Suggested method for parsing NextFlow time fields

tests/test_parse_time_field.py

@@ -0,0 +1,57 @@
+import pytest
+import re
+from datetime import timedelta
+
+
+TIMEDELTA_NAMES = {
+    "w": "weeks",
+    "d": "days",
+    "h": "hours",
+    "m": "minutes",
+    "s": "seconds",
+    "ms": "milliseconds",
+    "us": "microseconds",
+}
+
+
+def parse_time_field(txt: str) -> float:
+    """
+    Parses a time field from a NextFlow report and returns time in seconds
+    """
+    total_td = None
+    for m in re.finditer(r"\b([\+\-\d\.e]+)([a-df-z]+)\b", txt):
+        number, unit = m.groups()
+        td_name = TIMEDELTA_NAMES.get(unit)
+        if not td_name:
+            msg = f"Unexpected time unit '{unit}' in time field '{txt}'"
+            raise ValueError(msg)
+        td = timedelta(**{td_name: float(number)})
+        if total_td:
+            total_td += td
+        else:
+            total_td = td
+
+    if total_td is None:
+        msg = f"Failed to parse time from field '{txt}'"
+        raise ValueError(msg)
+
+    return total_td.total_seconds()
+
+
+def test_parse_time_field():
+    field_seconds = {
+        "0s": 0.0,
+        "1m": 60.0,
+        "1ms": 0.001,
+        "3.7s": 3.7,
+        "5.0e-3s": 0.005,
+        "3m 4s": 184.0,
+        "2d 4h 36ms": 187200.036,
+        "4us 1m 1m": 120.000004,
+    }
+    for txt, secs in field_seconds.items():
+        assert parse_time_field(txt) == secs
+    with pytest.raises(ValueError, match=r"Unexpected time unit"):
+        parse_time_field("3y")
+    with pytest.raises(ValueError, match=r"Failed to parse time from field"):
+        parse_time_field("")