archive/jdk
1
0
Fork 0
mirror of https://github.com/openjdk/jdk.git synced 2025-08-27 06:45:07 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

8281146: Replace StringCoding.hasNegatives with countPositives

Browse source 
Co-authored-by: Lutz Schmidt <lucy@openjdk.org>
Co-authored-by: Martin Doerr <mdoerr@openjdk.org>
Reviewed-by: kvn, lucy, rriggs
This commit is contained in:
Claes Redestad 2022-03-17 09:20:24 +00:00
parent 249d553659
commit beedae1141
36 changed files with 552 additions and 335 deletions
Download patch file Download diff file Expand all files Collapse all files

166
src/java.base/share/classes/java/lang/String.java
View file

@ -525,56 +525,63 @@ public final class String
this.value = "".value;
this.coder = "".coder;
} else if (charset == UTF_8.INSTANCE) {
if (COMPACT_STRINGS && !StringCoding.hasNegatives(bytes, offset, length)) {
this.value = Arrays.copyOfRange(bytes, offset, offset + length);
this.coder = LATIN1;
} else {
if (COMPACT_STRINGS) {
int dp = StringCoding.countPositives(bytes, offset, length);
if (dp == length) {
this.value = Arrays.copyOfRange(bytes, offset, offset + length);
this.coder = LATIN1;
return;
}
int sl = offset + length;
int dp = 0;
byte[] dst = null;
if (COMPACT_STRINGS) {
dst = new byte[length];
while (offset < sl) {
int b1 = bytes[offset];
if (b1 >= 0) {
dst[dp++] = (byte)b1;
byte[] dst = new byte[length];
if (dp > 0) {
System.arraycopy(bytes, offset, dst, 0, dp);
offset += dp;
}
while (offset < sl) {
int b1 = bytes[offset++];
if (b1 >= 0) {
dst[dp++] = (byte)b1;
continue;
}
if ((b1 & 0xfe) == 0xc2 && offset < sl) { // b1 either 0xc2 or 0xc3
int b2 = bytes[offset];
if (b2 < -64) { // continuation bytes are always negative values in the range -128 to -65
dst[dp++] = (byte)decode2(b1, b2);
offset++;
continue;
}
if ((b1 & 0xfe) == 0xc2 && offset + 1 < sl) { // b1 either 0xc2 or 0xc3
int b2 = bytes[offset + 1];
if (!isNotContinuation(b2)) {
dst[dp++] = (byte)decode2(b1, b2);
offset += 2;
continue;
}
}
// anything not a latin1, including the repl
// we have to go with the utf16
break;
}
if (offset == sl) {
if (dp != dst.length) {
dst = Arrays.copyOf(dst, dp);
}
this.value = dst;
this.coder = LATIN1;
return;
// anything not a latin1, including the REPL
// we have to go with the utf16
offset--;
break;
}
if (offset == sl) {
if (dp != dst.length) {
dst = Arrays.copyOf(dst, dp);
}
this.value = dst;
this.coder = LATIN1;
return;
}
if (dp == 0 || dst == null) {
dst = new byte[length << 1];
} else {
byte[] buf = new byte[length << 1];
StringLatin1.inflate(dst, 0, buf, 0, dp);
dst = buf;
}
byte[] buf = new byte[length << 1];
StringLatin1.inflate(dst, 0, buf, 0, dp);
dst = buf;
dp = decodeUTF8_UTF16(bytes, offset, sl, dst, dp, true);
if (dp != length) {
dst = Arrays.copyOf(dst, dp << 1);
}
this.value = dst;
this.coder = UTF16;
} else { // !COMPACT_STRINGS
byte[] dst = new byte[length << 1];
int dp = decodeUTF8_UTF16(bytes, offset, offset + length, dst, 0, true);
if (dp != length) {
dst = Arrays.copyOf(dst, dp << 1);
}
this.value = dst;
this.coder = UTF16;
}
} else if (charset == ISO_8859_1.INSTANCE) {
if (COMPACT_STRINGS) {
@ -682,41 +689,43 @@ public final class String
if (length == 0) {
return "";
}
if (COMPACT_STRINGS && !StringCoding.hasNegatives(bytes, offset, length)) {
return new String(Arrays.copyOfRange(bytes, offset, offset + length), LATIN1);
} else {
int dp;
byte[] dst;
if (COMPACT_STRINGS) {
dp = StringCoding.countPositives(bytes, offset, length);
int sl = offset + length;
int dp = 0;
byte[] dst = null;
if (COMPACT_STRINGS) {
dst = new byte[length];
while (offset < sl) {
int b1 = bytes[offset];
if (b1 >= 0) {
dst[dp++] = (byte) b1;
if (dp == length) {
return new String(Arrays.copyOfRange(bytes, offset, offset + length), LATIN1);
}
dst = new byte[length];
System.arraycopy(bytes, offset, dst, 0, dp);
offset += dp;
while (offset < sl) {
int b1 = bytes[offset++];
if (b1 >= 0) {
dst[dp++] = (byte)b1;
continue;
}
if ((b1 & 0xfe) == 0xc2 && offset < sl) { // b1 either 0xc2 or 0xc3
int b2 = bytes[offset];
if (b2 < -64) { // continuation bytes are always negative values in the range -128 to -65
dst[dp++] = (byte)decode2(b1, b2);
offset++;
continue;
}
if ((b1 & 0xfe) == 0xc2 && offset + 1 < sl) { // b1 either 0xc2 or 0xc3
int b2 = bytes[offset + 1];
if (!isNotContinuation(b2)) {
dst[dp++] = (byte) decode2(b1, b2);
offset += 2;
continue;
}
}
// anything not a latin1, including the REPL
// we have to go with the utf16
break;
}
if (offset == sl) {
if (dp != dst.length) {
dst = Arrays.copyOf(dst, dp);
}
return new String(dst, LATIN1);
}
// anything not a latin1, including the REPL
// we have to go with the utf16
offset--;
break;
}
if (dp == 0 || dst == null) {
if (offset == sl) {
if (dp != dst.length) {
dst = Arrays.copyOf(dst, dp);
}
return new String(dst, LATIN1);
}
if (dp == 0) {
dst = new byte[length << 1];
} else {
byte[] buf = new byte[length << 1];
@ -724,11 +733,14 @@ public final class String
dst = buf;
}
dp = decodeUTF8_UTF16(bytes, offset, sl, dst, dp, false);
if (dp != length) {
dst = Arrays.copyOf(dst, dp << 1);
}
return new String(dst, UTF16);
} else { // !COMPACT_STRINGS
dst = new byte[length << 1];
dp = decodeUTF8_UTF16(bytes, offset, offset + length, dst, 0, false);
}
if (dp != length) {
dst = Arrays.copyOf(dst, dp << 1);
}
return new String(dst, UTF16);
}
static String newStringNoRepl(byte[] src, Charset cs) throws CharacterCodingException {
@ -1019,17 +1031,9 @@ public final class String
*/
/* package-private */
static int decodeASCII(byte[] sa, int sp, char[] da, int dp, int len) {
if (!StringCoding.hasNegatives(sa, sp, len)) {
StringLatin1.inflate(sa, sp, da, dp, len);
return len;
} else {
int start = sp;
int end = sp + len;
while (sp < end && sa[sp] >= 0) {
da[dp++] = (char) sa[sp++];
}
return sp - start;
}
int count = StringCoding.countPositives(sa, sp, len);
StringLatin1.inflate(sa, sp, da, dp, count);
return count;
}
private static boolean isNotContinuation(int b) {