From c1e79c22fb83330eaadf81ad07a5cb3b5d2e8f27 Mon Sep 17 00:00:00 2001
From: Ben Webb <ben@salilab.org>
Date: Tue, 5 Dec 2023 18:08:57 -0800
Subject: [PATCH] Allow duplicate datasets in input

Even if two Datasets that are read in have
exactly the same information, do not treat them
as equal. This preserves more of the input file
structure (rather than combining identical datasets
into a single ID). Closes #127.
---
 ihm/dataset.py       | 12 ++++++++----
 ihm/reader.py        |  1 +
 test/test_dataset.py | 11 +++++++++++
 test/test_reader.py  |  3 +++
 4 files changed, 23 insertions(+), 4 deletions(-)

diff --git a/ihm/dataset.py b/ihm/dataset.py
index c1504b4..3352af4 100644
--- a/ihm/dataset.py
+++ b/ihm/dataset.py
@@ -18,12 +18,16 @@ class Dataset(object):
     """
 
     _eq_keys = ['location']
+    _allow_duplicates = False
 
-    # Datasets compare equal iff they are the same class and have the
-    # same attributes
+    # Datasets compare equal iff they are the same class, have the
+    # same attributes, and allow_duplicates=False
     def _eq_vals(self):
-        return tuple([self.__class__]
-                     + [getattr(self, x) for x in self._eq_keys])
+        if self._allow_duplicates:
+            return id(self)
+        else:
+            return tuple([self.__class__]
+                         + [getattr(self, x) for x in self._eq_keys])
 
     def __eq__(self, other):
         return self._eq_vals() == other._eq_vals()
diff --git a/ihm/reader.py b/ihm/reader.py
index f2f2807..c0333cb 100644
--- a/ihm/reader.py
+++ b/ihm/reader.py
@@ -1384,6 +1384,7 @@ def __call__(self, data_type, id, details):
         f = self.sysr.datasets.get_by_id(
             id, self.type_map.get(typ, ihm.dataset.Dataset))
         f.details = details
+        f._allow_duplicates = True
 
 
 class _DatasetGroupHandler(Handler):
diff --git a/test/test_dataset.py b/test/test_dataset.py
index 43bd489..9b461c0 100644
--- a/test/test_dataset.py
+++ b/test/test_dataset.py
@@ -40,6 +40,17 @@ def test_dataset(self):
         d.parents.append(td)
         self.assertEqual(len(d.parents), 2)
 
+    def test_dataset_allow_duplicates(self):
+        """Test Dataset base class with allow_duplicates=True"""
+        loc = ihm.location.PDBLocation('1abc', version='foo', details='bar')
+        d = ihm.dataset.Dataset(loc)
+        d._allow_duplicates = True
+        self.assertEqual(d._eq_vals(), id(d))
+
+        d2 = ihm.dataset.Dataset(loc)
+        d2._allow_duplicates = True
+        self.assertNotEqual(d, d2)
+
     def test_add_primary_no_parents(self):
         """Test add_primary() method, no parents"""
         l1 = ihm.location.PDBLocation('1abc', version='foo', details='bar')
diff --git a/test/test_reader.py b/test/test_reader.py
index 7d9c14f..cf0c81a 100644
--- a/test/test_reader.py
+++ b/test/test_reader.py
@@ -971,6 +971,9 @@ def test_dataset_list_handler(self):
             self.assertEqual(d1.__class__, ihm.dataset.PDBDataset)
             self.assertEqual(d2.__class__, ihm.dataset.ComparativeModelDataset)
             self.assertEqual(d3.__class__, ihm.dataset.EMMicrographsDataset)
+            self.assertTrue(d1._allow_duplicates)
+            self.assertTrue(d2._allow_duplicates)
+            self.assertTrue(d3._allow_duplicates)
             # No specified data type - use base class
             self.assertEqual(d4.__class__, ihm.dataset.Dataset)
             self.assertIsNone(d1.details)