Thu, 22 Jan 2015 13:21:57 +0100
Incorporate requested changes from Mozilla in review:
https://bugzilla.mozilla.org/show_bug.cgi?id=1123480#c6
michael@0 | 1 | /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */ |
michael@0 | 2 | /* This Source Code Form is subject to the terms of the Mozilla Public |
michael@0 | 3 | * License, v. 2.0. If a copy of the MPL was not distributed with this |
michael@0 | 4 | * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ |
michael@0 | 5 | |
michael@0 | 6 | #include "nsString.h" |
michael@0 | 7 | #include "nsICharsetConverterManager.h" |
michael@0 | 8 | #include "nsIScriptableUConv.h" |
michael@0 | 9 | #include "nsScriptableUConv.h" |
michael@0 | 10 | #include "nsIStringStream.h" |
michael@0 | 11 | #include "nsComponentManagerUtils.h" |
michael@0 | 12 | #include "nsCharsetAlias.h" |
michael@0 | 13 | #include "nsServiceManagerUtils.h" |
michael@0 | 14 | |
michael@0 | 15 | /* Implementation file */ |
michael@0 | 16 | NS_IMPL_ISUPPORTS(nsScriptableUnicodeConverter, nsIScriptableUnicodeConverter) |
michael@0 | 17 | |
michael@0 | 18 | nsScriptableUnicodeConverter::nsScriptableUnicodeConverter() |
michael@0 | 19 | : mIsInternal(false) |
michael@0 | 20 | { |
michael@0 | 21 | } |
michael@0 | 22 | |
michael@0 | 23 | nsScriptableUnicodeConverter::~nsScriptableUnicodeConverter() |
michael@0 | 24 | { |
michael@0 | 25 | } |
michael@0 | 26 | |
michael@0 | 27 | nsresult |
michael@0 | 28 | nsScriptableUnicodeConverter::ConvertFromUnicodeWithLength(const nsAString& aSrc, |
michael@0 | 29 | int32_t* aOutLen, |
michael@0 | 30 | char **_retval) |
michael@0 | 31 | { |
michael@0 | 32 | if (!mEncoder) |
michael@0 | 33 | return NS_ERROR_FAILURE; |
michael@0 | 34 | |
michael@0 | 35 | nsresult rv = NS_OK; |
michael@0 | 36 | int32_t inLength = aSrc.Length(); |
michael@0 | 37 | const nsAFlatString& flatSrc = PromiseFlatString(aSrc); |
michael@0 | 38 | rv = mEncoder->GetMaxLength(flatSrc.get(), inLength, aOutLen); |
michael@0 | 39 | if (NS_SUCCEEDED(rv)) { |
michael@0 | 40 | *_retval = (char*)moz_malloc(*aOutLen+1); |
michael@0 | 41 | if (!*_retval) |
michael@0 | 42 | return NS_ERROR_OUT_OF_MEMORY; |
michael@0 | 43 | |
michael@0 | 44 | rv = mEncoder->Convert(flatSrc.get(), &inLength, *_retval, aOutLen); |
michael@0 | 45 | if (NS_SUCCEEDED(rv)) |
michael@0 | 46 | { |
michael@0 | 47 | (*_retval)[*aOutLen] = '\0'; |
michael@0 | 48 | return NS_OK; |
michael@0 | 49 | } |
michael@0 | 50 | moz_free(*_retval); |
michael@0 | 51 | } |
michael@0 | 52 | *_retval = nullptr; |
michael@0 | 53 | return NS_ERROR_FAILURE; |
michael@0 | 54 | } |
michael@0 | 55 | |
michael@0 | 56 | /* ACString ConvertFromUnicode (in AString src); */ |
michael@0 | 57 | NS_IMETHODIMP |
michael@0 | 58 | nsScriptableUnicodeConverter::ConvertFromUnicode(const nsAString& aSrc, |
michael@0 | 59 | nsACString& _retval) |
michael@0 | 60 | { |
michael@0 | 61 | int32_t len; |
michael@0 | 62 | char* str; |
michael@0 | 63 | nsresult rv = ConvertFromUnicodeWithLength(aSrc, &len, &str); |
michael@0 | 64 | if (NS_SUCCEEDED(rv)) { |
michael@0 | 65 | // No Adopt on nsACString :( |
michael@0 | 66 | if (!_retval.Assign(str, len, mozilla::fallible_t())) { |
michael@0 | 67 | rv = NS_ERROR_OUT_OF_MEMORY; |
michael@0 | 68 | } |
michael@0 | 69 | moz_free(str); |
michael@0 | 70 | } |
michael@0 | 71 | return rv; |
michael@0 | 72 | } |
michael@0 | 73 | |
michael@0 | 74 | nsresult |
michael@0 | 75 | nsScriptableUnicodeConverter::FinishWithLength(char **_retval, int32_t* aLength) |
michael@0 | 76 | { |
michael@0 | 77 | if (!mEncoder) |
michael@0 | 78 | return NS_ERROR_FAILURE; |
michael@0 | 79 | |
michael@0 | 80 | int32_t finLength = 32; |
michael@0 | 81 | |
michael@0 | 82 | *_retval = (char *)moz_malloc(finLength); |
michael@0 | 83 | if (!*_retval) |
michael@0 | 84 | return NS_ERROR_OUT_OF_MEMORY; |
michael@0 | 85 | |
michael@0 | 86 | nsresult rv = mEncoder->Finish(*_retval, &finLength); |
michael@0 | 87 | if (NS_SUCCEEDED(rv)) |
michael@0 | 88 | *aLength = finLength; |
michael@0 | 89 | else |
michael@0 | 90 | moz_free(*_retval); |
michael@0 | 91 | |
michael@0 | 92 | return rv; |
michael@0 | 93 | |
michael@0 | 94 | } |
michael@0 | 95 | |
michael@0 | 96 | /* ACString Finish(); */ |
michael@0 | 97 | NS_IMETHODIMP |
michael@0 | 98 | nsScriptableUnicodeConverter::Finish(nsACString& _retval) |
michael@0 | 99 | { |
michael@0 | 100 | int32_t len; |
michael@0 | 101 | char* str; |
michael@0 | 102 | nsresult rv = FinishWithLength(&str, &len); |
michael@0 | 103 | if (NS_SUCCEEDED(rv)) { |
michael@0 | 104 | // No Adopt on nsACString :( |
michael@0 | 105 | if (!_retval.Assign(str, len, mozilla::fallible_t())) { |
michael@0 | 106 | rv = NS_ERROR_OUT_OF_MEMORY; |
michael@0 | 107 | } |
michael@0 | 108 | moz_free(str); |
michael@0 | 109 | } |
michael@0 | 110 | return rv; |
michael@0 | 111 | } |
michael@0 | 112 | |
michael@0 | 113 | /* AString ConvertToUnicode (in ACString src); */ |
michael@0 | 114 | NS_IMETHODIMP |
michael@0 | 115 | nsScriptableUnicodeConverter::ConvertToUnicode(const nsACString& aSrc, nsAString& _retval) |
michael@0 | 116 | { |
michael@0 | 117 | nsACString::const_iterator i; |
michael@0 | 118 | aSrc.BeginReading(i); |
michael@0 | 119 | return ConvertFromByteArray(reinterpret_cast<const uint8_t*>(i.get()), |
michael@0 | 120 | aSrc.Length(), |
michael@0 | 121 | _retval); |
michael@0 | 122 | } |
michael@0 | 123 | |
michael@0 | 124 | /* AString convertFromByteArray([const,array,size_is(aCount)] in octet aData, |
michael@0 | 125 | in unsigned long aCount); |
michael@0 | 126 | */ |
michael@0 | 127 | NS_IMETHODIMP |
michael@0 | 128 | nsScriptableUnicodeConverter::ConvertFromByteArray(const uint8_t* aData, |
michael@0 | 129 | uint32_t aCount, |
michael@0 | 130 | nsAString& _retval) |
michael@0 | 131 | { |
michael@0 | 132 | if (!mDecoder) |
michael@0 | 133 | return NS_ERROR_FAILURE; |
michael@0 | 134 | |
michael@0 | 135 | nsresult rv = NS_OK; |
michael@0 | 136 | int32_t inLength = aCount; |
michael@0 | 137 | int32_t outLength; |
michael@0 | 138 | rv = mDecoder->GetMaxLength(reinterpret_cast<const char*>(aData), |
michael@0 | 139 | inLength, &outLength); |
michael@0 | 140 | if (NS_SUCCEEDED(rv)) |
michael@0 | 141 | { |
michael@0 | 142 | char16_t* buf = (char16_t*)moz_malloc((outLength+1)*sizeof(char16_t)); |
michael@0 | 143 | if (!buf) |
michael@0 | 144 | return NS_ERROR_OUT_OF_MEMORY; |
michael@0 | 145 | |
michael@0 | 146 | rv = mDecoder->Convert(reinterpret_cast<const char*>(aData), |
michael@0 | 147 | &inLength, buf, &outLength); |
michael@0 | 148 | if (NS_SUCCEEDED(rv)) |
michael@0 | 149 | { |
michael@0 | 150 | buf[outLength] = 0; |
michael@0 | 151 | if (!_retval.Assign(buf, outLength, mozilla::fallible_t())) { |
michael@0 | 152 | rv = NS_ERROR_OUT_OF_MEMORY; |
michael@0 | 153 | } |
michael@0 | 154 | } |
michael@0 | 155 | moz_free(buf); |
michael@0 | 156 | return rv; |
michael@0 | 157 | } |
michael@0 | 158 | return NS_ERROR_FAILURE; |
michael@0 | 159 | |
michael@0 | 160 | } |
michael@0 | 161 | |
michael@0 | 162 | /* void convertToByteArray(in AString aString, |
michael@0 | 163 | [optional] out unsigned long aLen, |
michael@0 | 164 | [array, size_is(aLen),retval] out octet aData); |
michael@0 | 165 | */ |
michael@0 | 166 | NS_IMETHODIMP |
michael@0 | 167 | nsScriptableUnicodeConverter::ConvertToByteArray(const nsAString& aString, |
michael@0 | 168 | uint32_t* aLen, |
michael@0 | 169 | uint8_t** _aData) |
michael@0 | 170 | { |
michael@0 | 171 | char* data; |
michael@0 | 172 | int32_t len; |
michael@0 | 173 | nsresult rv = ConvertFromUnicodeWithLength(aString, &len, &data); |
michael@0 | 174 | if (NS_FAILED(rv)) |
michael@0 | 175 | return rv; |
michael@0 | 176 | nsXPIDLCString str; |
michael@0 | 177 | str.Adopt(data, len); // NOTE: This uses the XPIDLCString as a byte array |
michael@0 | 178 | |
michael@0 | 179 | rv = FinishWithLength(&data, &len); |
michael@0 | 180 | if (NS_FAILED(rv)) |
michael@0 | 181 | return rv; |
michael@0 | 182 | |
michael@0 | 183 | str.Append(data, len); |
michael@0 | 184 | moz_free(data); |
michael@0 | 185 | // NOTE: this being a byte array, it needs no null termination |
michael@0 | 186 | *_aData = reinterpret_cast<uint8_t*>(moz_malloc(str.Length())); |
michael@0 | 187 | if (!*_aData) |
michael@0 | 188 | return NS_ERROR_OUT_OF_MEMORY; |
michael@0 | 189 | memcpy(*_aData, str.get(), str.Length()); |
michael@0 | 190 | *aLen = str.Length(); |
michael@0 | 191 | return NS_OK; |
michael@0 | 192 | } |
michael@0 | 193 | |
michael@0 | 194 | /* nsIInputStream convertToInputStream(in AString aString); */ |
michael@0 | 195 | NS_IMETHODIMP |
michael@0 | 196 | nsScriptableUnicodeConverter::ConvertToInputStream(const nsAString& aString, |
michael@0 | 197 | nsIInputStream** _retval) |
michael@0 | 198 | { |
michael@0 | 199 | nsresult rv; |
michael@0 | 200 | nsCOMPtr<nsIStringInputStream> inputStream = |
michael@0 | 201 | do_CreateInstance("@mozilla.org/io/string-input-stream;1", &rv); |
michael@0 | 202 | if (NS_FAILED(rv)) |
michael@0 | 203 | return rv; |
michael@0 | 204 | |
michael@0 | 205 | uint8_t* data; |
michael@0 | 206 | uint32_t dataLen; |
michael@0 | 207 | rv = ConvertToByteArray(aString, &dataLen, &data); |
michael@0 | 208 | if (NS_FAILED(rv)) |
michael@0 | 209 | return rv; |
michael@0 | 210 | |
michael@0 | 211 | rv = inputStream->AdoptData(reinterpret_cast<char*>(data), dataLen); |
michael@0 | 212 | if (NS_FAILED(rv)) { |
michael@0 | 213 | moz_free(data); |
michael@0 | 214 | return rv; |
michael@0 | 215 | } |
michael@0 | 216 | |
michael@0 | 217 | NS_ADDREF(*_retval = inputStream); |
michael@0 | 218 | return rv; |
michael@0 | 219 | } |
michael@0 | 220 | |
michael@0 | 221 | /* attribute string charset; */ |
michael@0 | 222 | NS_IMETHODIMP |
michael@0 | 223 | nsScriptableUnicodeConverter::GetCharset(char * *aCharset) |
michael@0 | 224 | { |
michael@0 | 225 | *aCharset = ToNewCString(mCharset); |
michael@0 | 226 | if (!*aCharset) |
michael@0 | 227 | return NS_ERROR_OUT_OF_MEMORY; |
michael@0 | 228 | |
michael@0 | 229 | return NS_OK; |
michael@0 | 230 | } |
michael@0 | 231 | |
michael@0 | 232 | NS_IMETHODIMP |
michael@0 | 233 | nsScriptableUnicodeConverter::SetCharset(const char * aCharset) |
michael@0 | 234 | { |
michael@0 | 235 | mCharset.Assign(aCharset); |
michael@0 | 236 | return InitConverter(); |
michael@0 | 237 | } |
michael@0 | 238 | |
michael@0 | 239 | NS_IMETHODIMP |
michael@0 | 240 | nsScriptableUnicodeConverter::GetIsInternal(bool *aIsInternal) |
michael@0 | 241 | { |
michael@0 | 242 | *aIsInternal = mIsInternal; |
michael@0 | 243 | return NS_OK; |
michael@0 | 244 | } |
michael@0 | 245 | |
michael@0 | 246 | NS_IMETHODIMP |
michael@0 | 247 | nsScriptableUnicodeConverter::SetIsInternal(const bool aIsInternal) |
michael@0 | 248 | { |
michael@0 | 249 | mIsInternal = aIsInternal; |
michael@0 | 250 | return NS_OK; |
michael@0 | 251 | } |
michael@0 | 252 | |
michael@0 | 253 | nsresult |
michael@0 | 254 | nsScriptableUnicodeConverter::InitConverter() |
michael@0 | 255 | { |
michael@0 | 256 | nsresult rv = NS_OK; |
michael@0 | 257 | mEncoder = nullptr; |
michael@0 | 258 | |
michael@0 | 259 | nsCOMPtr<nsICharsetConverterManager> ccm = do_GetService(NS_CHARSETCONVERTERMANAGER_CONTRACTID, &rv); |
michael@0 | 260 | if (NS_FAILED(rv) || !ccm) { |
michael@0 | 261 | return rv; |
michael@0 | 262 | } |
michael@0 | 263 | |
michael@0 | 264 | // get an unicode converter |
michael@0 | 265 | rv = ccm->GetUnicodeEncoder(mCharset.get(), getter_AddRefs(mEncoder)); |
michael@0 | 266 | if (NS_FAILED(rv)) { |
michael@0 | 267 | return rv; |
michael@0 | 268 | } |
michael@0 | 269 | |
michael@0 | 270 | rv = mEncoder->SetOutputErrorBehavior(nsIUnicodeEncoder::kOnError_Replace, nullptr, (char16_t)'?'); |
michael@0 | 271 | if (NS_FAILED(rv)) { |
michael@0 | 272 | return rv; |
michael@0 | 273 | } |
michael@0 | 274 | |
michael@0 | 275 | nsAutoCString charset; |
michael@0 | 276 | rv = mIsInternal ? nsCharsetAlias::GetPreferredInternal(mCharset, charset) |
michael@0 | 277 | : nsCharsetAlias::GetPreferred(mCharset, charset); |
michael@0 | 278 | if (NS_FAILED(rv)) { |
michael@0 | 279 | return rv; |
michael@0 | 280 | } |
michael@0 | 281 | |
michael@0 | 282 | rv = ccm->GetUnicodeDecoderRaw(charset.get(), getter_AddRefs(mDecoder)); |
michael@0 | 283 | if (NS_FAILED(rv)) { |
michael@0 | 284 | return rv; |
michael@0 | 285 | } |
michael@0 | 286 | |
michael@0 | 287 | // The UTF-8 decoder used to throw regardless of the error behavior. |
michael@0 | 288 | // Simulating the old behavior for compatibility with legacy callers |
michael@0 | 289 | // (including addons). If callers want a control over the behavior, |
michael@0 | 290 | // they should switch to TextDecoder. |
michael@0 | 291 | if (charset.EqualsLiteral("UTF-8")) { |
michael@0 | 292 | mDecoder->SetInputErrorBehavior(nsIUnicodeDecoder::kOnError_Signal); |
michael@0 | 293 | } |
michael@0 | 294 | |
michael@0 | 295 | return rv ; |
michael@0 | 296 | } |