1 // Copyright 2020 The Chromium Authors 2 // Use of this source code is governed by a BSD-style license that can be 3 // found in the LICENSE file. 4 5 #ifndef BASE_STRINGS_ESCAPE_H_ 6 #define BASE_STRINGS_ESCAPE_H_ 7 8 #include <stdint.h> 9 10 #include <set> 11 #include <string> 12 13 #include "base/base_export.h" 14 #include "base/strings/string_piece.h" 15 #include "base/strings/utf_offset_string_conversions.h" 16 17 namespace base { 18 19 // Escaping -------------------------------------------------------------------- 20 21 // Escapes all characters except unreserved characters. Unreserved characters, 22 // as defined in RFC 3986, include alphanumerics and -._~ 23 BASE_EXPORT std::string EscapeAllExceptUnreserved(StringPiece text); 24 25 // Escapes characters in text suitable for use as a query parameter value. 26 // We %XX everything except alphanumerics and -_.!~*'() 27 // Spaces change to "+" unless you pass usePlus=false. 28 // This is basically the same as encodeURIComponent in javascript. 29 BASE_EXPORT std::string EscapeQueryParamValue(StringPiece text, bool use_plus); 30 31 // Escapes a partial or complete file/pathname. This includes: 32 // non-printable, non-7bit, and (including space) "#%:<>?[\]^`{|} 33 BASE_EXPORT std::string EscapePath(StringPiece path); 34 35 #if BUILDFLAG(IS_APPLE) 36 // Escapes characters as per expectations of NSURL. This includes: 37 // non-printable, non-7bit, and (including space) "#%<>[\]^`{|} 38 BASE_EXPORT std::string EscapeNSURLPrecursor(StringPiece precursor); 39 #endif // BUILDFLAG(IS_APPLE) 40 41 // Escapes application/x-www-form-urlencoded content. This includes: 42 // non-printable, non-7bit, and (including space) ?>=<;+'&%$#"![\]^`{|} 43 // Space is escaped as + (if use_plus is true) and other special characters 44 // as %XX (hex). 45 BASE_EXPORT std::string EscapeUrlEncodedData(StringPiece path, bool use_plus); 46 47 // Escapes all non-ASCII input, as well as escaping % to %25. 48 BASE_EXPORT std::string EscapeNonASCIIAndPercent(StringPiece input); 49 50 // Escapes all non-ASCII input. Note this function leaves % unescaped, which 51 // means the unescaping the resulting string will not give back the original 52 // input. 53 BASE_EXPORT std::string EscapeNonASCII(StringPiece input); 54 55 // Escapes characters in text suitable for use as an external protocol handler 56 // command. 57 // We %XX everything except alphanumerics and -_.!~*'() and the restricted 58 // characters (;/?:@&=+$,#[]) and a valid percent escape sequence (%XX). 59 BASE_EXPORT std::string EscapeExternalHandlerValue(StringPiece text); 60 61 // Appends the given character to the output string, escaping the character if 62 // the character would be interpreted as an HTML delimiter. 63 BASE_EXPORT void AppendEscapedCharForHTML(char c, std::string* output); 64 65 // Escapes chars that might cause this text to be interpreted as HTML tags. 66 BASE_EXPORT std::string EscapeForHTML(StringPiece text); 67 BASE_EXPORT std::u16string EscapeForHTML(StringPiece16 text); 68 69 // Unescaping ------------------------------------------------------------------ 70 71 class UnescapeRule { 72 public: 73 // A combination of the following flags that is passed to the unescaping 74 // functions. 75 typedef uint32_t Type; 76 77 // Don't unescape anything at all. 78 static constexpr Type NONE = 0; 79 80 // Don't unescape anything special, but all normal unescaping will happen. 81 // This is a placeholder and can't be combined with other flags (since it's 82 // just the absence of them). All other unescape rules imply "normal" in 83 // addition to their special meaning. Things like escaped letters, digits, 84 // and most symbols will get unescaped with this mode. 85 static constexpr Type NORMAL = 1 << 0; 86 87 // Convert %20 to spaces. In some places where we're showing URLs, we may 88 // want this. In places where the URL may be copied and pasted out, then 89 // you wouldn't want this since it might not be interpreted in one piece 90 // by other applications. Other UTF-8 spaces will not be unescaped. 91 static constexpr Type SPACES = 1 << 1; 92 93 // Unescapes '/' and '\\'. If these characters were unescaped, the resulting 94 // URL won't be the same as the source one. Moreover, they are dangerous to 95 // unescape in strings that will be used as file paths or names. This value 96 // should only be used when slashes don't have special meaning, like data 97 // URLs. 98 static constexpr Type PATH_SEPARATORS = 1 << 2; 99 100 // Unescapes various characters that will change the meaning of URLs, 101 // including '%', '+', '&', '#'. Does not unescape path separators. 102 // If these characters were unescaped, the resulting URL won't be the same 103 // as the source one. This flag is used when generating final output like 104 // filenames for URLs where we won't be interpreting as a URL and want to do 105 // as much unescaping as possible. 106 static constexpr Type URL_SPECIAL_CHARS_EXCEPT_PATH_SEPARATORS = 1 << 3; 107 108 // URL queries use "+" for space. This flag controls that replacement. 109 static constexpr Type REPLACE_PLUS_WITH_SPACE = 1 << 4; 110 }; 111 112 // Unescapes |escaped_text| and returns the result. 113 // Unescaping consists of looking for the exact pattern "%XX", where each X is 114 // a hex digit, and converting to the character with the numerical value of 115 // those digits. Thus "i%20=%203%3b" unescapes to "i = 3;", if the 116 // "UnescapeRule::SPACES" used. 117 // 118 // This method does not ensure that the output is a valid string using any 119 // character encoding. However, it does leave escaped certain byte sequences 120 // that would be dangerous to display to the user, because if interpreted as 121 // UTF-8, they could be used to mislead the user. Callers that want to 122 // unconditionally unescape everything for uses other than displaying data to 123 // the user should use UnescapeBinaryURLComponent(). 124 BASE_EXPORT std::string UnescapeURLComponent(StringPiece escaped_text, 125 UnescapeRule::Type rules); 126 127 // Unescapes the given substring as a URL, and then tries to interpret the 128 // result as being encoded as UTF-8. If the result is convertible into UTF-8, it 129 // will be returned as converted. If it is not, the original escaped string will 130 // be converted into a std::u16string and returned. |adjustments| provides 131 // information on how the original string was adjusted to get the string 132 // returned. 133 BASE_EXPORT std::u16string UnescapeAndDecodeUTF8URLComponentWithAdjustments( 134 StringPiece text, 135 UnescapeRule::Type rules, 136 OffsetAdjuster::Adjustments* adjustments); 137 138 // Unescapes a component of a URL for use as binary data. Unlike 139 // UnescapeURLComponent, leaves nothing unescaped, including nulls, invalid 140 // characters, characters that are unsafe to display, etc. This should *not* 141 // be used when displaying the decoded data to the user. 142 // 143 // Only the NORMAL and REPLACE_PLUS_WITH_SPACE rules are allowed. 144 BASE_EXPORT std::string UnescapeBinaryURLComponent( 145 StringPiece escaped_text, 146 UnescapeRule::Type rules = UnescapeRule::NORMAL); 147 148 // Variant of UnescapeBinaryURLComponent(). Writes output to |unescaped_text|. 149 // Returns true on success, returns false and clears |unescaped_text| on 150 // failure. Fails on characters escaped that are unsafe to unescape in some 151 // contexts, which are defined as characters "\0" through "\x1F" (Which includes 152 // CRLF but not space), and optionally path separators. Path separators include 153 // both forward and backward slashes on all platforms. Does not fail if any of 154 // those characters appear unescaped in the input string. 155 BASE_EXPORT bool UnescapeBinaryURLComponentSafe(StringPiece escaped_text, 156 bool fail_on_path_separators, 157 std::string* unescaped_text); 158 159 // Returns true if |escaped_text| contains any element of |bytes| in 160 // percent-encoded form. 161 // 162 // For example, if |bytes| is {'%', '/'}, returns true if |escaped_text| 163 // contains "%25" or "%2F", but not if it just contains bare '%' or '/' 164 // characters. 165 BASE_EXPORT bool ContainsEncodedBytes(StringPiece escaped_text, 166 const std::set<unsigned char>& bytes); 167 168 // Unescapes the following ampersand character codes from |text|: 169 // < > & " ' 170 BASE_EXPORT std::u16string UnescapeForHTML(StringPiece16 text); 171 172 } // namespace base 173 174 #endif // BASE_STRINGS_ESCAPE_H_ 175