Merge branch 'main' into poly-seq-scheme

Get changes to  ReplaceCategoryFilter and
support for the ihm_entry_collection_mapping table.

Author: benmwebb

ihm/__init__.py

@@ -1908,11 +1908,20 @@ class Collection:
        :param str id: Unique identifier (assigned by the archive).
        :param str name: Short name for the collection.
        :param str details: Longer description of the collection.
+       :param list entries: Explicit list of entry IDs in this collection;
+              if empty, just the current entry is assumed.
 
        See also :attr:`System.collections`.
     """
-    def __init__(self, id, name=None, details=None):
+    def __init__(self, id, name=None, details=None, entries=None):
         self.id, self.name, self.details = id, name, details
+        self.entries = entries or []
+
+    def __set_id(self, val):
+        self.id = val
+
+    # SystemReader wants to set _id, not id
+    _id = property(lambda s: s.id, __set_id)
 
 
 class BranchDescriptor:

ihm/dumper.py

@@ -92,11 +92,22 @@ def dump(self, system, writer):
 
 class _CollectionDumper(Dumper):
     def dump(self, system, writer):
+        self.dump_summary(system, writer)
+        self.dump_mapping(system, writer)
+
+    def dump_summary(self, system, writer):
         with writer.loop("_ihm_entry_collection",
                          ["id", "name", "details"]) as lp:
             for c in system.collections:
                 lp.write(id=c.id, name=c.name, details=c.details)
 
+    def dump_mapping(self, system, writer):
+        with writer.loop("_ihm_entry_collection_mapping",
+                         ["collection_id", "entry_id"]) as lp:
+            for c in system.collections:
+                for eid in (c.entries or [system.id]):
+                    lp.write(collection_id=c.id, entry_id=eid)
+
 
 class _AuditConformDumper(Dumper):
     URL = ("https://raw.githubusercontent.com/" +

ihm/format.py

@@ -893,7 +893,13 @@ def _get_replacement_token(self):
             writer = CifWriter(fh)
             self.dumper.finalize(self.system)
             self.dumper.dump(self.system, writer)
-            return self._RawCifToken(fh.getvalue())
+            cif = fh.getvalue()
+            # Strip _CifLoopWriter prefix and suffix, if any
+            if cif.startswith('#\n'):
+                cif = cif[2:]
+            if cif.endswith('#\n'):
+                cif = cif[:-2]
+            return self._RawCifToken(cif)
 
     def filter_category(self, tok):
         if self.match_token_category(tok):

ihm/reader.py

@@ -481,6 +481,10 @@ def __init__(self, model_class, starting_model_class, system=None):
         self.revisions = IDMapper(self.system.revisions, ihm.Revision,
                                   *(None,) * 4)
 
+        #: Mapping from ID to :class:`ihm.Collection` objects
+        self.collections = IDMapper(self.system.collections, ihm.Collection,
+                                    None)
+
         #: Mapping from ID to :class:`ihm.Entity` objects
         self.entities = IDMapper(self.system.entities, _make_new_entity)
 
@@ -1060,8 +1064,17 @@ class _CollectionHandler(Handler):
     category = '_ihm_entry_collection'
 
     def __call__(self, id, name, details):
-        c = ihm.Collection(id=id, name=name, details=details)
-        self.system.collections.append(c)
+        c = self.sysr.collections.get_by_id(id)
+        c.name = name
+        c.details = details
+
+
+class _CollectionMappingHandler(Handler):
+    category = '_ihm_entry_collection_mapping'
+
+    def __call__(self, collection_id, entry_id):
+        c = self.sysr.collections.get_by_id(collection_id)
+        c.entries.append(entry_id)
 
 
 class _StructHandler(Handler):
@@ -4154,7 +4167,8 @@ class IHMVariant(Variant):
     system_reader = SystemReader
 
     _handlers = [
-        _CollectionHandler, _StructHandler, _SoftwareHandler, _CitationHandler,
+        _CollectionHandler, _CollectionMappingHandler, _StructHandler,
+        _SoftwareHandler, _CitationHandler,
         _DatabaseHandler, _DatabaseStatusHandler,
         _AuditAuthorHandler, _AuditRevisionHistoryHandler,
         _AuditRevisionDetailsHandler, _AuditRevisionGroupHandler,

setup.py

@@ -17,6 +17,20 @@
     build_ext = False
     copy_args.remove('--without-ext')
 
+# Allow building with Python limited API
+if '--py_limited_api' in copy_args:
+    copy_args.remove('--py_limited_api')
+    # We require Python 3.11 or later since we use PyBUF_WRITE which only
+    # became part of the stable API in 3.11
+    ext_limited_args = {
+        'define_macros': [("Py_LIMITED_API", "0x030B0000")],
+        'py_limited_api': True}
+    setup_limited_args = {
+        'options': {"bdist_wheel": {"py_limited_api": "cp311"}}}
+else:
+    ext_limited_args = {}
+    setup_limited_args = {}
+
 if sys.platform == 'win32':
     # Our use of strdup, strerror should be safe - no need for the Windows
     # compiler to warn about it; we want to use the POSIX name for strdup too
@@ -36,7 +50,8 @@
                      extra_compile_args=cargs,
                      swig_opts=['-keyword', '-nodefaultctor',
                                 '-nodefaultdtor', '-noproxy'],
-                     optional=True)]
+                     optional=True,
+                     **ext_limited_args)]
 else:
     mod = []
 
@@ -61,4 +76,5 @@
           "Operating System :: OS Independent",
           "Intended Audience :: Science/Research",
           "Topic :: Scientific/Engineering",
-      ])
+      ],
+      **setup_limited_args)

src/ihm_format.i

@@ -411,7 +411,7 @@ static struct category_handler_data *do_add_handler(
     const char *key_name;
     PyObject *o = PySequence_GetItem(keywords, i);
     if (PyUnicode_Check(o)) {
-      key_name = PyUnicode_AsUTF8(o);
+      key_name = PyUnicode_AsUTF8AndSize(o, NULL);
       if (PySet_Contains(int_keywords, o) == 1) {
         hd->keywords[i] = ihm_keyword_int_new(category, key_name);
       } else if (PySet_Contains(float_keywords, o) == 1) {

test/test_dumper.py

@@ -1472,9 +1472,14 @@ def test_nonpoly_scheme_dumper(self):
 
     def test_collection_dumper(self):
         """Test CollectionDumper"""
-        system = ihm.System()
+        system = ihm.System(id='1abc')
+        # Collection with no explicit entries (will use system ID)
         c = ihm.Collection('foo', name='bar', details='more text')
         system.collections.append(c)
+        # Collection with explicit entries
+        c = ihm.Collection('coll2', name='cname2', details='cdetails2',
+                           entries=['2xyz', '3hjk'])
+        system.collections.append(c)
         dumper = ihm.dumper._CollectionDumper()
         out = _get_dumper_output(dumper, system)
         self.assertEqual(out, """#
@@ -1483,6 +1488,15 @@ def test_collection_dumper(self):
 _ihm_entry_collection.name
 _ihm_entry_collection.details
 foo bar 'more text'
+coll2 cname2 cdetails2
+#
+#
+loop_
+_ihm_entry_collection_mapping.collection_id
+_ihm_entry_collection_mapping.entry_id
+foo 1abc
+coll2 2xyz
+coll2 3hjk
 #
 """)
 

test/test_format.py

@@ -1263,14 +1263,29 @@ def test_cif_token_reader_replace_category_filter(self):
 _cat5.bar
 _cat5.baz
 a b
+#
+_cat6.a 1
+#
+loop_
+_cat7.bar
+_cat7.baz
+a b
+#
 """
         d = ihm.dumper._CommentDumper()
+        colld = ihm.dumper._CollectionDumper()
         s = ihm.System()
         s.comments.extend(['comment1', 'comment2'])
+        c = ihm.Collection('foo', name='bar', details='more text')
+        s.collections.append(c)
         r = ihm.format.CifTokenReader(StringIO(cif))
         filters = [ihm.format.ReplaceCategoryFilter("cat1"),
                    ihm.format.ReplaceCategoryFilter("_cat2", raw_cif='FOO'),
                    ihm.format.ReplaceCategoryFilter("cat3", dumper=d,
+                                                    system=s),
+                   ihm.format.ReplaceCategoryFilter("cat6", dumper=colld,
+                                                    system=s),
+                   ihm.format.ReplaceCategoryFilter("cat7", dumper=colld,
                                                     system=s)]
         tokens = list(r.read_file(filters))
         new_cif = "".join(x.as_mmcif() for x in tokens)
@@ -1289,6 +1304,32 @@ def test_cif_token_reader_replace_category_filter(self):
 _cat5.bar
 _cat5.baz
 a b
+#
+loop_
+_ihm_entry_collection.id
+_ihm_entry_collection.name
+_ihm_entry_collection.details
+foo bar 'more text'
+#
+#
+loop_
+_ihm_entry_collection_mapping.collection_id
+_ihm_entry_collection_mapping.entry_id
+foo model
+#
+loop_
+_ihm_entry_collection.id
+_ihm_entry_collection.name
+_ihm_entry_collection.details
+foo bar 'more text'
+#
+#
+loop_
+_ihm_entry_collection_mapping.collection_id
+_ihm_entry_collection_mapping.entry_id
+foo model
+
+#
 """)
 
     def test_category_token_group(self):

test/test_reader.py

@@ -223,13 +223,23 @@ def test_collection_handler(self):
 _ihm_entry_collection.name
 _ihm_entry_collection.details
 foo bar 'more text'
+c2id c2name c2details
+#
+_ihm_entry_collection_mapping.collection_id c2id
+_ihm_entry_collection_mapping.entry_id      entry1
 """
         for fh in cif_file_handles(cif):
             s, = ihm.reader.read(fh)
-            c, = s.collections
-            self.assertEqual(c.id, 'foo')
-            self.assertEqual(c.name, 'bar')
-            self.assertEqual(c.details, 'more text')
+            c1, c2 = s.collections
+            self.assertEqual(c1.id, 'foo')
+            self.assertEqual(c1.name, 'bar')
+            self.assertEqual(c1.details, 'more text')
+            self.assertEqual(c1.entries, [])
+
+            self.assertEqual(c2.id, 'c2id')
+            self.assertEqual(c2.name, 'c2name')
+            self.assertEqual(c2.details, 'c2details')
+            self.assertEqual(c2.entries, ['entry1'])
 
     def test_software_handler(self):
         """Test SoftwareHandler"""