mirror of
https://github.com/gbevin/urlencoder.git
synced 2025-04-25 07:17:11 -07:00
First commit
This commit is contained in:
commit
2d7f91cbd7
12 changed files with 913 additions and 0 deletions
76
lib/build.gradle.kts
Normal file
76
lib/build.gradle.kts
Normal file
|
|
@ -0,0 +1,76 @@
|
|||
plugins {
|
||||
`java-library`
|
||||
`maven-publish`
|
||||
signing
|
||||
}
|
||||
|
||||
base {
|
||||
archivesName.set("urlencoder")
|
||||
version = 0.9
|
||||
}
|
||||
java {
|
||||
withJavadocJar()
|
||||
withSourcesJar()
|
||||
toolchain {
|
||||
languageVersion.set(JavaLanguageVersion.of(17))
|
||||
}
|
||||
}
|
||||
|
||||
repositories {
|
||||
mavenCentral()
|
||||
}
|
||||
|
||||
dependencies {
|
||||
testImplementation("org.junit.jupiter:junit-jupiter:5.9.0")
|
||||
}
|
||||
|
||||
tasks.named<Test>("test") {
|
||||
useJUnitPlatform()
|
||||
}
|
||||
|
||||
publishing {
|
||||
publications {
|
||||
create<MavenPublication>("mavenJava") {
|
||||
artifactId = "urlencoder"
|
||||
from(components["java"])
|
||||
pom {
|
||||
name.set("URL Encoder")
|
||||
description.set("A simple library to encode/decode URL parameters.")
|
||||
url.set("https://github.com/gbevin/urlencoder")
|
||||
licenses {
|
||||
license {
|
||||
name.set("The Apache License, Version 2.0")
|
||||
url.set("http://www.apache.org/licenses/LICENSE-2.0.txt")
|
||||
}
|
||||
}
|
||||
developers {
|
||||
developer {
|
||||
id.set("gbevin")
|
||||
name.set("Geert Bevin")
|
||||
email.set("gbevin@uwyn.com")
|
||||
}
|
||||
}
|
||||
scm {
|
||||
connection.set("scm:git:https://github.com/gbevin/urlencoder.git")
|
||||
developerConnection.set("scm:git:git@github.com:gbevin/urlencoder.git")
|
||||
url.set("https://github.com/gbevin/urlencoder")
|
||||
}
|
||||
}
|
||||
repositories {
|
||||
maven {
|
||||
credentials {
|
||||
username = project.properties["ossrhUsername"].toString()
|
||||
password = project.properties["ossrhPassword"].toString()
|
||||
}
|
||||
val releasesRepoUrl = uri("https://s01.oss.sonatype.org/service/local/staging/deploy/maven2/")
|
||||
val snapshotsRepoUrl = uri("https://s01.oss.sonatype.org/content/repositories/snapshots/")
|
||||
url = if (version.toString().endsWith("SNAPSHOT")) snapshotsRepoUrl else releasesRepoUrl
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
signing {
|
||||
sign(publishing.publications["mavenJava"])
|
||||
}
|
||||
187
lib/src/main/java/com/uwyn/urlencoder/UrlEncoder.java
Normal file
187
lib/src/main/java/com/uwyn/urlencoder/UrlEncoder.java
Normal file
|
|
@ -0,0 +1,187 @@
|
|||
/*
|
||||
* Copyright 2001-2022 Geert Bevin (gbevin[remove] at uwyn dot com)
|
||||
* Licensed under the Apache License, Version 2.0 (the "License")
|
||||
*/
|
||||
package com.uwyn.urlencoder;
|
||||
|
||||
import java.nio.charset.StandardCharsets;
|
||||
import java.util.BitSet;
|
||||
|
||||
/**
|
||||
* URL encoding and decoding.
|
||||
* <p>
|
||||
* Rules determined by <a href="https://www.rfc-editor.org/rfc/rfc3986#page-13">RFC 3986</a>.
|
||||
*
|
||||
* @author Geert Bevin (gbevin[remove] at uwyn dot com)
|
||||
* @since 1.0
|
||||
*/
|
||||
public class UrlEncoder {
|
||||
/**
|
||||
* Transforms a provided <code>String</code> object into a new string,
|
||||
* containing only valid URL characters in the UTF-8 encoding.
|
||||
*
|
||||
* @param source The string that has to be transformed into a valid URL
|
||||
* string.
|
||||
* @return The encoded <code>String</code> object.
|
||||
* @see #decode(String)
|
||||
* @since 1.0
|
||||
*/
|
||||
public static String encode(String source) {
|
||||
if (source == null) {
|
||||
return null;
|
||||
}
|
||||
|
||||
StringBuilder out = null;
|
||||
char ch;
|
||||
for (var i = 0; i < source.length(); ) {
|
||||
ch = source.charAt(i);
|
||||
if (isUnreservedUriChar(ch)) {
|
||||
if (out != null) {
|
||||
out.append(ch);
|
||||
}
|
||||
i += 1;
|
||||
} else {
|
||||
if (out == null) {
|
||||
out = new StringBuilder(source.length());
|
||||
out.append(source, 0, i);
|
||||
}
|
||||
|
||||
var cp = source.codePointAt(i);
|
||||
if (cp < 0x80) {
|
||||
appendUrlEncodedByte(out, cp);
|
||||
i += 1;
|
||||
} else if (Character.isBmpCodePoint(cp)) {
|
||||
for (var b : Character.toString(ch).getBytes(StandardCharsets.UTF_8)) {
|
||||
appendUrlEncodedByte(out, b);
|
||||
}
|
||||
i += 1;
|
||||
} else if (Character.isSupplementaryCodePoint(cp)) {
|
||||
var high = Character.highSurrogate(cp);
|
||||
var low = Character.lowSurrogate(cp);
|
||||
for (var b : new String(new char[]{high, low}).getBytes(StandardCharsets.UTF_8)) {
|
||||
appendUrlEncodedByte(out, b);
|
||||
}
|
||||
i += 2;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
if (out == null) {
|
||||
return source;
|
||||
}
|
||||
|
||||
return out.toString();
|
||||
}
|
||||
|
||||
static final BitSet UNRESERVED_URI_CHARS;
|
||||
|
||||
static {
|
||||
// see https://www.rfc-editor.org/rfc/rfc3986#page-13
|
||||
var unreserved = new BitSet('~' + 1);
|
||||
unreserved.set('-');
|
||||
unreserved.set('.');
|
||||
for (int c = '0'; c <= '9'; ++c) unreserved.set(c);
|
||||
for (int c = 'A'; c <= 'Z'; ++c) unreserved.set(c);
|
||||
unreserved.set('_');
|
||||
for (int c = 'a'; c <= 'z'; ++c) unreserved.set(c);
|
||||
unreserved.set('~');
|
||||
UNRESERVED_URI_CHARS = unreserved;
|
||||
}
|
||||
|
||||
// see https://www.rfc-editor.org/rfc/rfc3986#page-13
|
||||
private static boolean isUnreservedUriChar(char ch) {
|
||||
if (ch > '~') return false;
|
||||
return UNRESERVED_URI_CHARS.get(ch);
|
||||
}
|
||||
|
||||
private static final char[] HEX_DIGITS = "0123456789ABCDEF".toCharArray();
|
||||
|
||||
private static void appendUrlEncodedDigit(StringBuilder out, int digit) {
|
||||
out.append(HEX_DIGITS[digit & 0x0F]);
|
||||
}
|
||||
|
||||
private static void appendUrlEncodedByte(StringBuilder out, int ch) {
|
||||
out.append("%");
|
||||
appendUrlEncodedDigit(out, ch >> 4);
|
||||
appendUrlEncodedDigit(out, ch);
|
||||
}
|
||||
|
||||
/**
|
||||
* Transforms a provided <code>String</code> URL into a new string,
|
||||
* containing decoded URL characters in the UTF-8 encoding.
|
||||
*
|
||||
* @param source The string URL that has to be decoded
|
||||
* @return The decoded <code>String</code> object.
|
||||
* @see #encode(String)
|
||||
* @since 1.0
|
||||
*/
|
||||
public static String decode(String source) {
|
||||
if (source == null) {
|
||||
return source;
|
||||
}
|
||||
|
||||
var length = source.length();
|
||||
StringBuilder out = null;
|
||||
char ch;
|
||||
byte[] bytes_buffer = null;
|
||||
var bytes_pos = 0;
|
||||
for (var i = 0; i < length; ) {
|
||||
ch = source.charAt(i);
|
||||
|
||||
if (ch == '%') {
|
||||
if (out == null) {
|
||||
out = new StringBuilder(source.length());
|
||||
out.append(source, 0, i);
|
||||
}
|
||||
|
||||
if (bytes_buffer == null) {
|
||||
// the remaining characters divided by the length
|
||||
// of the encoding format %xx, is the maximum number of
|
||||
// bytes that can be extracted
|
||||
bytes_buffer = new byte[(length - i) / 3];
|
||||
bytes_pos = 0;
|
||||
}
|
||||
|
||||
i += 1;
|
||||
if (length < i + 2) {
|
||||
throw new IllegalArgumentException("Illegal escape sequence");
|
||||
}
|
||||
try {
|
||||
var v = Integer.parseInt(source, i, i + 2, 16);
|
||||
if (v < 0 || v > 0xFF) {
|
||||
throw new IllegalArgumentException("Illegal escape value");
|
||||
}
|
||||
|
||||
bytes_buffer[bytes_pos++] = (byte) v;
|
||||
|
||||
i += 2;
|
||||
} catch (NumberFormatException e) {
|
||||
throw new IllegalArgumentException("Illegal characters in escape sequence" + e.getMessage());
|
||||
}
|
||||
} else {
|
||||
if (bytes_buffer != null) {
|
||||
out.append(new String(bytes_buffer, 0, bytes_pos, StandardCharsets.UTF_8));
|
||||
|
||||
bytes_buffer = null;
|
||||
bytes_pos = 0;
|
||||
}
|
||||
|
||||
if (out != null) {
|
||||
out.append(ch);
|
||||
}
|
||||
|
||||
i += 1;
|
||||
}
|
||||
}
|
||||
|
||||
if (out == null) {
|
||||
return source;
|
||||
}
|
||||
|
||||
if (bytes_buffer != null) {
|
||||
out.append(new String(bytes_buffer, 0, bytes_pos, StandardCharsets.UTF_8));
|
||||
}
|
||||
|
||||
return out.toString();
|
||||
}
|
||||
}
|
||||
53
lib/src/test/java/com/uwyn/urlencoder/UrlEncoderTest.java
Normal file
53
lib/src/test/java/com/uwyn/urlencoder/UrlEncoderTest.java
Normal file
|
|
@ -0,0 +1,53 @@
|
|||
/*
|
||||
* Copyright 2001-2022 Geert Bevin (gbevin[remove] at uwyn dot com)
|
||||
* Licensed under the Apache License, Version 2.0 (the "License")
|
||||
*/
|
||||
package com.uwyn.urlencoder;
|
||||
|
||||
import org.junit.jupiter.api.Test;
|
||||
|
||||
import static org.junit.jupiter.api.Assertions.*;
|
||||
import static org.junit.jupiter.api.Assertions.assertTrue;
|
||||
|
||||
class UrlEncoderTest {
|
||||
@Test
|
||||
public void testEncodeURL() {
|
||||
assertNull(UrlEncoder.encode(null));
|
||||
assertEquals("a%20test%20%26", UrlEncoder.encode("a test &"));
|
||||
String valid = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVQXYZ0123456789-_.~";
|
||||
assertSame(valid, UrlEncoder.encode(valid));
|
||||
assertEquals("%21abcdefghijklmnopqrstuvwxyz%25%25ABCDEFGHIJKLMNOPQRSTUVQXYZ0123456789-_.~%3D", UrlEncoder.encode("!abcdefghijklmnopqrstuvwxyz%%ABCDEFGHIJKLMNOPQRSTUVQXYZ0123456789-_.~="));
|
||||
assertEquals("%25%23ok%C3%A9k%C3%89%C8%A2%20smile%21%F0%9F%98%81", UrlEncoder.encode("%#okékÉȢ smile!😁"));
|
||||
}
|
||||
|
||||
@Test
|
||||
public void testDecodeURL() {
|
||||
assertNull(UrlEncoder.decode(null));
|
||||
assertEquals("a test &", UrlEncoder.decode("a%20test%20%26"));
|
||||
String valid = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVQXYZ0123456789-_.~";
|
||||
assertSame(valid, UrlEncoder.decode(valid));
|
||||
assertEquals("!abcdefghijklmnopqrstuvwxyz%%ABCDEFGHIJKLMNOPQRSTUVQXYZ0123456789-_.~=", UrlEncoder.decode("%21abcdefghijklmnopqrstuvwxyz%25%25ABCDEFGHIJKLMNOPQRSTUVQXYZ0123456789-_.~%3D"));
|
||||
assertEquals("%#okékÉȢ smile!😁", UrlEncoder.decode("%25%23ok%C3%A9k%C3%89%C8%A2%20smile%21%F0%9F%98%81"));
|
||||
|
||||
try {
|
||||
UrlEncoder.decode("sdkjfh%");
|
||||
fail();
|
||||
} catch (Exception e) {
|
||||
assertTrue(e instanceof IllegalArgumentException);
|
||||
}
|
||||
|
||||
try {
|
||||
UrlEncoder.decode("sdkjfh%6");
|
||||
fail();
|
||||
} catch (Exception e) {
|
||||
assertTrue(e instanceof IllegalArgumentException);
|
||||
}
|
||||
|
||||
try {
|
||||
UrlEncoder.decode("sdkjfh%xx");
|
||||
fail();
|
||||
} catch (Exception e) {
|
||||
assertTrue(e instanceof IllegalArgumentException);
|
||||
}
|
||||
}
|
||||
}
|
||||
Loading…
Add table
Add a link
Reference in a new issue