data = None
@@ -479,6 +480,7 @@ def parse_beaker_job_log(args, start_dt_utc, taskout_url):
         "status": "RUNNING",
         "last_test": "N/A",
     }
+    role = None # Upstream-testsuite does not report role name in test_re
     for line in get_file_data(args, taskout_url):
         match = duration_re.search(line)
         if match:
@@ -487,9 +489,17 @@ def parse_beaker_job_log(args, start_dt_utc, taskout_url):
         if match:
             job_data["status"] = match.group(1)
             break
+        match = test_role_re.match(line)
+        if match:
+            role = match.group(1)
         match = test_re.search(line)
         if match:
             data = match.groupdict()
+            if data["role"] is None:
+                data["role"] = role
+            if data["role"] in job_data["roles"] and data["test_name"] in job_data["roles"][data["role"]]:
+                logging.debug("Already processed result for [%s/%s] - skipping [%s]", data["role"], data["test_name"], line)
+                continue
             if not start_dt:
                 # figure out TZ offset and set in start_dt
                 start_dt = start_dt_utc.replace(
@@ -506,7 +516,7 @@ def parse_beaker_job_log(args, start_dt_utc, taskout_url):
                 start_data = data
             elif start_data:
                 btr = BeakerTestRec(start_dt, start_data, data)
-                job_data["roles"].setdefault(data["role"], []).append(btr)
+                job_data["roles"].setdefault(data["role"], {})[data["test_name"]] = btr
                 if data["status"] == "PASS":
                     job_data["passed"].append(btr)
                 else:
@@ -567,8 +577,11 @@ def get_beaker_job_info(args, job):
             if key in ("start_time", "finish_time") and task_data[key]:
                 dt = datetime.datetime.fromisoformat(task_data[key] + "+00:00")
                 task_data[key] = dt
-        if task_data["name"].endswith("basic-smoke-test"):
-            task_data["name"] = "basic-smoke-test"
+        if task_data["name"].endswith("basic-smoke-test") or task_data["name"].endswith("Upstream-testsuite"):
+            if task_data["name"].endswith("basic-smoke-test"):
+                task_data["name"] = "basic-smoke-test"
+            else:
+                task_data["name"] = "Upstream-testsuite"
             log_urls = []
             for log in task.find("logs"):
                 if hasattr(log, "get"):
@@ -580,13 +593,21 @@ def get_beaker_job_info(args, job):
                             args, task_data["start_time"], link
                         )
                     elif name.startswith("SYSTEM-ROLE-"):
-                        if args.gather_errors:
-                            logging.debug("    Processing test log [%s]", link)
-                            task_data["errors"].extend(
-                                get_errors_from_ansible_log(args, link)
-                            )
                         log_urls.append(link)
             task_data["logs"] = log_urls
+            if args.gather_errors:
+                for log in log_urls:
+                    match = SYSTEM_ROLE_LOG_RE.search(log)
+                    if match:
+                        role = match.group("role")
+                        test_name = match.group("test_name")
+                        btr = task_data["job_data"].get("roles", {}).get(role, {}).get(test_name)
+                        if btr and btr.status != "PASS":
+                            logging.debug("    Processing test log [%s]", log)
+                            task_data["errors"].extend(
+                                get_errors_from_ansible_log(args, log)
+                            )
+
             role = None
             task_data["avcs"] = {}
             for result in task.find_all("result"):