github_mirrors/vectorscan
3
0
Fork 1
mirror of https://github.com/VectorCamp/vectorscan.git synced 2025-09-29 19:24:25 +03:00
Code Issues Packages Projects Releases Wiki Activity

character classes: handle \Q\E and utf8

Browse Source
This commit is contained in:
Alex Coyte
2017-06-20 10:19:32 +10:00
committed by Matthew Barr
parent a185be5a4f
commit d317d75615
2 changed files with 32 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
src/parser/Parser.rl
View File

@@ -1184,6 +1184,11 @@ unichar readUtf8CodePoint4c(const char *s) {
currentSeq->addComponent(move(cc));
};
hi_byte when is_utf8 => {
assert(mode.utf8);
throwInvalidUtf8();
};
# Literal character
any => {
addLiteral(currentSeq, *ts, mode);
@@ -1198,6 +1203,31 @@ unichar readUtf8CodePoint4c(const char *s) {
'\\E' => {
fret;
};
#unicode chars
utf8_2c when is_utf8 => {
assert(mode.utf8);
currentCls->add(readUtf8CodePoint2c(ts));
inCharClassEarly = false;
};
utf8_3c when is_utf8 => {
assert(mode.utf8);
currentCls->add(readUtf8CodePoint3c(ts));
inCharClassEarly = false;
};
utf8_4c when is_utf8 => {
assert(mode.utf8);
currentCls->add(readUtf8CodePoint4c(ts));
inCharClassEarly = false;
};
hi_byte when is_utf8 => {
assert(mode.utf8);
throwInvalidUtf8();
};
# Literal character
any => {
currentCls->add(*ts);