Support Unicode for saving and loading Pascal XML format

Use lxml to pretty the xml instead of minidom because minido does not support unicode
Save xml with unicode
Add tests dir to collect more images for unit tests in the future

labelImg.py

@@ -538,7 +538,6 @@ class MainWindow(QMainWindow, WindowMixin):
         self.actions.edit.setEnabled(selected)
         self.actions.shapeLineColor.setEnabled(selected)
         self.actions.shapeFillColor.setEnabled(selected)
-        print 'shapeSelectionChanged'
 
     def addLabel(self, shape):
         item = QListWidgetItem(shape.label)
@@ -736,8 +735,8 @@ class MainWindow(QMainWindow, WindowMixin):
             ## Label xml file and show bound box according to its filename
             if self.usingPascalVocFormat is True and \
                     self.defaultSaveDir is not None:
-                    basename = os.path.basename(os.path.splitext(self.filename)[0])
-                    xmlPath = os.path.join(self.defaultSaveDir, basename + '.xml')
+                    basename = os.path.basename(os.path.splitext(self.filename)[0]) + '.xml'
+                    xmlPath = os.path.join(self.defaultSaveDir, basename)
                     self.loadPascalXMLByFilename(xmlPath)
 
             return True
@@ -1069,13 +1068,13 @@ class MainWindow(QMainWindow, WindowMixin):
                     else:
                         self.labelHist.append(line)
 
-    def loadPascalXMLByFilename(self, filename):
+    def loadPascalXMLByFilename(self, xmlPath):
         if self.filename is None:
             return
-        if os.path.exists(filename) is False:
+        if os.path.isfile(xmlPath) is False:
             return
 
-        tVocParseReader = PascalVocReader(filename)
+        tVocParseReader = PascalVocReader(xmlPath)
         shapes = tVocParseReader.getShapes()
         self.loadLabels(shapes)
 

libs/pascal_voc_io.py

@@ -1,10 +1,14 @@
+#!/usr/bin/env python
+# -*- coding: utf8 -*-
+import _init_path
 import sys
 from xml.etree import ElementTree
 from xml.etree.ElementTree import Element, SubElement
-from xml.dom import minidom
 from lxml import etree
 
+
 class PascalVocWriter:
+
     def __init__(self, foldername, filename, imgSize, databaseSrc='Unknown', localImgPath=None):
         self.foldername = foldername
         self.filename = filename
@@ -18,8 +22,8 @@ class PascalVocWriter:
             Return a pretty-printed XML string for the Element.
         """
         rough_string = ElementTree.tostring(elem, 'utf8')
-        reparsed = minidom.parseString(rough_string)
-        return reparsed.toprettyxml(indent="\t")
+        root = etree.fromstring(rough_string)
+        return etree.tostring(root, pretty_print=True)
 
     def genXML(self):
         """
@@ -59,13 +63,12 @@ class PascalVocWriter:
 
         segmented = SubElement(top, 'segmented')
         segmented.text = '0'
-
         return top
 
     def addBndBox(self, xmin, ymin, xmax, ymax, name):
         bndbox = {'xmin': xmin, 'ymin': ymin, 'xmax': xmax, 'ymax': ymax}
         bndbox['name'] = name
-        self.boxlist.append(bndbox);
+        self.boxlist.append(bndbox)
 
     def appendObjects(self, top):
         for each_object in self.boxlist:
@@ -97,15 +100,16 @@ class PascalVocWriter:
         else:
             out_file = open(targetFile, 'w')
 
-        out_file.write(self.prettify(root))
+        prettifyResult = self.prettify(root)
+        out_file.write(prettifyResult)
         out_file.close()
 
 
 class PascalVocReader:
 
     def __init__(self, filepath):
-        ## shapes type:
-        ## [labbel, [(x1,y1), (x2,y2), (x3,y3), (x4,y4)], color, color]
+        # shapes type:
+        # [labbel, [(x1,y1), (x2,y2), (x3,y3), (x4,y4)], color, color]
         self.shapes = []
         self.filepath = filepath
         self.parseXML()
@@ -123,7 +127,8 @@ class PascalVocReader:
 
     def parseXML(self):
         assert self.filepath.endswith('.xml'), "Unsupport file format"
-        xmltree = ElementTree.parse(self.filepath).getroot()
+        parser = etree.XMLParser(encoding='utf-8')
+        xmltree = ElementTree.parse(self.filepath, parser=parser).getroot()
         filename = xmltree.find('filename').text
 
         for object_iter in xmltree.findall('object'):
@@ -146,4 +151,3 @@ tmp.addBndBox(10,10,20,30,'chair')
 tmp.addBndBox(1,1,600,600,'car')
 tmp.save()
 """
-