Merge pull request #3 from rchristie/fix-alignment

Improve segment automatic alignment

Author: rchristie

src/segmentationstitcher/annotation.py

@@ -5,6 +5,10 @@
 from cmlibs.utils.zinc.field import get_group_list
 from cmlibs.utils.zinc.group import group_get_highest_dimension, groups_have_same_local_contents
 from cmlibs.zinc.field import Field
+import logging
+
+
+logger = logging.getLogger(__name__)
 
 
 class AnnotationCategory(Enum):
@@ -24,9 +28,18 @@ def get_group_name(self):
         """
         return '.' + self.name
 
+    def get_lower_name(self):
+        """
+        :return: Lower case category name.
+        """
+        return self.name.lower()
+
     def is_connectable(self):
         return self in (self.INDEPENDENT_NETWORK, self.NETWORK_GROUP_1, self.NETWORK_GROUP_2)
 
+    def is_connectable_different_annotation(self):
+        return self.is_connectable() and not (self == self.INDEPENDENT_NETWORK)
+
 
 class Annotation:
     """
@@ -56,8 +69,8 @@ def decode_settings(self, settings_in: dict):
         assert (settings_in.get("name") == self._name) and (settings_in.get("term") == self._term)
         settings_dimension = settings_in.get("dimension")
         if settings_dimension != self._dimension:
-            print("WARNING: Segmentation Stitcher.  Annotation with name", self._name, "term", self._term,
-                  "was dimension ", settings_dimension, "in settings, is now ", self._dimension,
+            logger.warning("Segmentation Stitcher.  Annotation with name " + self._name, " term " + str(self._term) +
+                  " was dimension " + str(settings_dimension), "in settings, is now " + str(self._dimension) +
                   ". Have input files changed?")
             settings_in["dimension"] = self._dimension
         # update current settings to gain new ones and override old ones
@@ -125,6 +138,31 @@ def set_term(self, term):
         assert self._term is None
         self._term = term
 
+    def clear_term(self):
+        """
+        Clear term to None, call in cases of mismatched terms for the same group name.
+        """
+        self._term = None
+
+    def is_connectable(self):
+        """
+        :return: True if annotation's category is connectable.
+        """
+        return self._category.is_connectable()
+
+    def is_connectable_with(self, other_annotation):
+        """
+        Query whether ends annotated with self and other_annotation can be connected.
+        :param other_annotation: Another Annotation object.
+        :return: True if self and other_annotation are allowed to be connected by a link.
+        """
+        if self._category.is_connectable() and (self._category == other_annotation.get_category()):
+            if other_annotation is self:
+                return True
+            elif self._category.is_connectable_different_annotation():
+                return True
+        return False
+
 
 def region_get_annotations(region, network_group1_keywords, network_group2_keywords, term_keywords):
     """
@@ -146,6 +184,17 @@ def region_get_annotations(region, network_group1_keywords, network_group2_keywo
     annotations = []
     term_annotations = []
     datapoints = fieldmodule.findNodesetByFieldDomainType(Field.DOMAIN_TYPE_DATAPOINTS)
+    # these terms have had slight mismatches with contents of url group, so explicitly matching:
+    segment_name = region.getParent().getName()
+    known_terms = {
+        "epineurium": "http://uri.interlex.org/base/ilx_0103892",
+        "left cervical vagus nerve": "http://uri.interlex.org/base/ilx_0794142",
+        "right cervical vagus nerve": "http://uri.interlex.org/base/ilx_0794141",
+        "left thoracic vagus nerve": "http://uri.interlex.org/base/ilx_0787543",
+        "right thoracic vagus nerve": "http://uri.interlex.org/base/ilx_0786664",
+        "left vagus x nerve trunk": "http://uri.interlex.org/base/ilx_0736691",
+        "right vagus x nerve trunk": "http://uri.interlex.org/base/ilx_0730515"
+    }
     for group in groups:
         # clean up name to remove case and leading/trailing whitespace
         name = group.getName().strip()
@@ -159,7 +208,9 @@ def region_get_annotations(region, network_group1_keywords, network_group2_keywo
                 continue  # empty group
         if lower_name.isdigit():
             continue  # ignore as these can never be valid annotation names
-        category = AnnotationCategory.GENERAL
+        if '<property name=' in lower_name:
+            continue  # don't want these groups output by mbfxml2ex
+        category = AnnotationCategory.EXCLUDE if ('_' in lower_name) else AnnotationCategory.GENERAL
         for keyword in network_group1_keywords:
             if keyword in lower_name:
                 category = AnnotationCategory.NETWORK_GROUP_1
@@ -169,9 +220,10 @@ def region_get_annotations(region, network_group1_keywords, network_group2_keywo
                 if keyword in lower_name:
                     category = AnnotationCategory.NETWORK_GROUP_2
                     break
-        annotation = Annotation(name, None, dimension, category)
+        term = known_terms.get(lower_name)
+        annotation = Annotation(name, term, dimension, category)
         is_term = False
-        if category == AnnotationCategory.GENERAL:
+        if category in (AnnotationCategory.GENERAL, AnnotationCategory.EXCLUDE):
             for keyword in term_keywords:
                 if keyword in lower_name:
                     is_term = True
@@ -180,23 +232,55 @@ def region_get_annotations(region, network_group1_keywords, network_group2_keywo
             term_annotations.append(annotation)
         else:
             annotations.append(annotation)
+
+    # sort annotations to have networks first, then general, lastly EXCLUDE to associate terms with earlier ones first
+    ordered_categories = (
+        AnnotationCategory.NETWORK_GROUP_1,
+        AnnotationCategory.NETWORK_GROUP_2,
+        AnnotationCategory.INDEPENDENT_NETWORK,
+        AnnotationCategory.GENERAL,
+        AnnotationCategory.EXCLUDE)
+    assert len(ordered_categories) == len(AnnotationCategory)  # in case new categories added, expand the above
+    sorted_annotations = []
+    for category in ordered_categories:
+        for annotation in annotations:
+            if annotation.get_category() == category:
+                sorted_annotations.append(annotation)
+
     for term_annotation in term_annotations:
         term = term_annotation.get_name()
         term_group = fieldmodule.findFieldByName(term).castGroup()
         dimension = term_annotation.get_dimension()
-        for annotation in annotations:
-            if annotation.get_term() is not None:
-                continue
+        term_matched = False
+        for annotation in sorted_annotations:
             if annotation.get_dimension() != dimension:
                 continue
+            if annotation.get_category() == AnnotationCategory.EXCLUDE:
+                continue
             name = annotation.get_name()
             name_group = fieldmodule.findFieldByName(name).castGroup()
             if groups_have_same_local_contents(name_group, term_group):
-                annotation.set_term(term)
-                break
-        else:
-            print("WARNING: Segmentation Stitcher.  Did not find matching annotation name for term", term,
-                  ". Adding separate annotation.")
+                old_term = annotation.get_term()
+                if old_term:
+                    if old_term != term:
+                        logger.warning("Segment " + segment_name + ": " +
+                            "Annotation name " + name + " already has term " + old_term +
+                            " but matched group with term " + term + ". Keeping original term.")
+                else:
+                    # logger.info("Segment " + segment_name + ": " + "Annotation name " + name +
+                    #             " discovered term " + term + ".")
+                    annotation.set_term(term)
+                term_matched = True
+                # do not break to allow all groups with matching contents to get the term
+            else:
+                known_term = known_terms.get(name.lower())
+                if known_term == term:
+                    logger.warning("Segment " + segment_name + ": " +
+                        "Known annotation name " + name + " and term " + term + " groups differ. Using name group.")
+                    term_matched = True
+        if not term_matched:
+            logger.warning("Segment " + segment_name + ": " +
+                  ".  Did not find matching annotation name for term " + term + ". Adding separate annotation.")
             term_annotation.set_term(term)
             index = 0
             for annotation in annotations: