Backport: Feature [feature-requests:#1259]. Add SCI_SETCHARACTERCATEGORYOPTIMIZATION API to optimize speed of character category features.

Backport of changeset 7392:2832adedd0f4, but with added includes for Sci::clamp().

1 files changed, 33 insertions, 0 deletions

diff --git a/lexlib/CharacterCategory.cxx b/lexlib/CharacterCategory.cxx
index bc2fa2336..19c44cabe 100644
--- a/lexlib/CharacterCategory.cxx
+++ b/lexlib/CharacterCategory.cxx
@@ -7,10 +7,13 @@
 // Copyright 2013 by Neil Hodgson <neilh@scintilla.org>
 // The License.txt file describes the conditions under which this software may be distributed.
 
+#include <vector>
 #include <algorithm>
 #include <iterator>
 
+#include "Scintilla.h" // for ptrdiff_t in Position.h
 #include "CharacterCategory.h"
+#include "Position.h" // for Sci::clamp
 
 namespace Scintilla {
 
@@ -3790,6 +3793,7 @@ const int catRanges[] = {
 33554397,
 33554460,
 35651549,
+35651613,
 //--Autogenerated -- end of section automatically generated
 };
 
@@ -3963,4 +3967,33 @@ bool IsXidContinue(int character) {
 	}
 }
 
+CharacterCategoryMap::CharacterCategoryMap() noexcept {
+	Optimize(256);
+}
+
+int CharacterCategoryMap::Size() const noexcept {
+	return static_cast<int>(dense.size());
+}
+
+void CharacterCategoryMap::Optimize(int countCharacters) {
+	const int characters = Sci::clamp(countCharacters, 256, maxUnicode + 1);
+	dense.resize(characters);
+
+	int end = 0;
+	int index = 0;
+	int current = catRanges[index];
+	++index;
+	do {
+		const int next = catRanges[index];
+		const unsigned char category = current & maskCategory;
+		current >>= 5;
+		end = std::min(characters, next >> 5);
+		while (current < end) {
+			dense[current++] = category;
+		}
+		current = next;
+		++index;
+	} while (characters > end);
+}
+
 }