Merge pull request #1147 from BioGeek/fix-datapoint-id-collision

Give generated datapoint IDs sub-second resolution

Author: rodvrees

proteobench/datapoint/denovo_datapoint.py

@@ -197,9 +197,12 @@ def generate_id(self) -> None:
         """
         Generate a unique ID for the benchmark run by combining the software name and a timestamp.
 
-        This ID is used to uniquely identify each run of the benchmark.
+        This ID is used to uniquely identify each run of the benchmark. The timestamp carries
+        microseconds because datapoints are also generated in batches -- the resubmission
+        script regenerates many in a loop -- where whole-second resolution hands several runs
+        the same ID.
         """
-        time_stamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+        time_stamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
         self.id = "_".join([self.software_name, str(time_stamp)])
         logging.info(f"Assigned the following ID to this run: {self.id}")
 

proteobench/datapoint/entrapment_datapoint.py

@@ -89,9 +89,12 @@ def generate_id(self) -> None:
         """
         Generate a unique ID for the benchmark run by combining the software name and a timestamp.
 
-        This ID is used to uniquely identify each run of the benchmark.
+        This ID is used to uniquely identify each run of the benchmark. The timestamp carries
+        microseconds because datapoints are also generated in batches -- the resubmission
+        script regenerates many in a loop -- where whole-second resolution hands several runs
+        the same ID.
         """
-        time_stamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+        time_stamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
         self.id = "_".join([self.software_name, str(time_stamp)])
         logging.info(f"Assigned the following ID to this run: {self.id}")
 

proteobench/datapoint/quant_datapoint.py

@@ -327,9 +327,12 @@ def generate_id(self) -> None:
         """
         Generate a unique ID for the benchmark run by combining the software name and a timestamp.
 
-        This ID is used to uniquely identify each run of the benchmark.
+        This ID is used to uniquely identify each run of the benchmark. The timestamp carries
+        microseconds because datapoints are also generated in batches -- the resubmission
+        script regenerates many in a loop -- where whole-second resolution hands several runs
+        the same ID.
         """
-        time_stamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+        time_stamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
         self.id = "_".join([self.software_name, str(time_stamp)])
         logging.info(f"Assigned the following ID to this run: {self.id}")
 

test/test_datapoint_id.py

@@ -0,0 +1,50 @@
+"""Tests that generated datapoint IDs stay unique when datapoints are created in a batch."""
+
+import re
+
+import pytest
+
+from proteobench.datapoint.denovo_datapoint import DenovoDatapoint
+from proteobench.datapoint.entrapment_datapoint import EntrapmentDatapoint
+from proteobench.datapoint.quant_datapoint import QuantDatapointHYE
+
+DATAPOINT_CLASSES = [QuantDatapointHYE, DenovoDatapoint, EntrapmentDatapoint]
+
+# software name, then date_time_microseconds
+ID_PATTERN = re.compile(r"^(?P<software>.+)_\d{8}_\d{6}_\d{6}$")
+
+
+@pytest.mark.parametrize("datapoint_class", DATAPOINT_CLASSES)
+def test_generated_id_starts_with_the_software_name(datapoint_class):
+    datapoint = datapoint_class(software_name="TestTool")
+    datapoint.generate_id()
+    assert datapoint.id.startswith("TestTool_")
+
+
+@pytest.mark.parametrize("datapoint_class", DATAPOINT_CLASSES)
+def test_generated_id_carries_sub_second_resolution(datapoint_class):
+    datapoint = datapoint_class(software_name="TestTool")
+    datapoint.generate_id()
+    match = ID_PATTERN.match(datapoint.id)
+    assert match is not None, f"unexpected ID format: {datapoint.id}"
+    assert match.group("software") == "TestTool"
+
+
+@pytest.mark.parametrize("datapoint_class", DATAPOINT_CLASSES)
+def test_ids_generated_in_a_tight_loop_are_unique(datapoint_class):
+    """A batch (e.g. the resubmission script) must not hand several runs the same ID."""
+    ids = []
+    for _ in range(50):
+        datapoint = datapoint_class(software_name="TestTool")
+        datapoint.generate_id()
+        ids.append(datapoint.id)
+    assert len(set(ids)) == len(ids)
+
+
+def test_ids_stay_distinct_across_datapoint_types():
+    ids = []
+    for datapoint_class in DATAPOINT_CLASSES:
+        datapoint = datapoint_class(software_name="TestTool")
+        datapoint.generate_id()
+        ids.append(datapoint.id)
+    assert len(set(ids)) == len(ids)