resolve abnormal urls in compliance with rfc3986 (#1482)

jhy · Jul 9, 2021 · 8db724e · 8db724e
1 parent 661523f
commit 8db724e
Show file tree

Hide file tree

Showing 2 changed files with 26 additions and 3 deletions.
diff --git a/src/main/java/org/jsoup/internal/StringUtil.java b/src/main/java/org/jsoup/internal/StringUtil.java
@@ -9,6 +9,7 @@
 import java.util.Collection;
 import java.util.Iterator;
 import java.util.Stack;
+import java.util.regex.Pattern;
 
 /**
  A minimal String utility class. Designed for <b>internal</b> jsoup use only - the API and outcome may change without
@@ -259,6 +260,7 @@ public static boolean isAscii(String string) {
         return true;
     }
 
+    private static Pattern extraDotSegmentsPattern = Pattern.compile("^/((\\.{1,2}/)+)");
     /**
      * Create a new absolute URL, from a provided existing absolute URL and a relative URL component.
      * @param base the existing absolute base URL
@@ -271,10 +273,12 @@ public static URL resolve(URL base, String relUrl) throws MalformedURLException
         if (relUrl.startsWith("?"))
             relUrl = base.getPath() + relUrl;
         // workaround: //example.com + ./foo = //example.com/./foo, not //example.com/foo
-        if (relUrl.indexOf('.') == 0 && base.getFile().indexOf('/') != 0) {
-            base = new URL(base.getProtocol(), base.getHost(), base.getPort(), "/" + base.getFile());
+        URL url = new URL(base, relUrl);
+        String fixedFile = extraDotSegmentsPattern.matcher(url.getFile()).replaceFirst("/");
+        if (url.getRef() != null) {
+            fixedFile = fixedFile + "#" + url.getRef();
         }
-        return new URL(base, relUrl);
+        return new URL(url.getProtocol(), url.getHost(), url.getPort(), fixedFile);
     }
 
     /**

diff --git a/src/test/java/org/jsoup/internal/StringUtilTest.java b/src/test/java/org/jsoup/internal/StringUtilTest.java
@@ -99,6 +99,25 @@ public void join() {
         assertEquals("ftp://example.com/one", resolve("ftp://example.com/two/", "../one"));
         assertEquals("ftp://example.com/one/two.c", resolve("ftp://example.com/one/", "./two.c"));
         assertEquals("ftp://example.com/one/two.c", resolve("ftp://example.com/one/", "two.c"));
+        // examples taken from rfc3986 section 5.4.2
+        assertEquals("http://example.com/g", resolve("http://example.com/b/c/d;p?q", "../../../g"));
+        assertEquals("http://example.com/g", resolve("http://example.com/b/c/d;p?q", "../../../../g"));
+        assertEquals("http://example.com/g", resolve("http://example.com/b/c/d;p?q", "/./g"));
+        assertEquals("http://example.com/g", resolve("http://example.com/b/c/d;p?q", "/../g"));
+        assertEquals("http://example.com/b/c/g.", resolve("http://example.com/b/c/d;p?q", "g."));
+        assertEquals("http://example.com/b/c/.g", resolve("http://example.com/b/c/d;p?q", ".g"));
+        assertEquals("http://example.com/b/c/g..", resolve("http://example.com/b/c/d;p?q", "g.."));
+        assertEquals("http://example.com/b/c/..g", resolve("http://example.com/b/c/d;p?q", "..g"));
+        assertEquals("http://example.com/b/g", resolve("http://example.com/b/c/d;p?q", "./../g"));
+        assertEquals("http://example.com/b/c/g/", resolve("http://example.com/b/c/d;p?q", "./g/."));
+        assertEquals("http://example.com/b/c/g/h", resolve("http://example.com/b/c/d;p?q", "g/./h"));
+        assertEquals("http://example.com/b/c/h", resolve("http://example.com/b/c/d;p?q", "g/../h"));
+        assertEquals("http://example.com/b/c/g;x=1/y", resolve("http://example.com/b/c/d;p?q", "g;x=1/./y"));
+        assertEquals("http://example.com/b/c/y", resolve("http://example.com/b/c/d;p?q", "g;x=1/../y"));
+        assertEquals("http://example.com/b/c/g?y/./x", resolve("http://example.com/b/c/d;p?q", "g?y/./x"));
+        assertEquals("http://example.com/b/c/g?y/../x", resolve("http://example.com/b/c/d;p?q", "g?y/../x"));
+        assertEquals("http://example.com/b/c/g#s/./x", resolve("http://example.com/b/c/d;p?q", "g#s/./x"));
+        assertEquals("http://example.com/b/c/g#s/../x", resolve("http://example.com/b/c/d;p?q", "g#s/../x"));
     }
 
     @Test