Author: Carl Friedrich Bolz-Tereick <[email protected]>
Branch:
Changeset: r97415:b2971d2576c1
Date: 2019-09-10 13:18 +0200
http://bitbucket.org/pypy/pypy/changeset/b2971d2576c1/
Log: stop using codepoints_in_utf8 for the result of find
even for very small strings it's a lot slower than using binary
search
diff --git a/pypy/objspace/std/unicodeobject.py
b/pypy/objspace/std/unicodeobject.py
--- a/pypy/objspace/std/unicodeobject.py
+++ b/pypy/objspace/std/unicodeobject.py
@@ -891,6 +891,15 @@
return end - start
return rutf8.codepoints_in_utf8(self._utf8, start, end)
+ def _byte_to_index(self, bytepos):
+ """ this returns index such that self._index_to_byte(index) == bytepos
+ NB: this is slow! roughly logarithmic with a big constant
+ """
+ if self.is_ascii():
+ return bytepos
+ return rutf8.codepoint_index_at_byte_position(
+ self._utf8, self._get_index_storage(), bytepos)
+
@always_inline
def _unwrap_and_search(self, space, w_sub, w_start, w_end, forward=True):
w_sub = self.convert_arg_to_w_unicode(space, w_sub)
@@ -912,16 +921,14 @@
res_index = self._utf8.find(w_sub._utf8, start_index, end_index)
if res_index < 0:
return None
- skip = self._codepoints_in_utf8(start_index, res_index)
- res = start + skip
+ res = self._byte_to_index(res_index)
assert res >= 0
return space.newint(res)
else:
res_index = self._utf8.rfind(w_sub._utf8, start_index, end_index)
if res_index < 0:
return None
- skip = self._codepoints_in_utf8(res_index, end_index)
- res = end - skip
+ res = self._byte_to_index(res_index)
assert res >= 0
return space.newint(res)
_______________________________________________
pypy-commit mailing list
[email protected]
https://mail.python.org/mailman/listinfo/pypy-commit
