Merge branch 'main' of github.com:lilab-bcb/altocumulus

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.5.0
+    rev: v4.6.0
     hooks:
       # - id: double-quote-string-fixer  # for single quotes: uncomment and add black config “skip-string-normalization”
       - id: trailing-whitespace
@@ -11,14 +11,14 @@ repos:
       - id: docformatter
         args: ["--in-place", "--wrap-summaries=100", "--wrap-descriptions=100", "--config=./pyproject.toml"]
   - repo: https://github.com/psf/black
-    rev: 23.12.1
+    rev: 24.4.2
     hooks:
       - id: black
   - repo: https://github.com/pycqa/isort
     rev: 5.13.2
     hooks:
       - id: isort
   - repo: https://github.com/csachs/pyproject-flake8
-    rev: v6.1.0
+    rev: v7.0.0
     hooks:
       - id: pyproject-flake8

alto/commands/parse_monitoring_log.py

@@ -59,13 +59,12 @@ def execute(input_path, report_filename, plot_filename=None):
     results = []
     for i in range(len(log_paths)):
         log_path = log_paths[i]
-
-        task, shard = get_task_and_shard(log_path) if is_dir else None, None
         result = parse_log(
             scheme + "://" + log_path,
             details=generate_plot,
         )
-        if task is not None:
+        if is_dir:
+            task, shard = get_task_and_shard(log_path)
             result["task"] = task
             result["shard"] = shard
 
@@ -88,21 +87,23 @@ def _figsize(nrow=1, ncol=1, aspect=1, size=3):
 def get_task_and_shard(log_path):
     p = Path(log_path)
     shard_name = p.parent.name
-    if shard_name == "cacheCopy" or shard_name.startswith("attempt-") or shard_name == "execution":
+    if shard_name.startswith("attempt-") or shard_name in ("execution", "cacheCopy", "work"):
         p = p.parent
         shard_name = p.parent.name
 
-    if shard_name.startswith("shard-"):
+    if shard_name.startswith("shard-") or shard_name.isdigit():
         task_name = p.parent.parent.name
-
     else:
         task_name = shard_name
         shard_name = ""
 
     if task_name.startswith("call-"):
         task_name = task_name[len("call-") :]
+    elif task_name.endswith("_monitoring_log"):
+        task_name = task_name[: -len("_monitoring_log")]
     # gs://output/cromwell_execution/xxx_workflow/92f48dc5-6/call-xxx_task/shard-0/monitoring.log
     # gs://output/cromwell_execution/xxx_workflow/92f48dc5-6/call-xxx_task/shard-0/cacheCopy/monitoring.log
+    # aws-omics: task_monitoring_log/0/monitoring.log
     return task_name, shard_name