archive/jdk
1
0
Fork 0
mirror of https://github.com/openjdk/jdk.git synced 2025-08-27 14:54:52 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

8302871: Speed up StringLatin1.regionMatchesCI

Browse source 
Reviewed-by: redestad, martin, alanb
This commit is contained in:
Eirik Bjorsnos 2023-02-25 07:48:03 +00:00 • committed by Alan Bateman
parent b4ea80731c
commit 17e3769ed7
4 changed files with 161 additions and 18 deletions
Download patch file Download diff file Expand all files Collapse all files

39
src/java.base/share/classes/java/lang/CharacterDataLatin1.java.template
View file

@ -138,10 +138,11 @@ class CharacterDataLatin1 extends CharacterData {
if (ch < 'A') { // Fast path for low code points
return ch;
}
int l = ch | 0x20; // Lowercase using 'oldest ASCII trick in the book'
if (l <= 'z' // In range a-z
|| (l >= 0xE0 && l <= 0xFE && l != 0xF7)) { // ..or agrave-thorn, excluding division
return l;
// ASCII and Latin-1 were designed to optimize case-twiddling operations
int lower = ch | 0x20;
if (lower <= 'z' // In range a-z
|| (lower >= 0xE0 && lower <= 0xFE && lower != 0xF7)) { // ..or agrave-thorn, excluding division
return lower;
}
return ch;
}
@ -150,10 +151,11 @@ class CharacterDataLatin1 extends CharacterData {
if (ch < 'a') { // Fast path for low code points
return ch;
}
int U = ch & 0xDF; // Uppercase using 'oldest ASCII trick in the book'
if (U <= 'Z' // In range A-Z
|| (U >= 0xC0 && U <= 0xDE && U != 0xD7)) { // ..or Agrave-Thorn, excluding multiplication
return U;
// ASCII and Latin-1 were designed to optimize case-twiddling operations
int upper = ch & 0xDF;
if (upper <= 'Z' // In range A-Z
|| (upper >= 0xC0 && upper <= 0xDE && upper != 0xD7)) { // ..or Agrave-Thorn, not multiplication
return upper;
}
// Special-case for 'y with Diaeresis' which uppercases out of latin1
@ -167,6 +169,27 @@ class CharacterDataLatin1 extends CharacterData {
return ch;
}
/**
* Compares two latin1 code points, ignoring case considerations
*
* @param b1 byte representing a latin1 code point
* @param b2 another byte representing a latin1 code point
* @return true if the two bytes are considered equals ignoring case in latin1
*/
static boolean equalsIgnoreCase(byte b1, byte b2) {
if (b1 == b2) {
return true;
}
// ASCII and Latin-1 were designed to optimize case-twiddling operations
int upper = b1 & 0xDF;
if (upper < 'A') {
return false; // Low ASCII
}
return (upper <= 'Z' // In range A-Z
|| (upper >= 0xC0 && upper <= 0XDE && upper != 0xD7)) // ..or A-grave-Thorn, not multiplication
&& upper == (b2 & 0xDF); // b2 has same uppercase
}
int toTitleCase(int ch) {
return toUpperCase(ch);
}

11
src/java.base/share/classes/java/lang/StringLatin1.java
View file

@ -384,14 +384,9 @@ final class StringLatin1 {
byte[] other, int ooffset, int len) {
int last = toffset + len;
while (toffset < last) {
char c1 = (char)(value[toffset++] & 0xff);
char c2 = (char)(other[ooffset++] & 0xff);
if (c1 == c2) {
continue;
}
int u1 = CharacterDataLatin1.instance.toUpperCase(c1);
int u2 = CharacterDataLatin1.instance.toUpperCase(c2);
if (u1 == u2) {
byte b1 = value[toffset++];
byte b2 = other[ooffset++];
if (CharacterDataLatin1.equalsIgnoreCase(b1, b2)) {
continue;
}
return false;