From: Alex Coyte <a.coyte@intel.com>
Date: Mon, 19 Jun 2017 01:03:05 +0000 (+1000)
Subject: Treat characters between \Q \E as codepoints in UTF8 mode.
X-Git-Tag: v4.5.2^2~5
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=a185be5a4f684c9bdbd90a2b9716ca02dde9e7b2;p=thirdparty%2Fvectorscan.git

Treat characters between \Q \E as codepoints in UTF8 mode.

fixes github issue #57
---

diff --git a/src/parser/Parser.rl b/src/parser/Parser.rl
index 52b3340c..05a084bb 100644
--- a/src/parser/Parser.rl
+++ b/src/parser/Parser.rl
@@ -1155,6 +1155,35 @@ unichar readUtf8CodePoint4c(const char *s) {
               '\\E' => {
                   fgoto main;
               };
+
+              #unicode chars
+              utf8_2c when is_utf8 => {
+                  assert(mode.utf8);
+                  /* leverage ComponentClass to generate the vertices */
+                  auto cc = getComponentClass(mode);
+                  cc->add(readUtf8CodePoint2c(ts));
+                  cc->finalize();
+                  currentSeq->addComponent(move(cc));
+              };
+
+              utf8_3c when is_utf8 => {
+                  assert(mode.utf8);
+                  /* leverage ComponentClass to generate the vertices */
+                  auto cc = getComponentClass(mode);
+                  cc->add(readUtf8CodePoint3c(ts));
+                  cc->finalize();
+                  currentSeq->addComponent(move(cc));
+              };
+
+              utf8_4c when is_utf8 => {
+                  assert(mode.utf8);
+                  /* leverage ComponentClass to generate the vertices */
+                  auto cc = getComponentClass(mode);
+                  cc->add(readUtf8CodePoint4c(ts));
+                  cc->finalize();
+                  currentSeq->addComponent(move(cc));
+              };
+
               # Literal character
               any => {
                   addLiteral(currentSeq, *ts, mode);