Author: Amaury Forgeot d'Arc <amaur...@gmail.com>
Branch: py3k
Changeset: r49979:5a8c73ef4e8f
Date: 2011-11-29 23:25 +0100
http://bitbucket.org/pypy/pypy/changeset/5a8c73ef4e8f/
Log: (chronitis) Add bytes.maketrans, and allow buffer-compatible objects
to be used as arguments of most bytes methods
diff --git a/pypy/objspace/std/stringobject.py
b/pypy/objspace/std/stringobject.py
--- a/pypy/objspace/std/stringobject.py
+++ b/pypy/objspace/std/stringobject.py
@@ -250,10 +250,10 @@
return space.newlist(res_w)
-def str_split__String_String_ANY(space, w_self, w_by, w_maxsplit=-1):
+def str_split__String_ANY_ANY(space, w_self, w_by, w_maxsplit=-1):
maxsplit = space.int_w(w_maxsplit)
value = w_self._value
- by = w_by._value
+ by = space.bufferstr_w(w_by)
bylen = len(by)
if bylen == 0:
raise OperationError(space.w_ValueError, space.wrap("empty separator"))
@@ -324,12 +324,19 @@
def make_rsplit_with_delim(funcname, sliced):
from pypy.tool.sourcetools import func_with_new_name
+ if 'Unicode' in funcname:
+ def unwrap_sep(space, w_by):
+ return w_by.value
+ else:
+ def unwrap_sep(space, w_by):
+ return space.bufferstr_w(w_by)
+
def fn(space, w_self, w_by, w_maxsplit=-1):
maxsplit = space.int_w(w_maxsplit)
res_w = []
value = w_self._value
end = len(value)
- by = w_by._value
+ by = unwrap_sep(space, w_by)
bylen = len(by)
if bylen == 0:
raise OperationError(space.w_ValueError, space.wrap("empty
separator"))
@@ -348,8 +355,8 @@
return func_with_new_name(fn, funcname)
-str_rsplit__String_String_ANY =
make_rsplit_with_delim('str_rsplit__String_String_ANY',
- sliced)
+str_rsplit__String_ANY_ANY = make_rsplit_with_delim(
+ 'str_rsplit__String_ANY_ANY', sliced)
def str_join__String_ANY(space, w_self, w_list):
list_w = space.listview(w_list)
@@ -372,18 +379,22 @@
reslen = len(self) * (size - 1)
for i in range(size):
w_s = list_w[i]
- if not space.isinstance_w(w_s, space.w_bytes):
+ try:
+ item = space.bufferstr_w(w_s)
+ except OperationError, e:
+ if not e.match(space, space.w_TypeError):
+ raise
raise operationerrfmt(
space.w_TypeError,
"sequence item %d: expected bytes, %s "
"found", i, space.type(w_s).getname(space))
- reslen += len(space.bytes_w(w_s))
+ reslen += len(item)
sb = StringBuilder(reslen)
for i in range(size):
if self and i != 0:
sb.append(self)
- sb.append(space.bytes_w(list_w[i]))
+ sb.append(space.bufferstr_w(list_w[i]))
return space.wrapbytes(sb.build())
def str_rjust__String_ANY_ANY(space, w_self, w_arg, w_fillchar):
@@ -426,6 +437,11 @@
space, lenself, w_start, w_end, upper_bound=upper_bound)
return (self, start, end)
+def contains__String_ANY(space, w_self, w_sub):
+ self = w_self._value
+ sub = space.bufferstr_w(w_sub)
+ return space.newbool(self.find(sub) >= 0)
+
def contains__String_String(space, w_self, w_sub):
self = w_self._value
sub = w_sub._value
@@ -436,13 +452,23 @@
char = w_char.intval
return space.newbool(self.find(chr(char)) >= 0)
+def str_find__String_ANY_ANY_ANY(space, w_self, w_sub, w_start, w_end):
+ (self, start, end) = _convert_idx_params(space, w_self, w_start, w_end)
+ res = self.find(space.bufferstr_w(w_sub), start, end)
+ return space.wrap(res)
+
def str_find__String_String_ANY_ANY(space, w_self, w_sub, w_start, w_end):
- (self, start, end) = _convert_idx_params(space, w_self, w_start, w_end)
+ (self, start, end) = _convert_idx_params(space, w_self, w_start, w_end)
res = self.find(w_sub._value, start, end)
return space.wrap(res)
+def str_rfind__String_ANY_ANY_ANY(space, w_self, w_sub, w_start, w_end):
+ (self, start, end) = _convert_idx_params(space, w_self, w_start, w_end)
+ res = self.rfind(space.bufferstr_w(w_sub), start, end)
+ return space.wrap(res)
+
def str_rfind__String_String_ANY_ANY(space, w_self, w_sub, w_start, w_end):
- (self, start, end) = _convert_idx_params(space, w_self, w_start, w_end)
+ (self, start, end) = _convert_idx_params(space, w_self, w_start, w_end)
res = self.rfind(w_sub._value, start, end)
return space.wrap(res)
@@ -554,7 +580,7 @@
def _strip(space, w_self, w_chars, left, right):
"internal function called by str_xstrip methods"
u_self = w_self._value
- u_chars = w_chars._value
+ u_chars = space.bufferstr_w(w_chars)
lpos = 0
rpos = len(u_self)
@@ -590,20 +616,20 @@
assert rpos >= lpos # annotator hint, don't remove
return sliced(space, u_self, lpos, rpos, w_self)
-def str_strip__String_String(space, w_self, w_chars):
+def str_strip__String_ANY(space, w_self, w_chars):
return _strip(space, w_self, w_chars, left=1, right=1)
def str_strip__String_None(space, w_self, w_chars):
return _strip_none(space, w_self, left=1, right=1)
-def str_rstrip__String_String(space, w_self, w_chars):
+def str_rstrip__String_ANY(space, w_self, w_chars):
return _strip(space, w_self, w_chars, left=0, right=1)
def str_rstrip__String_None(space, w_self, w_chars):
return _strip_none(space, w_self, left=0, right=1)
-def str_lstrip__String_String(space, w_self, w_chars):
+def str_lstrip__String_ANY(space, w_self, w_chars):
return _strip(space, w_self, w_chars, left=1, right=0)
def str_lstrip__String_None(space, w_self, w_chars):
@@ -633,6 +659,12 @@
u_self, u_start, u_end = _convert_idx_params(space, w_self, w_start, w_end)
return wrapint(space, u_self.count(w_arg._value, u_start, u_end))
+def str_endswith__String_ANY_ANY_ANY(space, w_self, w_suffix, w_start, w_end):
+ (u_self, start, end) = _convert_idx_params(space, w_self, w_start,
+ w_end, True)
+ return space.newbool(stringendswith(u_self, space.bufferstr_w(w_suffix),
+ start, end))
+
def str_endswith__String_String_ANY_ANY(space, w_self, w_suffix, w_start,
w_end):
(u_self, start, end) = _convert_idx_params(space, w_self, w_start,
w_end, True)
@@ -642,15 +674,17 @@
(u_self, start, end) = _convert_idx_params(space, w_self, w_start,
w_end, True)
for w_suffix in space.fixedview(w_suffixes):
- if space.isinstance_w(w_suffix, space.w_unicode):
- w_u = space.call_function(space.w_unicode, w_self)
- return space.call_method(w_u, "endswith", w_suffixes, w_start,
- w_end)
- suffix = space.bytes_w(w_suffix)
+ suffix = space.bufferstr_w(w_suffix)
if stringendswith(u_self, suffix, start, end):
return space.w_True
return space.w_False
+def str_startswith__String_ANY_ANY_ANY(space, w_self, w_prefix, w_start,
w_end):
+ (u_self, start, end) = _convert_idx_params(space, w_self, w_start,
+ w_end, True)
+ return space.newbool(stringstartswith(u_self, space.bufferstr_w(w_prefix),
+ start, end))
+
def str_startswith__String_String_ANY_ANY(space, w_self, w_prefix, w_start,
w_end):
(u_self, start, end) = _convert_idx_params(space, w_self, w_start,
w_end, True)
@@ -660,11 +694,7 @@
(u_self, start, end) = _convert_idx_params(space, w_self,
w_start, w_end, True)
for w_prefix in space.fixedview(w_prefixes):
- if space.isinstance_w(w_prefix, space.w_unicode):
- w_u = space.call_function(space.w_unicode, w_self)
- return space.call_method(w_u, "startswith", w_prefixes, w_start,
- w_end)
- prefix = space.bytes_w(w_prefix)
+ prefix = space.bufferstr_w(w_prefix)
if stringstartswith(u_self, prefix, start, end):
return space.w_True
return space.w_False
diff --git a/pypy/objspace/std/stringtype.py b/pypy/objspace/std/stringtype.py
--- a/pypy/objspace/std/stringtype.py
+++ b/pypy/objspace/std/stringtype.py
@@ -365,6 +365,38 @@
W_StringObject.__init__(w_obj, chars)
return w_obj
+def descr_maketrans(space, w_type, w_from, w_to):
+ """bytes.maketrans(frm, to) -> translation table
+
+ Return a translation table (a bytes object of length 256) suitable
+ for use in the bytes or bytearray translate method where each byte
+ in frm is mapped to the byte at the same position in to.
+ The bytes objects frm and to must be of the same length."""
+ base_table = [chr(i) for i in range(256)]
+ list_from = makebytesdata_w(space, w_from)
+ list_to = makebytesdata_w(space, w_to)
+
+ if len(list_from) != len(list_to):
+ raise OperationError(space.w_ValueError, space.wrap(
+ "maketrans arguments must have same length"))
+
+ for i in range(len(list_from)):
+ pos_from = ord(list_from[i])
+ char_to = list_to[i]
+ base_table[pos_from] = char_to
+
+ chars = ''.join(base_table)
+ if space.config.objspace.std.withrope:
+ from pypy.objspace.std.ropeobject import rope, W_RopeObject
+ w_obj = space.allocate_instance(W_RopeObject, w_type)
+ W_RopeObject.__init__(w_obj, rope.LiteralStringNode(chars))
+ return w_obj
+ else:
+ from pypy.objspace.std.stringobject import W_StringObject
+ w_obj = space.allocate_instance(W_StringObject, w_type)
+ W_StringObject.__init__(w_obj, chars)
+ return w_obj
+
# ____________________________________________________________
str_typedef = StdTypeDef("bytes",
@@ -378,7 +410,8 @@
' - a text string encoded using the specified encoding\n'
' - a bytes or a buffer object\n'
' - any object implementing the buffer API.',
- fromhex = gateway.interp2app(descr_fromhex, as_classmethod=True)
+ fromhex = gateway.interp2app(descr_fromhex, as_classmethod=True),
+ maketrans = gateway.interp2app(descr_maketrans, as_classmethod=True),
)
str_typedef.registermethods(globals())
diff --git a/pypy/objspace/std/test/test_stringobject.py
b/pypy/objspace/std/test/test_stringobject.py
--- a/pypy/objspace/std/test/test_stringobject.py
+++ b/pypy/objspace/std/test/test_stringobject.py
@@ -728,6 +728,34 @@
x = b"A" * (2**16)
raises(OverflowError, x.replace, b'', x)
+ def test_compatibility(self):
+ #a whole bunch of methods should accept bytearray/memoryview without
complaining...
+ #I don't know how slavishly we should follow the cpython spec here,
since it appears
+ #quite arbitrary in which methods accept only bytes as secondary
arguments or
+ #anything with the buffer protocol
+
+ b = b'hello world'
+ b2 = b'ello'
+ #not testing result, just lack of TypeError
+ for bb in (b2, bytearray(b2), memoryview(b2)):
+ assert b.split(bb)
+ assert b.rsplit(bb)
+ assert b.split(bb[:1])
+ assert b.rsplit(bb[:1])
+ assert b.join((bb, bb)) # cpython accepts bytes and
+ # bytearray only, not buffer
+ assert bb in b
+ assert b.find(bb)
+ assert b.rfind(bb)
+ assert b.strip(bb)
+ assert b.rstrip(bb)
+ assert b.lstrip(bb)
+ assert not b.startswith(bb)
+ assert not b.startswith((bb, bb))
+ assert not b.endswith(bb)
+ assert not b.endswith((bb, bb))
+ assert b.maketrans(bb, bb)
+
class AppTestPrebuilt(AppTestStringObject):
def setup_class(cls):
cls.space = gettestobjspace(**{"objspace.std.withprebuiltchar": True})
_______________________________________________
pypy-commit mailing list
pypy-commit@python.org
http://mail.python.org/mailman/listinfo/pypy-commit
