Escaping special symbols now working correct when dumping; typo on RE…

…ADME corrected; bad assertition removed on test_loads_dumps

README.rst

@@ -1,5 +1,5 @@
 =========
-LIAF-ARFF
+LIAC-ARFF
 =========
 
 The liac-arff module implements functions to read and write ARFF files in

arff.py

@@ -160,6 +160,18 @@
 _RE_RELATION     = re.compile(r'^(\".*\"|\'.*\'|\S*)$', re.UNICODE)
 _RE_ATTRIBUTE    = re.compile(r'^(\".*\"|\'.*\'|\S*)\s+(.+)$', re.UNICODE)
 _RE_TYPE_NOMINAL = re.compile(r'^\{\s*((\".*\"|\'.*\'|\S*)\s*,\s*)*(\".*\"|\'.*\'|\S*)}$', re.UNICODE)
+_RE_ESCAPE = re.compile(r'\\\'|\\\"|\\\%|[\\"\'%]')
+
+_ESCAPE_DCT = {
+    ' ': ' ',
+    "'": "\\'",
+    '"': '\\"',
+    '%': '\\%',
+    '\\': '\\',
+    '\\\'': '\\\'',
+    '\\"': '\\"',
+    '\\%': '\\%',
+}
 # =============================================================================
 
 # COMPATIBILITY WITH PYTHON 3.3 ===============================================
@@ -232,6 +244,11 @@ def __str__(self):
 # =============================================================================
 
 # INTERNAL ====================================================================
+def encode_string(s):
+    def replace(match):
+        return _ESCAPE_DCT[match.group(0)]
+    return u"'" + _RE_ESCAPE.sub(replace, s) + u"'"
+
 class Conversor(object):
     '''Conversor is a helper used for converting ARFF types to Python types.'''
 
@@ -592,8 +609,10 @@ def _encode_data(self, data):
         new_data = []
         for v in data:
             s = unicode(v)
-            if u' ' in s:
-                s = u'"%s"'%s
+            for escape_char in _ESCAPE_DCT:
+                if escape_char in s:
+                    s = encode_string(s)
+                    break
             new_data.append(s)
 
         return u','.join(new_data)

tests/test_dump_escape.py

@@ -0,0 +1,54 @@
+import unittest
+import arff
+
+ARFF_SOURCE = '''% 
+@RELATION teste
+
+@ATTRIBUTE a STRING
+@ATTRIBUTE b {a, b, c, d}
+@ATTRIBUTE c STRING
+
+@DATA
+lorem, b, thisisavalidstatement
+lorem, b, 'this is a valid statement with an % symbol'
+lorem2, d, 'this is a valid statement'
+lorem3, c, 'this is a valid statement with double quotes included """""""! '
+lorem4, a, 'this is a valid statement with singlequotes included \' lol \'! '
+'''
+
+ARFF_DESTINY = '''@RELATION teste
+
+@ATTRIBUTE a STRING
+@ATTRIBUTE b {a, b, c, d}
+@ATTRIBUTE c STRING
+
+@DATA
+lorem,b,thisisavalidstatement
+lorem,b,'this is a valid statement with an \\% symbol'
+lorem2,d,'this is a valid statement'
+lorem3,c,'this is a valid statement with double quotes included \\"\\"\\"\\"\\"\\"\\"! '
+lorem4,a,'this is a valid statement with singlequotes included \\\' lol \\\'! '
+% 
+% 
+% '''
+
+
+class TestDumpEscape(unittest.TestCase):
+    def test_encode_source(self):
+        obj = arff.loads(ARFF_SOURCE)
+        result = arff.dumps(obj)
+        expected = ARFF_DESTINY
+
+        self.assertEqual(result, expected)
+
+    def test_encode_destiny(self):
+        src = ARFF_DESTINY
+
+        count = 0
+        while count < 10:
+            count += 1
+
+            obj = arff.loads(src)
+            src = arff.dumps(obj)
+            self.assertEqual(src, ARFF_DESTINY)
+

tests/test_encode_data.py

@@ -11,7 +11,7 @@ def test_simple(self):
 
         fixture = [1, 3, 'Renato', 'Name with spaces']
         result = encoder._encode_data(fixture)
-        expected = u'1,3,Renato,"Name with spaces"'
+        expected = u"1,3,Renato,'Name with spaces'"
 
         self.assertEqual(result, expected)
 

tests/test_loads_dumps.py

@@ -61,4 +61,3 @@ def test_simple(self):
             arff = dumps(obj)
             self.assertEqual(arff, ARFF)
 
-        assert False