BaseUrlValidateFunction.java revision 56ed4167b942ec265f9cee70ac4d71d10b3835ce 
1/*
2 * Copyright (C) 2010 Google Inc.
3 *
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
7 *
8 * http://www.apache.org/licenses/LICENSE-2.0
9 *
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
15 */
16
17package com.google.clearsilver.jsilver.functions.html;
18
19import com.google.clearsilver.jsilver.functions.TextFilter;
20
21import java.io.IOException;
22import java.lang.Character.UnicodeBlock;
23
24/**
25 * Validates that a given string is either something that looks like a relative URI, or looks like
26 * an absolute URI using one of a set of allowed schemes (http, https, ftp, mailto). If the string
27 * is valid according to these criteria, the string is escaped with an appropriate escaping
28 * function. Otherwise, the string "#" is returned.
29 *
30 * Subclasses will apply the necessary escaping function to the string by overriding {@code
31 * applyEscaping}.
32 *
33 * <p>
34 * Note: this function does <em>not</em> validate that the URI is well-formed beyond the scheme part
35 * (and if the URI appears to be relative, not even then). Note in particular that this function
36 * considers strings of the form "www.google.com:80" to be invalid.
37 */
38public abstract class BaseUrlValidateFunction implements TextFilter {
39
40  @Override
41  public void filter(String in, Appendable out) throws IOException {
42    if (!isValidUri(in)) {
43      out.append('#');
44      return;
45    }
46    applyEscaping(in, out);
47  }
48
49  /**
50   * Called by {@code filter} after verifying that the input is a valid URI. Should apply any
51   * appropriate escaping to the input string.
52   *
53   * @throws IOException
54   */
55  protected abstract void applyEscaping(String in, Appendable out) throws IOException;
56
57  /**
58   * @return true if a given string either looks like a relative URI, or like an absolute URI with
59   *         an allowed scheme.
60   */
61  protected boolean isValidUri(String in) {
62    // Quick check for the allowed absolute URI schemes.
63    String maybeScheme = toLowerCaseAsciiOnly(in.substring(0, Math.min(in.length(), 8)));
64    if (maybeScheme.startsWith("http://") || maybeScheme.startsWith("https://")
65        || maybeScheme.startsWith("ftp://") || maybeScheme.startsWith("mailto:")) {
66      return true;
67    }
68
69    // If it's an absolute URI with a different scheme, it's invalid.
70    // ClearSilver defines an absolute URI as one that contains a colon prior
71    // to any slash.
72    int slashPos = in.indexOf('/');
73    if (slashPos != -1) {
74      // only colons before this point are bad.
75      return in.lastIndexOf(':', slashPos - 1) == -1;
76    } else {
77      // then any colon is bad.
78      return in.indexOf(':') == -1;
79    }
80  }
81
82  /**
83   * Converts an ASCII string to lowercase. Non-ASCII characters are replaced with '?'.
84   */
85  private String toLowerCaseAsciiOnly(String string) {
86    char[] ca = string.toCharArray();
87    for (int i = 0; i < ca.length; i++) {
88      char ch = ca[i];
89      ca[i] =
90          (Character.UnicodeBlock.of(ch) == UnicodeBlock.BASIC_LATIN)
91              ? Character.toLowerCase(ch)
92              : '?';
93    }
94    return new String(ca);
95  }
96}
97