bump version, resolve http://code.google.com/p/simplejson/issues/detail?id=3

git-svn-id: http://simplejson.googlecode.com/svn/trunk@80 a4795897-2c25-0410-b006-0d3caba88fa1

5 files changed, 32 insertions, 14 deletions

diff --git a/setup.py b/setup.py
index 7fbd436..082023f 100644
--- a/setup.py
+++ b/setup.py
@@ -18,7 +18,7 @@ from distutils.command.build_ext import build_ext
 from distutils.errors import CCompilerError, DistutilsExecError, \
     DistutilsPlatformError
 
-VERSION = '1.8.1'
+VERSION = '1.8.2'
 DESCRIPTION = "Simple, fast, extensible JSON encoder/decoder for Python"
 LONG_DESCRIPTION = """
 simplejson is a simple, fast, complete, correct and extensible
diff --git a/simplejson/__init__.py b/simplejson/__init__.py
index be9e87a..94b94f8 100644
--- a/simplejson/__init__.py
+++ b/simplejson/__init__.py
@@ -99,7 +99,7 @@ pretty-print::
 Note that the JSON produced by this module's default settings
 is a subset of YAML, so it may be used as a serializer for that as well.
 """
-__version__ = '1.8.1'
+__version__ = '1.8.2'
 __all__ = [
     'dump', 'dumps', 'load', 'loads',
     'JSONDecoder', 'JSONEncoder',
diff --git a/simplejson/_speedups.c b/simplejson/_speedups.c
index c7a7a7c..7021b21 100644
--- a/simplejson/_speedups.c
+++ b/simplejson/_speedups.c
@@ -239,7 +239,7 @@ join_list_unicode(PyObject *lst) {
 }
 
 static PyObject *
-scanstring_str(PyObject *pystr, Py_ssize_t end, char *encoding) {
+scanstring_str(PyObject *pystr, Py_ssize_t end, char *encoding, int strict) {
     PyObject *rval;
     Py_ssize_t len = PyString_GET_SIZE(pystr);
     Py_ssize_t begin = end - 1;
@@ -257,6 +257,9 @@ scanstring_str(PyObject *pystr, Py_ssize_t end, char *encoding) {
             c = buf[next];
             if (c == '"' || c == '\\') {
                 break;
+            } else if (strict && c <= 0x1f) {
+                raise_errmsg("Invalid control character at", pystr, begin);
+                goto bail;
             }
         }
         if (!(c == '"' || c == '\\')) {
@@ -397,7 +400,7 @@ bail:
 
 
 static PyObject *
-scanstring_unicode(PyObject *pystr, Py_ssize_t end) {
+scanstring_unicode(PyObject *pystr, Py_ssize_t end, int strict) {
     PyObject *rval;
     Py_ssize_t len = PyUnicode_GET_SIZE(pystr);
     Py_ssize_t begin = end - 1;
@@ -415,6 +418,9 @@ scanstring_unicode(PyObject *pystr, Py_ssize_t end) {
             c = buf[next];
             if (c == '"' || c == '\\') {
                 break;
+            } else if (strict && c <= 0x1f) {
+                raise_errmsg("Invalid control character at", pystr, begin);
+                goto bail;
             }
         }
         if (!(c == '"' || c == '\\')) {
@@ -559,11 +565,12 @@ static PyObject *
 py_scanstring(PyObject* self UNUSED, PyObject *args) {
     PyObject *pystr;
     Py_ssize_t end;
-    char *encoding;
+    char *encoding = NULL;
+    int strict = 0;
 #if PY_VERSION_HEX < 0x02050000 
-    if (!PyArg_ParseTuple(args, "Oiz:scanstring", &pystr, &end, &encoding)) {
+    if (!PyArg_ParseTuple(args, "Oi|zi:scanstring", &pystr, &end, &encoding, &strict)) {
 #else
-    if (!PyArg_ParseTuple(args, "Onz:scanstring", &pystr, &end, &encoding)) {
+    if (!PyArg_ParseTuple(args, "On|zi:scanstring", &pystr, &end, &encoding, &strict)) {
 #endif
         return NULL;
     }
@@ -571,9 +578,9 @@ py_scanstring(PyObject* self UNUSED, PyObject *args) {
         encoding = DEFAULT_ENCODING;
     }
     if (PyString_Check(pystr)) {
-        return scanstring_str(pystr, end, encoding);
+        return scanstring_str(pystr, end, encoding, strict);
     } else if (PyUnicode_Check(pystr)) {
-        return scanstring_unicode(pystr, end);
+        return scanstring_unicode(pystr, end, strict);
     }
     PyErr_SetString(PyExc_TypeError, "first argument must be a string");
     return NULL;
diff --git a/simplejson/decoder.py b/simplejson/decoder.py
index 6a73e9a..6d4c5d7 100644
--- a/simplejson/decoder.py
+++ b/simplejson/decoder.py
@@ -70,7 +70,7 @@ def JSONNumber(match, context):
     return res, None
 pattern(r'(-?(?:0|[1-9]\d*))(\.\d+)?([eE][-+]?\d+)?')(JSONNumber)
 
-STRINGCHUNK = re.compile(r'(.*?)(["\\])', FLAGS)
+STRINGCHUNK = re.compile(r'(.*?)(["\\\x00-\x1f])', FLAGS)
 BACKSLASH = {
     '"': u'"', '\\': u'\\', '/': u'/',
     'b': u'\b', 'f': u'\f', 'n': u'\n', 'r': u'\r', 't': u'\t',
@@ -78,7 +78,7 @@ BACKSLASH = {
 
 DEFAULT_ENCODING = "utf-8"
 
-def scanstring(s, end, encoding=None, _b=BACKSLASH, _m=STRINGCHUNK.match):
+def scanstring(s, end, encoding=None, strict=True, _b=BACKSLASH, _m=STRINGCHUNK.match):
     if encoding is None:
         encoding = DEFAULT_ENCODING
     chunks = []
@@ -97,6 +97,12 @@ def scanstring(s, end, encoding=None, _b=BACKSLASH, _m=STRINGCHUNK.match):
             _append(content)
         if terminator == '"':
             break
+        elif terminator != '\\':
+            if strict:
+                raise ValueError(errmsg("Invalid control character %r at", s, end))
+            else:
+                _append(terminator)
+                continue
         try:
             esc = s[end]
         except IndexError:
@@ -140,7 +146,8 @@ if _speedups is not None:
 
 def JSONString(match, context):
     encoding = getattr(context, 'encoding', None)
-    return scanstring(match.string, match.end(), encoding)
+    strict = getattr(context, 'strict', True)
+    return scanstring(match.string, match.end(), encoding, strict)
 pattern(r'"')(JSONString)
 
 WHITESPACE = re.compile(r'\s*', FLAGS)
@@ -157,9 +164,10 @@ def JSONObject(match, context, _w=WHITESPACE.match):
         raise ValueError(errmsg("Expecting property name", s, end))
     end += 1
     encoding = getattr(context, 'encoding', None)
+    strict = getattr(context, 'strict', True)
     iterscan = JSONScanner.iterscan
     while True:
-        key, end = scanstring(s, end, encoding)
+        key, end = scanstring(s, end, encoding, strict)
         end = _w(s, end).end()
         if s[end:end + 1] != ':':
             raise ValueError(errmsg("Expecting : delimiter", s, end))
@@ -257,7 +265,7 @@ class JSONDecoder(object):
     __all__ = ['__init__', 'decode', 'raw_decode']
 
     def __init__(self, encoding=None, object_hook=None, parse_float=None,
-            parse_int=None, parse_constant=None):
+            parse_int=None, parse_constant=None, strict=True):
         """
         ``encoding`` determines the encoding used to interpret any ``str``
         objects decoded by this instance (utf-8 by default).  It has no
@@ -291,6 +299,7 @@ class JSONDecoder(object):
         self.parse_float = parse_float
         self.parse_int = parse_int
         self.parse_constant = parse_constant
+        self.strict = strict
 
     def decode(self, s, _w=WHITESPACE.match):
         """
diff --git a/simplejson/tests/test_fail.py b/simplejson/tests/test_fail.py
index a99d9c4..9534a94 100644
--- a/simplejson/tests/test_fail.py
+++ b/simplejson/tests/test_fail.py
@@ -48,6 +48,8 @@ JSONDOCS = [
     '["Bad value", truth]',
     # http://json.org/JSON_checker/test/fail24.json
     "['single quote']",
+    # http://code.google.com/p/simplejson/issues/detail?id=3
+    u'["A\u001FZ control characters in string"]',
 ]
 
 SKIPS = {