Found yet another bug in chapter processing (during merges). Changed merge...

Found yet another bug in chapter processing (during merges). Changed merge behavior to favor incoming object structure. Refactored base class merge to not rely on __dict__

src/domain.py

@@ -6,7 +6,7 @@ from copy import deepcopy
 import json
 from dateutil.parser import parse
 import re
-
+import inspect
 
 class Gcisbase(object):
     def __init__(self, data, fields=[], trans={}):
@@ -40,9 +40,13 @@ class Gcisbase(object):
                     setattr(self, k, data[k])
 
     def merge(self, other):
-        for k in self.__dict__:
-            if self.__dict__[k] in (None, '') and hasattr(other, k):
-                self.__dict__[k] = getattr(other, k)
+        #This sucks
+        attrs_we_care_about = [(attr, v) for attr, v in inspect.getmembers(self, lambda a: not (inspect.isroutine(a)))
+                               if not attr.startswith('__')]
+
+        for attr, value in attrs_we_care_about:
+            if value in (None, '') and hasattr(other, attr):
+                setattr(self, attr, value)
         return self
 
     def as_json(self, indent=0, omit_fields=[]):
@@ -69,7 +73,7 @@ class Figure(Gcisbase):
 
         #Special case for chapter
         chap_tree = data.pop('chapter', None)
-        self.chapter = Chapter(chap_tree) if chap_tree else None
+        self.chapter = Chapter(chap_tree) if chap_tree else self.chapter
 
         #Special case for images
         image_list = data.pop('images', None)
@@ -83,7 +87,7 @@ class Figure(Gcisbase):
         if isinstance(self.chapter, Chapter) and self.chapter.number and self.ordinal:
             return '{0}.{1}'.format(self.chapter.number, self.ordinal)
         else:
-            return self.ordinal
+            return '{0}.{1}'.format(self.chapter, self.ordinal)
 
     #TODO: Ordinal handling is unnecessarily complex
     @figure_num.setter
@@ -114,7 +118,21 @@ class Figure(Gcisbase):
         return string
 
     def __repr__(self):
-        return super(Figure, self).__repr__()
+        # return super(Figure, self).__repr__()
+        return self.__str__()
+
+    def merge(self, other):
+        #Special handling for Chapters
+        if isinstance(other.chapter, Chapter) and isinstance(self.chapter, Chapter):
+            self.chapter.merge(other.chapter)
+
+        #This might want to move to Chapter's merge()
+        elif isinstance(other.chapter, Chapter) and not isinstance(self.chapter, Chapter):
+            chapter_num = self.chapter
+            self.chapter = other.chapter
+            self.chapter.number = chapter_num
+
+        return super(Figure, self).merge(other)
 
 
 class Chapter(Gcisbase):

src/problems.py

+__author__ = 'abuddenberg'

src/test_suite.py

@@ -55,14 +55,16 @@ def test_chapter_parsing():
     gcis_fig = Figure(json.loads(test_figure_json))
 
     assert isinstance(webform_fig.ordinal, int)
-    assert webform_fig.figure_num == 17
+    assert webform_fig.figure_num == '2.17'
     assert webform_fig.ordinal == 17
-    assert webform_fig.chapter is None
+    assert webform_fig.chapter == 2
 
     merged_figure = webform_fig.merge(gcis_fig)
     #FYI, these are identical; I just wanted the variable name to reflect the merge
     assert id(merged_figure) == id(webform_fig)
 
+    assert isinstance(gcis_fig.chapter, Chapter)
+
     assert merged_figure.figure_num == '2.17'
     assert merged_figure.ordinal == 17
     assert isinstance(merged_figure.chapter, Chapter)
@@ -92,4 +94,4 @@ def test_dataset_special_properties():
 
 
 if __name__ == '__main__':
-    test_dataset_special_properties()
\ No newline at end of file
+    test_chapter_parsing()
\ No newline at end of file