Merge pull request #70 from kba/test-points-xywh

fix/test utility coordination translation functions

ocrd/model/ocrd_xml_base.py

@@ -4,8 +4,8 @@
 )
 from ocrd.utils import (
     xmllint_format,
-    xywh_from_coordinate_string,
-    coordinate_string_from_xywh
+    xywh_from_points,
+    points_from_xywh
 )
 
 from lxml import etree as ET
@@ -22,14 +22,14 @@ def get_coords(el):
     coords = el.find('page:Coords', NAMESPACES)
     if coords is not None:
         points = coords.get('points')
-        return xywh_from_coordinate_string(points)
+        return xywh_from_points(points)
 
 def set_coords(el, box):
     if box is not None:
         coords = el.find('page:Coords', NAMESPACES)
         if coords is None:
             coords = ET.SubElement(el, TAG_PAGE_COORDS)
-        coords.set("points", coordinate_string_from_xywh(box))
+        coords.set("points", points_from_xywh(box))
 
 class OcrdXmlDocument(object):
 

ocrd/utils.py

@@ -10,23 +10,32 @@
 def getLogger(*args, **kwargs):
     return logging.getLogger(*args, **kwargs)
 
-def coordinate_string_from_xywh(box):
+def points_from_xywh(box):
     """
     Constructs a polygon representation from a rectangle described as a dict with keys x, y, w, h.
     """
+    x, y, w, h = box['x'], box['y'], box['w'], box['h']
     # tesseract uses a different region representation format
     return "%i,%i %i,%i %i,%i %i,%i" % (
-        box['x'],
-        box['y'],
-        box['x'] + box['w'],
-        box['y'] + box['w'],
-        box['x'] + box['w'] + box['h'],
-        box['y'] + box['w'] + box['h'],
-        box['x'] + box['h'],
-        box['y'] + box['h']
+        x, y,
+        x + w, y,
+        x + w, y + h,
+        x, y + h
     )
 
-def xywh_from_coordinate_string(points):
+def xywh_from_points(points):
+    """
+    Constructs an dict representing a rectangle with keys x, y, w, h
+    """
+    [tl, tr, br] = [[float(p) for p in pair.split(',')] for pair in points.split(' ')[:3]]
+    return {
+        'x': tl[0],
+        'y': tl[1],
+        'w': tr[0] - tl[0],
+        'h': br[1] - tr[1],
+    }
+
+def polygon_from_points(points):
     """
     Constructs a numpy-compatible polygon from a page representation.
     """
@@ -46,7 +55,11 @@ def xmllint_format(xml):
     output, _ = proc.communicate(xml)
     return output
 
+# TODO better name
 def mets_file_id(grp, n):
+    """
+    Concatenate string and zero-padded 4 digit number
+    """
     return "%s_%04i"  % (grp, n + 1)
 
 def safe_filename(url):

test/test_utils.py

@@ -0,0 +1,29 @@
+from test.base import TestCase, main
+from ocrd.utils import (
+    points_from_xywh,
+    xywh_from_points,
+    polygon_from_points
+)
+
+class TestResolver(TestCase):
+
+    #  def runTest(self):
+
+    def test_points_from_xywh(self):
+        self.assertEqual(
+            points_from_xywh({'x': 100, 'y': 100, 'w': 100, 'h': 100}),
+            '100,100 200,100 200,200 100,200'
+        )
+
+    def test_xywh_from_points(self):
+        self.assertEqual(
+            xywh_from_points('100,100 200,100 200,200 100,200'),
+            {'x': 100, 'y': 100, 'w': 100, 'h': 100})
+
+    def test_polygon_from_points(self):
+        self.assertEqual(
+            polygon_from_points('100,100 200,100 200,200 100,200'),
+            [[100, 100], [200, 100], [200, 200], [100, 200]])
+
+if __name__ == '__main__':
+    main()