Implement escapeURIComponent for JRuby

ruby · Nov 7, 2023 · c90c9ea · c90c9ea
1 parent a60754b
commit c90c9ea
Showing 1 changed file with 48 additions and 7 deletions.
diff --git a/ext/java/org/jruby/ext/cgi/escape/CGIEscape.java b/ext/java/org/jruby/ext/cgi/escape/CGIEscape.java
@@ -244,7 +244,7 @@ static boolean ISXDIGIT(byte[] cstrBytes, int i) {
         return (cstrByte >= '0' && cstrByte <= '9') || (cstrByte >= 'a' && cstrByte <= 'f') || (cstrByte >= 'A' && cstrByte <= 'F');
     }
 
-    static boolean url_unreserved_char(int c) {
+    static boolean url_unreserved_char(int c, boolean escapePlus) {
         switch (c) {
             case '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': case '8': case '9':
             case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g': case 'h': case 'i': case 'j':
@@ -255,6 +255,8 @@ static boolean url_unreserved_char(int c) {
             case 'U': case 'V': case 'W': case 'X': case 'Y': case 'Z':
             case '-': case '.': case '_': case '~':
                 return true;
+            case ' ':
+                return !escapePlus;
             default:
                 break;
         }
@@ -263,7 +265,7 @@ static boolean url_unreserved_char(int c) {
 
     static final byte[] upper_hexdigits = "0123456789ABCDEF".getBytes(RubyEncoding.UTF8);
 
-    static IRubyObject optimized_escape(Ruby runtime, RubyString str) {
+    static IRubyObject optimized_escape(Ruby runtime, RubyString str, boolean escapePlus) {
         int i, len, beg = 0;
         RubyString dest = null;
         byte[] cstrBytes;
@@ -277,7 +279,7 @@ static IRubyObject optimized_escape(Ruby runtime, RubyString str) {
 
         for (i = 0; i < len; ++i) {
             int c = cstrBytes[cstr + i] & 0xFF;
-            if (!url_unreserved_char(c)) {
+            if (!url_unreserved_char(c, escapePlus)) {
                 if (dest == null) {
                     dest = RubyString.newStringLight(runtime, len);
                 }
@@ -305,7 +307,7 @@ static IRubyObject optimized_escape(Ruby runtime, RubyString str) {
     }
 
     static IRubyObject
-    optimized_unescape(ThreadContext context, RubyString str, IRubyObject encoding) {
+    optimized_unescape(ThreadContext context, RubyString str, IRubyObject encoding, boolean unescapePlus) {
         int i, len, beg = 0;
         RubyString dest = null;
         byte[] cstrBytes;
@@ -331,7 +333,7 @@ static IRubyObject optimized_escape(Ruby runtime, RubyString str) {
                 buf = ((char_to_number(cstrBytes[cstr + i + 1]) << 4)
                         | char_to_number(cstrBytes[cstr + i + 2]));
                 clen = 2;
-            } else if (c == '+') {
+            } else if (unescapePlus && c == '+') {
                 buf = ' ';
             } else {
                 continue;
@@ -416,7 +418,25 @@ public static IRubyObject cgiesc_escape(ThreadContext context, IRubyObject self,
         RubyString str = _str.convertToString();
 
         if (str.getEncoding().isAsciiCompatible()) {
-            return optimized_escape(context.runtime, str);
+            return optimized_escape(context.runtime, str, true);
+        } else {
+            return Helpers.invokeSuper(context, self, _str, Block.NULL_BLOCK);
+        }
+    }
+
+    /*
+     *  call-seq:
+     *     CGI.escapeURIComponent(string) -> string
+     *
+     *  Returns URL-escaped string following RFC 3986.
+     *
+     */
+    @JRubyMethod(name = "escapeURIComponent", alias = { "escape_uri_component" }, module = true, frame = true)
+    public static IRubyObject cgiesc_escape_uri_component(ThreadContext context, IRubyObject self, IRubyObject _str) {
+        RubyString str = _str.convertToString();
+
+        if (str.getEncoding().isAsciiCompatible()) {
+            return optimized_escape(context.runtime, str, false);
         } else {
             return Helpers.invokeSuper(context, self, _str, Block.NULL_BLOCK);
         }
@@ -443,7 +463,28 @@ public static IRubyObject cgiesc_unescape(ThreadContext context, IRubyObject sel
 
         if (str.getEncoding().isAsciiCompatible()) {
             IRubyObject enc = accept_charset(argv, argv.length - 1, 1, self);
-            return optimized_unescape(context, str, enc);
+            return optimized_unescape(context, str, enc, true);
+        } else {
+            return Helpers.invokeSuper(context, self, argv, Block.NULL_BLOCK);
+        }
+    }
+
+    /*
+     *  call-seq:
+     *     CGI.unescapeURIComponent(string, encoding=@@accept_charset) -> string
+     *
+     *  Returns URL-unescaped string following RFC 3986.
+     *
+     */
+    @JRubyMethod(name = "unescapeURIComponent", alias = { "unescape_uri_component" }, required = 1, optional = 1, module = true, frame = true)
+    public static IRubyObject cgiesc_unescape_uri_component(ThreadContext context, IRubyObject self, IRubyObject[] argv) {
+        IRubyObject _str = argv[0];
+
+        RubyString str = _str.convertToString();
+
+        if (str.getEncoding().isAsciiCompatible()) {
+            IRubyObject enc = accept_charset(argv, argv.length - 1, 1, self);
+            return optimized_unescape(context, str, enc, false);
         } else {
             return Helpers.invokeSuper(context, self, argv, Block.NULL_BLOCK);
         }