[GR-17590] [GR-17591][GR-16846] regex problem with None values and radd problem with custom string class

PullRequest: graalpython/605

Author: Franziska Geiger

README.md

@@ -21,7 +21,7 @@ wrapper scripts and makes the implementation usable from shell as standard Pytho
 execute the following command in the project directory:
 
 ```
-mx graalpython -m venv <dir-to-venv>
+mx python -m venv <dir-to-venv>
 ```
 
 To activate the environment in your shell session call:

graalpython/com.oracle.graal.python.test/src/tests/test_re.py

@@ -82,6 +82,15 @@ def test_json_bytes_re_compile():
     else:
         assert False, "searching a bytes-pattern in a str did not raise"
 
+def test_none_value():
+    regex_find = re.compile(
+        r"(//?| ==?)|([[]]+)").findall
+    stream = iter([ (special,text)
+                    for (special,text) in regex_find('[]')
+                    if special or text ])
+    n = next(stream)
+    assert not n[0]
+    assert str(n[0]) == 'None'
 
 class S(str):
     def __getitem__(self, index):
@@ -370,7 +379,7 @@ def test_getattr(self):
         self.assertEqual(re.compile("(?i)(a)(b)").pattern, "(?i)(a)(b)")
         # TODO at the moment, we use slightly different default flags
         #self.assertEqual(re.compile("(?i)(a)(b)").flags, re.I | re.U)
-        
+
         # TODO re-enable this test once TRegex provides this property
         #self.assertEqual(re.compile("(?i)(a)(b)").groups, 2)
         self.assertEqual(re.compile("(?i)(a)(b)").groupindex, {})
@@ -390,7 +399,7 @@ def test_getattr(self):
         p = re.compile(r'(?i)(?P<first>a)(?P<other>b)')
         self.assertEqual(sorted(p.groupindex), ['first', 'other'])
         self.assertEqual(p.groupindex['other'], 2)
-        
+
         if sys.version_info.minor >= 6:
             with self.assertRaises(TypeError):
                 p.groupindex['other'] = 0
@@ -438,7 +447,15 @@ def test_escaping(self):
         self.assertTrue(match)
         assert "frac" in match.groupdict()
         assert match.groupdict()["frac"] == "1"
-        
-    
+
+
     def test_escape(self):
         self.assertEqual(re.escape(" ()"), "\\ \\(\\)")
+
+    def test_finditer_empty_string(self):
+        regex = re.compile(
+            r"(//?| ==?)|([[]]+)")
+        for m in regex.finditer(''):
+            self.fail()
+
+

graalpython/com.oracle.graal.python.test/src/tests/test_string.py

@@ -1096,3 +1096,34 @@ def test_strip_with_sep():
     assertRaises(TypeError, 'hello', 'strip', 42, 42)
     assertRaises(TypeError, 'hello', 'lstrip', 42, 42)
     assertRaises(TypeError, 'hello', 'rstrip', 42, 42)
+
+class EncodedString(str):
+    # unicode string subclass to keep track of the original encoding.
+    # 'encoding' is None for unicode strings and the source encoding
+    # otherwise
+    encoding = None
+
+    def __deepcopy__(self, memo):
+        return self
+
+    def byteencode(self):
+        assert self.encoding is not None
+        return self.encode(self.encoding)
+
+    def utf8encode(self):
+        assert self.encoding is None
+        return self.encode("UTF-8")
+
+    @property
+    def is_unicode(self):
+        return self.encoding is None
+
+    def contains_surrogates(self):
+        return string_contains_surrogates(self)
+
+    def as_utf8_string(self):
+        return bytes_literal(self.utf8encode(), 'utf8')
+
+def test_radd():
+    val = EncodedString('abc')
+    assert 'cde' + val == 'cdeabc'

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/builtins/objects/str/StringBuiltins.java

@@ -345,6 +345,10 @@ public abstract static class AddNode extends PythonBinaryBuiltinNode {
         protected final ConditionProfile rightProfile1 = ConditionProfile.createBinaryProfile();
         protected final ConditionProfile rightProfile2 = ConditionProfile.createBinaryProfile();
 
+        public static AddNode create() {
+            return StringBuiltinsFactory.AddNodeFactory.create();
+        }
+
         @Specialization(guards = "!concatGuard(self, other)")
         String doSSSimple(String self, String other) {
             if (LazyString.length(self, leftProfile1, leftProfile2) == 0) {
@@ -369,7 +373,7 @@ Object doSSSimple(PString self, String other) {
             return self;
         }
 
-        @Specialization(guards = "!concatGuard(self.getCharSequence(), self.getCharSequence())")
+        @Specialization(guards = "!concatGuard(self.getCharSequence(), other.getCharSequence())")
         PString doSSSimple(PString self, PString other) {
             if (LazyString.length(self.getCharSequence(), leftProfile1, leftProfile2) == 0) {
                 return other;
@@ -447,7 +451,12 @@ protected boolean concatGuard(CharSequence left, CharSequence right) {
 
     @Builtin(name = __RADD__, minNumOfPositionalArgs = 2)
     @GenerateNodeFactory
-    public abstract static class RAddNode extends AddNode {
+    public abstract static class RAddNode extends PythonBinaryBuiltinNode {
+        @Specialization
+        Object doAll(VirtualFrame frame, Object left, Object right,
+                        @Cached("create()") AddNode addNode) {
+            return addNode.execute(frame, right, left);
+        }
     }
 
     // str.startswith(prefix[, start[, end]])

graalpython/lib-graalpython/_sre.py

@@ -339,12 +339,14 @@ def __sanitize_out_type(self, elem):
            returned list if always either 'str' or 'bytes'."""
         if self.__binary:
             return bytes(elem)
+        elif elem is None:
+            return None
         else:
             return str(elem)
 
     def finditer(self, string, pos=0, endpos=-1):
         self.__check_input_type(string)
-        if endpos > len(string):
+        if endpos > len(string) or len(string) == 0:
             endpos = len(string)
         elif endpos < 0:
             endpos = endpos % len(string) + 1