summaryrefslogtreecommitdiffstats
path: root/vespajlib
diff options
context:
space:
mode:
authorHenning Baldersheim <balder@yahoo-inc.com>2022-04-05 14:41:06 +0200
committerHenning Baldersheim <balder@yahoo-inc.com>2022-04-05 14:41:06 +0200
commit24983142a4747d092a255062d588aad82bbc9490 (patch)
treea08385b48a93bb1fa9740457b4115a8244925b0e /vespajlib
parentee806dbf2e9240a255271ccb890d80468a7b05f1 (diff)
Update abi and split in hot/cold path.
Diffstat (limited to 'vespajlib')
-rw-r--r--vespajlib/abi-spec.json1
-rw-r--r--vespajlib/src/main/java/com/yahoo/text/Text.java6
-rw-r--r--vespajlib/src/test/java/com/yahoo/text/TextTestCase.java44
3 files changed, 50 insertions, 1 deletions
diff --git a/vespajlib/abi-spec.json b/vespajlib/abi-spec.json
index 20c7d435964..e69631e8375 100644
--- a/vespajlib/abi-spec.json
+++ b/vespajlib/abi-spec.json
@@ -3232,6 +3232,7 @@
"methods": [
"public static boolean isTextCharacter(int)",
"public static java.util.OptionalInt validateTextString(java.lang.String)",
+ "public static boolean isValidTextString(java.lang.String)",
"public static boolean isDisplayable(int)",
"public static java.lang.String stripInvalidCharacters(java.lang.String)",
"public static java.lang.String truncate(java.lang.String, int)",
diff --git a/vespajlib/src/main/java/com/yahoo/text/Text.java b/vespajlib/src/main/java/com/yahoo/text/Text.java
index be34a9e3ab0..30eba3ebd65 100644
--- a/vespajlib/src/main/java/com/yahoo/text/Text.java
+++ b/vespajlib/src/main/java/com/yahoo/text/Text.java
@@ -48,7 +48,11 @@ public final class Text {
// The link above notes that 0x7F-0x84 and 0x86-0x9F are discouraged, but they are still allowed -
// see http://www.w3.org/International/questions/qa-controls
- if (codepoint < 0x80) return allowedAsciiChars[codepoint];
+ return (codepoint < 0x80)
+ ? allowedAsciiChars[codepoint]
+ : isTextCharAboveUsAscii(codepoint);
+ }
+ private static boolean isTextCharAboveUsAscii(int codepoint) {
if (codepoint < 0xFDD0) return true;
if (codepoint <= 0xFDDF) return false;
if (codepoint < 0x1FFFE) return true;
diff --git a/vespajlib/src/test/java/com/yahoo/text/TextTestCase.java b/vespajlib/src/test/java/com/yahoo/text/TextTestCase.java
index a2cb2158278..33274380aad 100644
--- a/vespajlib/src/test/java/com/yahoo/text/TextTestCase.java
+++ b/vespajlib/src/test/java/com/yahoo/text/TextTestCase.java
@@ -1,6 +1,7 @@
// Copyright Yahoo. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
package com.yahoo.text;
+import org.junit.Ignore;
import org.junit.Test;
import java.util.OptionalInt;
@@ -76,4 +77,47 @@ public class TextTestCase {
public void testFormat() {
assertEquals("foo 3.14", Text.format("%s %.2f", "foo", 3.1415926536));
}
+
+ private static long isValid(String [] strings, int num) {
+ long sum = 0;
+ for (int i=0; i < num; i++) {
+ if (Text.isValidTextString(strings[i%strings.length])) {
+ sum++;
+ }
+ }
+ return sum;
+ }
+ private static long validate(String [] strings, int num) {
+ long sum = 0;
+ for (int i=0; i < num; i++) {
+ if (Text.validateTextString(strings[i%strings.length]).isEmpty()) {
+ sum++;
+ }
+ }
+ return sum;
+ }
+
+ @Ignore
+ @Test
+ public void benchmarkValidate() {
+ String [] strings = new String[100];
+ for (int i=0; i < strings.length; i++) {
+ strings[i] = new StringBuilder("some text ").append(i).append("of mine.").appendCodePoint(0xDFFFC).append("foo").toString();
+ }
+ long sum = validate(strings, 1000000);
+ System.out.println("Warmup num validate = " + sum);
+ sum = isValid(strings, 1000000);
+ System.out.println("Warmup num isValid = " + sum);
+
+ long start = System.nanoTime();
+ sum = validate(strings, 100000000);
+ long diff = System.nanoTime() - start;
+ System.out.println("Validation num validate = " + sum + ". Took " + diff + "ns");
+
+ start = System.nanoTime();
+ sum = isValid(strings, 100000000);
+ diff = System.nanoTime() - start;
+ System.out.println("Validation num isValid = " + sum + ". Took " + diff + "ns");
+
+ }
}