1 // Copyright (c) 2011 The Chromium Authors. All rights reserved. 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 NET_BASE_DATA_URL_H_ 6 #define NET_BASE_DATA_URL_H_ 7 8 #include <string> 9 10 #include "base/compiler_specific.h" 11 #include "base/memory/scoped_refptr.h" 12 #include "base/strings/string_piece.h" 13 #include "net/base/net_errors.h" 14 #include "net/base/net_export.h" 15 16 class GURL; 17 18 namespace net { 19 20 class HttpResponseHeaders; 21 22 // See RFC 2397 for a complete description of the 'data' URL scheme. 23 // 24 // Briefly, a 'data' URL has the form: 25 // 26 // data:[<mediatype>][;base64],<data> 27 // 28 // The <mediatype> is an Internet media type specification (with optional 29 // parameters.) The appearance of ";base64" means that the data is encoded as 30 // base64. Without ";base64", the data (as a sequence of octets) is represented 31 // using ASCII encoding for octets inside the range of safe URL characters and 32 // using the standard %xx hex encoding of URLs for octets outside that range. 33 // If <mediatype> is omitted, it defaults to text/plain;charset=US-ASCII. As a 34 // shorthand, "text/plain" can be omitted but the charset parameter supplied. 35 // 36 class NET_EXPORT DataURL { 37 public: 38 // This method can be used to parse a 'data' URL into its component pieces. 39 // 40 // |mime_type| and |charset| must be non-null and point to empty strings. 41 // 42 // If |data| is null, then the <data> section will not be parsed or validated. 43 // If non-null, it must point to an empty string. 44 // 45 // The resulting mime_type is normalized to lowercase. The data is the 46 // decoded data (e.g.., if the data URL specifies base64 encoding, then the 47 // returned data is base64 decoded, and any %-escaped bytes are unescaped). 48 // 49 // If the media type value doesn't match the media-type production defined in 50 // RFC 7231, mime_type will be set to the default value "text/plain". We 51 // don't simply fail for this grammar violation since Chromium had been 52 // accepting such invalid values. For example, <img> element with the src 53 // attribute set to a data URL with an invalid media type "image" (without a 54 // slash and subtype) had been displayed. However, the value this method will 55 // store in mime_type argument can be used for generating other headers, etc. 56 // This could lead to security vulnerability. We don't want to accept 57 // arbitrary value and ask each caller to validate the return value. 58 // 59 // If the charset parameter is specified but its value doesn't match the 60 // token production defined in RFC 7230, this method simply fails and returns 61 // false. 62 // 63 // If there's any other grammar violation in the URL, then this method will 64 // return false, and all passed in pointers will be unmodified. On success, 65 // true is returned. 66 static bool Parse(const GURL& url, 67 std::string* mime_type, 68 std::string* charset, 69 std::string* data) WARN_UNUSED_RESULT; 70 71 // Similar to parse, except that it also generates a bogus set of response 72 // headers, with Content-Type populated, and takes a method. Only the "HEAD" 73 // method modifies the response, resulting in a 0-length body. All arguments 74 // except must be non-null. All std::string pointers must point to empty 75 // strings, and |*headers| must be nullptr. Returns net::OK on success. 76 static Error BuildResponse(const GURL& url, 77 base::StringPiece method, 78 std::string* mime_type, 79 std::string* charset, 80 std::string* data, 81 scoped_refptr<HttpResponseHeaders>* headers) 82 WARN_UNUSED_RESULT; 83 }; 84 85 } // namespace net 86 87 #endif // NET_BASE_DATA_URL_H_ 88