#2650: Merge with 3.1.

This commit is contained in:
Ezio Melotti 2011-03-25 14:25:36 +02:00
commit 213eb96902
2 changed files with 55 additions and 23 deletions

View File

@ -224,8 +224,7 @@ def escape(pattern):
if isinstance(pattern, str): if isinstance(pattern, str):
alphanum = _alphanum_str alphanum = _alphanum_str
s = list(pattern) s = list(pattern)
for i in range(len(pattern)): for i, c in enumerate(pattern):
c = pattern[i]
if c not in alphanum: if c not in alphanum:
if c == "\000": if c == "\000":
s[i] = "\\000" s[i] = "\\000"

View File

@ -1,7 +1,9 @@
from test.support import verbose, run_unittest from test.support import verbose, run_unittest
import re import re
from re import Scanner from re import Scanner
import sys, traceback import sys
import string
import traceback
from weakref import proxy from weakref import proxy
# Misc tests from Tim Peters' re.doc # Misc tests from Tim Peters' re.doc
@ -411,31 +413,62 @@ class ReTests(unittest.TestCase):
self.assertEqual(re.search("\s(b)", " b").group(1), "b") self.assertEqual(re.search("\s(b)", " b").group(1), "b")
self.assertEqual(re.search("a\s", "a ").group(0), "a ") self.assertEqual(re.search("a\s", "a ").group(0), "a ")
def test_re_escape(self): def assertMatch(self, pattern, text, match=None, span=None,
p="" matcher=re.match):
self.assertEqual(re.escape(p), p) if match is None and span is None:
for i in range(0, 256): # the pattern matches the whole text
p = p + chr(i) match = text
self.assertEqual(re.match(re.escape(chr(i)), chr(i)) is not None, span = (0, len(text))
True) elif match is None or span is None:
self.assertEqual(re.match(re.escape(chr(i)), chr(i)).span(), (0,1)) raise ValueError('If match is not None, span should be specified '
'(and vice versa).')
m = matcher(pattern, text)
self.assertTrue(m)
self.assertEqual(m.group(), match)
self.assertEqual(m.span(), span)
pat=re.compile(re.escape(p)) def test_re_escape(self):
self.assertEqual(pat.match(p) is not None, True) alnum_chars = string.ascii_letters + string.digits
self.assertEqual(pat.match(p).span(), (0,256)) p = ''.join(chr(i) for i in range(256))
for c in p:
if c in alnum_chars:
self.assertEqual(re.escape(c), c)
elif c == '\x00':
self.assertEqual(re.escape(c), '\\000')
else:
self.assertEqual(re.escape(c), '\\' + c)
self.assertMatch(re.escape(c), c)
self.assertMatch(re.escape(p), p)
def test_re_escape_byte(self): def test_re_escape_byte(self):
p=b"" alnum_chars = (string.ascii_letters + string.digits).encode('ascii')
self.assertEqual(re.escape(p), p) p = bytes(range(256))
for i in range(0, 256): for i in p:
b = bytes([i]) b = bytes([i])
p += b if b in alnum_chars:
self.assertEqual(re.match(re.escape(b), b) is not None, True) self.assertEqual(re.escape(b), b)
self.assertEqual(re.match(re.escape(b), b).span(), (0,1)) elif i == 0:
self.assertEqual(re.escape(b), b'\\000')
else:
self.assertEqual(re.escape(b), b'\\' + b)
self.assertMatch(re.escape(b), b)
self.assertMatch(re.escape(p), p)
pat=re.compile(re.escape(p)) def test_re_escape_non_ascii(self):
self.assertEqual(pat.match(p) is not None, True) s = 'xxx\u2620\u2620\u2620xxx'
self.assertEqual(pat.match(p).span(), (0,256)) s_escaped = re.escape(s)
self.assertEqual(s_escaped, 'xxx\\\u2620\\\u2620\\\u2620xxx')
self.assertMatch(s_escaped, s)
self.assertMatch('.%s+.' % re.escape('\u2620'), s,
'x\u2620\u2620\u2620x', (2, 7), re.search)
def test_re_escape_non_ascii_bytes(self):
b = 'y\u2620y\u2620y'.encode('utf-8')
b_escaped = re.escape(b)
self.assertEqual(b_escaped, b'y\\\xe2\\\x98\\\xa0y\\\xe2\\\x98\\\xa0y')
self.assertMatch(b_escaped, b)
res = re.findall(re.escape('\u2620'.encode('utf-8')), b)
self.assertEqual(len(res), 2)
def pickle_test(self, pickle): def pickle_test(self, pickle):
oldpat = re.compile('a(?:b|(c|e){1,2}?|d)+?(.)') oldpat = re.compile('a(?:b|(c|e){1,2}?|d)+?(.)')