The Tor Browser / file comparison
comparison: intl/uconv/src/nsUTF8ConverterService.cpp
intl/uconv/src/nsUTF8ConverterService.cpp
- branch
- TOR_BUG_3246
- changeset 4
- fc2d59ddac77
equal
deleted
inserted
replaced
| -1:000000000000 | 0:1c64aebd1465 |
|---|---|
| 1 /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 2 -*- */ | |
| 2 /* vim:expandtab:shiftwidth=2:tabstop=4: | |
| 3 */ | |
| 4 /* This Source Code Form is subject to the terms of the Mozilla Public | |
| 5 * License, v. 2.0. If a copy of the MPL was not distributed with this | |
| 6 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ | |
| 7 #include "nsString.h" | |
| 8 #include "nsICharsetConverterManager.h" | |
| 9 #include "nsUTF8ConverterService.h" | |
| 10 #include "nsEscape.h" | |
| 11 #include "nsAutoPtr.h" | |
| 12 #include "nsServiceManagerUtils.h" | |
| 13 | |
| 14 NS_IMPL_ISUPPORTS(nsUTF8ConverterService, nsIUTF8ConverterService) | |
| 15 | |
| 16 static nsresult | |
| 17 ToUTF8(const nsACString &aString, const char *aCharset, | |
| 18 bool aAllowSubstitution, nsACString &aResult) | |
| 19 { | |
| 20 nsresult rv; | |
| 21 if (!aCharset || !*aCharset) | |
| 22 return NS_ERROR_INVALID_ARG; | |
| 23 | |
| 24 nsCOMPtr<nsICharsetConverterManager> ccm; | |
| 25 | |
| 26 ccm = do_GetService(NS_CHARSETCONVERTERMANAGER_CONTRACTID, &rv); | |
| 27 NS_ENSURE_SUCCESS(rv, rv); | |
| 28 | |
| 29 nsCOMPtr<nsIUnicodeDecoder> unicodeDecoder; | |
| 30 rv = ccm->GetUnicodeDecoder(aCharset, | |
| 31 getter_AddRefs(unicodeDecoder)); | |
| 32 NS_ENSURE_SUCCESS(rv, rv); | |
| 33 | |
| 34 if (!aAllowSubstitution) | |
| 35 unicodeDecoder->SetInputErrorBehavior(nsIUnicodeDecoder::kOnError_Signal); | |
| 36 | |
| 37 int32_t srcLen = aString.Length(); | |
| 38 int32_t dstLen; | |
| 39 const nsAFlatCString& inStr = PromiseFlatCString(aString); | |
| 40 rv = unicodeDecoder->GetMaxLength(inStr.get(), srcLen, &dstLen); | |
| 41 NS_ENSURE_SUCCESS(rv, rv); | |
| 42 | |
| 43 nsAutoArrayPtr<char16_t> ustr(new char16_t[dstLen]); | |
| 44 NS_ENSURE_TRUE(ustr, NS_ERROR_OUT_OF_MEMORY); | |
| 45 | |
| 46 rv = unicodeDecoder->Convert(inStr.get(), &srcLen, ustr, &dstLen); | |
| 47 if (NS_SUCCEEDED(rv)){ | |
| 48 // Tru64 Cxx needs an explicit get() | |
| 49 CopyUTF16toUTF8(Substring(ustr.get(), ustr + dstLen), aResult); | |
| 50 } | |
| 51 return rv; | |
| 52 } | |
| 53 | |
| 54 NS_IMETHODIMP | |
| 55 nsUTF8ConverterService::ConvertStringToUTF8(const nsACString &aString, | |
| 56 const char *aCharset, | |
| 57 bool aSkipCheck, | |
| 58 bool aAllowSubstitution, | |
| 59 uint8_t aOptionalArgc, | |
| 60 nsACString &aUTF8String) | |
| 61 { | |
| 62 bool allowSubstitution = (aOptionalArgc == 1) ? aAllowSubstitution : true; | |
| 63 | |
| 64 // return if ASCII only or valid UTF-8 providing that the ASCII/UTF-8 | |
| 65 // check is requested. It may not be asked for if a caller suspects | |
| 66 // that the input is in non-ASCII 7bit charset (ISO-2022-xx, HZ) or | |
| 67 // it's in a charset other than UTF-8 that can be mistaken for UTF-8. | |
| 68 if (!aSkipCheck && (IsASCII(aString) || IsUTF8(aString))) { | |
| 69 aUTF8String = aString; | |
| 70 return NS_OK; | |
| 71 } | |
| 72 | |
| 73 aUTF8String.Truncate(); | |
| 74 | |
| 75 nsresult rv = ToUTF8(aString, aCharset, allowSubstitution, aUTF8String); | |
| 76 | |
| 77 // additional protection for cases where check is skipped and the input | |
| 78 // is actually in UTF-8 as opposed to aCharset. (i.e. caller's hunch | |
| 79 // was wrong.) We don't check ASCIIness assuming there's no charset | |
| 80 // incompatible with ASCII (we don't support EBCDIC). | |
| 81 if (aSkipCheck && NS_FAILED(rv) && IsUTF8(aString)) { | |
| 82 aUTF8String = aString; | |
| 83 return NS_OK; | |
| 84 } | |
| 85 | |
| 86 return rv; | |
| 87 } | |
| 88 | |
| 89 NS_IMETHODIMP | |
| 90 nsUTF8ConverterService::ConvertURISpecToUTF8(const nsACString &aSpec, | |
| 91 const char *aCharset, | |
| 92 nsACString &aUTF8Spec) | |
| 93 { | |
| 94 // assume UTF-8 if the spec contains unescaped non-ASCII characters. | |
| 95 // No valid spec in Mozilla would break this assumption. | |
| 96 if (!IsASCII(aSpec)) { | |
| 97 aUTF8Spec = aSpec; | |
| 98 return NS_OK; | |
| 99 } | |
| 100 | |
| 101 aUTF8Spec.Truncate(); | |
| 102 | |
| 103 nsAutoCString unescapedSpec; | |
| 104 // NS_UnescapeURL does not fill up unescapedSpec unless there's at least | |
| 105 // one character to unescape. | |
| 106 bool written = NS_UnescapeURL(PromiseFlatCString(aSpec).get(), aSpec.Length(), | |
| 107 esc_OnlyNonASCII, unescapedSpec); | |
| 108 | |
| 109 if (!written) { | |
| 110 aUTF8Spec = aSpec; | |
| 111 return NS_OK; | |
| 112 } | |
| 113 // return if ASCII only or escaped UTF-8 | |
| 114 if (IsASCII(unescapedSpec) || IsUTF8(unescapedSpec)) { | |
| 115 aUTF8Spec = unescapedSpec; | |
| 116 return NS_OK; | |
| 117 } | |
| 118 | |
| 119 return ToUTF8(unescapedSpec, aCharset, true, aUTF8Spec); | |
| 120 } | |
| 121 |
