| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
git-svn-id: svn://vcs.exim.org/pcre/code/trunk@1133 2f5784b3-3f2a-0410-8824-cb99058d5e15
|
|
|
|
|
|
|
|
|
| |
The code replaced surrogates and characters > 10ffff with fffe, for no apparent
reason. Taking this code out still lets the test suite pass. We cannot assert
however that the character is a valid unicode character since we do need to handle
e.g. 10ffff here.
git-svn-id: svn://vcs.exim.org/pcre/code/trunk@1099 2f5784b3-3f2a-0410-8824-cb99058d5e15
|
|
|
|
|
|
|
|
|
|
| |
UTF-32 only uses 21 bits, so the upper bits may be used to store flags
etc. To allow passing the unmodified internal buffers to pcre32, make
pcre32 mask out those upper bits.
TODO: do the same for the JIT compiler, and add tests
git-svn-id: svn://vcs.exim.org/pcre/code/trunk@1083 2f5784b3-3f2a-0410-8824-cb99058d5e15
|
|
Create libpcre32 that operates on 32-bit characters (UTF-32).
This turned out to be surprisingly simple after the UTF-16 support
was introduced; mostly just extra ifdefs and adjusting and adding
some tests.
git-svn-id: svn://vcs.exim.org/pcre/code/trunk@1055 2f5784b3-3f2a-0410-8824-cb99058d5e15
|