@@ -10,6 +10,21 @@ if ! command -v jq >/dev/null 2>&1; then
1010 exit 1
1111fi
1212
13+ if ! command -v python3 > /dev/null 2>&1 ; then
14+ echo " python3 is required to process Hive client logs but was not found in PATH"
15+ exit 1
16+ fi
17+
18+ slugify () {
19+ local input=" ${1:- } "
20+ local lowered trimmed
21+ lowered=" $( printf ' %s' " ${input} " | tr ' [:upper:]' ' [:lower:]' ) "
22+ trimmed=" $( printf ' %s' " ${lowered} " | sed -E ' s/[^a-z0-9._-]+/-/g' ) "
23+ trimmed=" ${trimmed# -} "
24+ trimmed=" ${trimmed% -} "
25+ printf ' %s' " ${trimmed} "
26+ }
27+
1328results_dir=" ${1:- src/ results} "
1429
1530if [ ! -d " $results_dir " ; then
@@ -86,15 +101,13 @@ for json_file in "${json_files[@]}"; do
86101 {
87102 echo " ### Hive failures: ${suite_name:- $(basename " ${json_file} " "
88103 printf ' %s\n' " ${failure_list} "
104+ echo " Note: Hive scenarios may include multiple ethrex clients, so each failing case can have more than one log snippet."
89105 echo
90106 } >> " ${GITHUB_STEP_SUMMARY} "
91107 fi
92108
93109 suite_slug_raw=" ${suite_name:- $(basename " ${json_file} " "
94- suite_slug=" $( printf ' %s' " ${suite_slug_raw} " | tr ' [:upper:]' ' [:lower:]' ) "
95- suite_slug=" $( printf ' %s' " ${suite_slug} " | sed -E ' s/[^a-z0-9._-]+/-/g' ) "
96- suite_slug=" ${suite_slug# -} "
97- suite_slug=" ${suite_slug% -} "
110+ suite_slug=" $( slugify " ${suite_slug_raw} " ) "
98111 suite_dir=" ${failed_logs_root} /${suite_slug:- suite} "
99112 mkdir -p " ${suite_dir} "
100113
@@ -178,6 +191,198 @@ for json_file in "${json_files[@]}"; do
178191 done <<< " ${suite_logs_output}"
179192 fi
180193
194+ client_case_entries=" $(
195+ jq -r '
196+ .testCases
197+ | to_entries[]
198+ | select(.value.summaryResult.pass != true)
199+ | . as $case_entry
200+ | ($case_entry.value.clientInfo? // {}) | to_entries[]
201+ | [
202+ .value.logFile // "",
203+ ($case_entry.value.name // ("case-" + $case_entry.key)),
204+ $case_entry.key,
205+ ($case_entry.value.start // ""),
206+ ($case_entry.value.end // ""),
207+ .key
208+ ]
209+ | @tsv
210+ ' " ${json_file} " 2> /dev/null || true
211+ ) "
212+ generated_client_snippets=0
213+ if [ -n " ${client_case_entries} " ; then
214+ client_logs_dir=" ${suite_dir} /client_logs"
215+ mkdir -p " ${client_logs_dir} "
216+
217+ while IFS= read -r client_entry; do
218+ [ -n " ${client_entry} " || continue
219+ IFS=$' \t ' read -r client_log_rel raw_case_name case_id case_start case_end client_id <<< " ${client_entry}"
220+
221+ if [ -z " ${client_log_rel} " || [ -z " ${case_start} " || [ -z " ${case_end} " ; then
222+ continue
223+ fi
224+
225+ log_copy_path=" ${suite_dir} /${client_log_rel} "
226+ if [ ! -f " ${log_copy_path} " ; then
227+ continue
228+ fi
229+
230+ case_slug=" $( slugify " ${raw_case_name} " ) "
231+ if [ -n " ${case_slug} " ; then
232+ case_slug=" ${case_slug} -case-${case_id} "
233+ else
234+ case_slug=" case-${case_id} "
235+ fi
236+
237+ client_slug=" $( slugify " ${client_id} " ) "
238+ if [ -z " ${client_slug} " ; then
239+ client_slug=" client"
240+ fi
241+
242+ case_dir=" ${client_logs_dir} /${case_slug} "
243+ mkdir -p " ${case_dir} "
244+ snippet_path=" ${case_dir} /${client_slug} .log"
245+
246+ python3 - " ${log_copy_path} " " ${snippet_path} " " ${raw_case_name} " " ${case_start} " " ${case_end} " " ${client_id} " " ${client_log_rel} " << 'PY '
247+ import sys
248+ from datetime import datetime, timedelta
249+ from pathlib import Path
250+
251+ FORMATS = ("%Y-%m-%dT%H:%M:%S.%fZ", "%Y-%m-%dT%H:%M:%SZ")
252+ CONTEXT_SECONDS = 2
253+ PREFETCH_LIMIT = 50
254+
255+ def normalise_timestamp_str(value):
256+ if not value or not value.endswith("Z"):
257+ return value
258+ prefix = value[:-1]
259+ if "." not in prefix:
260+ return value
261+ base, frac = prefix.split(".", 1)
262+ frac_digits = "".join(ch for ch in frac if ch.isdigit())
263+ if not frac_digits:
264+ return f"{base}.000000Z"
265+ frac_digits = (frac_digits + "000000")[:6]
266+ return f"{base}.{frac_digits}Z"
267+
268+ def parse_timestamp(value):
269+ if not value:
270+ return None
271+ value = normalise_timestamp_str(value)
272+ for fmt in FORMATS:
273+ try:
274+ return datetime.strptime(value, fmt)
275+ except ValueError:
276+ continue
277+ return None
278+
279+ def timestamp_from_line(line):
280+ if not line:
281+ return None
282+ token = line.split(" ", 1)[0]
283+ if not token or not token[0].isdigit():
284+ return None
285+ token = normalise_timestamp_str(token)
286+ for fmt in FORMATS:
287+ try:
288+ return datetime.strptime(token, fmt)
289+ except ValueError:
290+ continue
291+ return None
292+
293+ log_path = Path(sys.argv[1])
294+ output_path = Path(sys.argv[2])
295+ case_name = sys.argv[3]
296+ case_start_raw = sys.argv[4]
297+ case_end_raw = sys.argv[5]
298+ client_id = sys.argv[6] or "unknown"
299+ client_log_rel = sys.argv[7]
300+
301+ try:
302+ log_content = log_path.read_text(encoding="utf-8", errors="replace").splitlines(keepends=True)
303+ except Exception as exc:
304+ output_path.parent.mkdir(parents=True, exist_ok=True)
305+ output_path.write_text(f"# Failed to read log '{log_path}': {exc}\n", encoding="utf-8")
306+ sys.exit(0)
307+
308+ start_ts = parse_timestamp(case_start_raw)
309+ end_ts = parse_timestamp(case_end_raw)
310+
311+ fallback_reason = None
312+ if not start_ts or not end_ts or end_ts < start_ts:
313+ fallback_reason = "Unable to determine reliable time window from test metadata."
314+ else:
315+ start_ts = start_ts - timedelta(seconds=CONTEXT_SECONDS)
316+ end_ts = end_ts + timedelta(seconds=CONTEXT_SECONDS)
317+
318+ captured_lines = []
319+ prefetch = []
320+ current_ts = None
321+ capturing = False
322+
323+ if not fallback_reason:
324+ for line in log_content:
325+ ts = timestamp_from_line(line)
326+ if ts is not None:
327+ current_ts = ts
328+
329+ if not capturing:
330+ prefetch.append(line)
331+ if len(prefetch) > PREFETCH_LIMIT:
332+ prefetch.pop(0)
333+
334+ in_window = current_ts is not None and start_ts <= current_ts <= end_ts
335+
336+ if in_window:
337+ if not capturing:
338+ captured_lines.extend(prefetch)
339+ capturing = True
340+ captured_lines.append(line)
341+ elif capturing and current_ts is not None and current_ts > end_ts:
342+ break
343+ elif capturing:
344+ captured_lines.append(line)
345+
346+ if not captured_lines:
347+ fallback_reason = "No timestamped log lines matched the computed time window."
348+
349+ if fallback_reason:
350+ captured_lines = log_content
351+
352+ header_lines = [
353+ f"# Test: {case_name}\n",
354+ f"# Client ID: {client_id}\n",
355+ f"# Source log: {client_log_rel}\n",
356+ ]
357+
358+ if start_ts and end_ts and not fallback_reason:
359+ header_lines.append(
360+ f"# Time window (UTC): {case_start_raw} .. {case_end_raw} (with ±{CONTEXT_SECONDS}s context)\n"
361+ )
362+ else:
363+ header_lines.append("# Time window (UTC): unavailable\n")
364+
365+ if fallback_reason:
366+ header_lines.append(f"# NOTE: {fallback_reason}\n")
367+
368+ header_lines.append("\n")
369+
370+ output_path.parent.mkdir(parents=True, exist_ok=True)
371+ with output_path.open("w", encoding="utf-8") as dst:
372+ dst.writelines(header_lines)
373+ dst.writelines(captured_lines)
374+ PY
375+
376+ if [ -s " ${snippet_path} " ; then
377+ generated_client_snippets=$(( generated_client_snippets + 1 ))
378+ fi
379+ done <<< " ${client_case_entries}"
380+ fi
381+
382+ if [ " ${generated_client_snippets} " -gt 0 ]; then
383+ echo " Generated ${generated_client_snippets} client log snippet(s) in ${client_logs_dir} "
384+ fi
385+
181386 echo " Saved Hive failure artifacts to ${suite_dir} "
182387
183388 failures=$(( failures + failed_cases))
0 commit comments