From 3e86b5d5318d757571076760097d22e3316f4b61 Mon Sep 17 00:00:00 2001
From: Benjamin Pritchard <ben@bennyp.org>
Date: Tue, 7 Jan 2025 22:02:01 -0500
Subject: [PATCH] Add function to convert duration string to int

---
 qcportal/qcportal/test_utils.py | 32 +++++++++++++++++++++++++++++++-
 qcportal/qcportal/utils.py      | 32 ++++++++++++++++++++++++++++++++
 2 files changed, 63 insertions(+), 1 deletion(-)

diff --git a/qcportal/qcportal/test_utils.py b/qcportal/qcportal/test_utils.py
index 4f7cf03fa..337b58d09 100644
--- a/qcportal/qcportal/test_utils.py
+++ b/qcportal/qcportal/test_utils.py
@@ -1,4 +1,4 @@
-from qcportal.utils import chunk_iterable, seconds_to_hms, is_included
+from qcportal.utils import chunk_iterable, seconds_to_hms, duration_to_seconds, is_included
 
 
 def test_chunk_iterable():
@@ -29,6 +29,36 @@ def test_seconds_to_hms():
     assert seconds_to_hms(3670.12) == "01:01:10.12"
 
 
+def test_duration_to_seconds():
+    assert duration_to_seconds(0) == 0
+    assert duration_to_seconds("0") == 0
+    assert duration_to_seconds(17) == 17
+    assert duration_to_seconds("17") == 17
+
+    assert duration_to_seconds("17s") == 17
+    assert duration_to_seconds("70s") == 70
+    assert duration_to_seconds("8m17s") == 497
+    assert duration_to_seconds("80m72s") == 4872
+    assert duration_to_seconds("3h8m17s") == 11297
+    assert duration_to_seconds("03h08m07s") == 11287
+    assert duration_to_seconds("03h08m070s") == 11350
+    assert duration_to_seconds("9d03h08m070s") == 788950
+
+    assert duration_to_seconds("9d") == 777600
+    assert duration_to_seconds("10m") == 600
+    assert duration_to_seconds("90m") == 5400
+    assert duration_to_seconds("04h") == 14400
+    assert duration_to_seconds("4h5s") == 14405
+    assert duration_to_seconds("1d9s") == 86409
+
+    assert duration_to_seconds("8:17") == 497
+    assert duration_to_seconds("80:72") == 4872
+    assert duration_to_seconds("3:8:17") == 11297
+    assert duration_to_seconds("03:08:07") == 11287
+    assert duration_to_seconds("03:08:070") == 11350
+    assert duration_to_seconds("9:03:08:07") == 788950
+
+
 def test_is_included():
     assert is_included("test", None, None, True) is True
     assert is_included("test", None, None, False) is False
diff --git a/qcportal/qcportal/utils.py b/qcportal/qcportal/utils.py
index 89806afd1..1192212e5 100644
--- a/qcportal/qcportal/utils.py
+++ b/qcportal/qcportal/utils.py
@@ -8,6 +8,7 @@
 import json
 import logging
 import math
+import re
 import time
 from contextlib import contextmanager, redirect_stderr, redirect_stdout
 from hashlib import sha256
@@ -261,6 +262,37 @@ def seconds_to_hms(seconds: Union[float, int]) -> str:
         return f"{hours:02d}:{minutes:02d}:{seconds+fraction:02.2f}"
 
 
+def duration_to_seconds(s: Union[int, str]) -> int:
+    """
+    Parses a string in dd:hh:mm:ss or 1d2h3m4s to an integer number of seconds
+    """
+
+    # Is already an int
+    if isinstance(s, int):
+        return s
+
+    # Plain number of seconds (as a string)
+    if s.isdigit():
+        return int(s)
+
+    # Handle dd:hh:mm:ss format
+    if ":" in s:
+        parts = list(map(int, s.split(":")))
+        while len(parts) < 4:  # Pad missing parts with zeros
+            parts.insert(0, 0)
+        days, hours, minutes, seconds = parts
+        return days * 86400 + hours * 3600 + minutes * 60 + seconds
+
+    # Handle format like 3d4h7m10s
+    pattern = re.compile(r"(?:(\d+)d)?(?:(\d+)h)?(?:(\d+)m)?(?:(\d+)s)?")
+    match = pattern.fullmatch(s)
+    if not match:
+        raise ValueError(f"Invalid duration format: {s}")
+
+    days, hours, minutes, seconds = map(lambda x: int(x) if x else 0, match.groups())
+    return days * 86400 + hours * 3600 + minutes * 60 + seconds
+
+
 def recursive_normalizer(value: Any, digits: int = 10, lowercase: bool = True) -> Any:
     """
     Prepare a structure for hashing by lowercasing all values and round all floats