1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
|
/* -*- Mode: C; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
#include "nsConverterInputStream.h"
#include "nsIInputStream.h"
#include "nsReadLine.h"
#include "nsStreamUtils.h"
#include <algorithm>
#include "mozilla/dom/EncodingUtils.h"
using mozilla::dom::EncodingUtils;
#define CONVERTER_BUFFER_SIZE 8192
NS_IMPL_ISUPPORTS(nsConverterInputStream, nsIConverterInputStream,
nsIUnicharInputStream, nsIUnicharLineInputStream)
NS_IMETHODIMP
nsConverterInputStream::Init(nsIInputStream* aStream,
const char *aCharset,
int32_t aBufferSize,
char16_t aReplacementChar)
{
nsAutoCString label;
if (!aCharset) {
label.AssignLiteral("UTF-8");
} else {
label = aCharset;
}
if (aBufferSize <=0) aBufferSize=CONVERTER_BUFFER_SIZE;
// get the decoder
nsAutoCString encoding;
if (label.EqualsLiteral("UTF-16")) {
// Compat with old test cases. Unclear if any extensions really care.
encoding.Assign(label);
} else if (!EncodingUtils::FindEncodingForLabelNoReplacement(label,
encoding)) {
return NS_ERROR_UCONV_NOCONV;
}
mConverter = EncodingUtils::DecoderForEncoding(encoding);
// set up our buffers
if (!mByteData.SetCapacity(aBufferSize, mozilla::fallible) ||
!mUnicharData.SetCapacity(aBufferSize, mozilla::fallible)) {
return NS_ERROR_OUT_OF_MEMORY;
}
mInput = aStream;
mReplacementChar = aReplacementChar;
if (!aReplacementChar ||
aReplacementChar != mConverter->GetCharacterForUnMapped()) {
mConverter->SetInputErrorBehavior(nsIUnicodeDecoder::kOnError_Signal);
}
return NS_OK;
}
NS_IMETHODIMP
nsConverterInputStream::Close()
{
nsresult rv = mInput ? mInput->Close() : NS_OK;
mLineBuffer = nullptr;
mInput = nullptr;
mConverter = nullptr;
mByteData.Clear();
mUnicharData.Clear();
return rv;
}
NS_IMETHODIMP
nsConverterInputStream::Read(char16_t* aBuf,
uint32_t aCount,
uint32_t *aReadCount)
{
NS_ASSERTION(mUnicharDataLength >= mUnicharDataOffset, "unsigned madness");
uint32_t readCount = mUnicharDataLength - mUnicharDataOffset;
if (0 == readCount) {
// Fill the unichar buffer
readCount = Fill(&mLastErrorCode);
if (readCount == 0) {
*aReadCount = 0;
return mLastErrorCode;
}
}
if (readCount > aCount) {
readCount = aCount;
}
memcpy(aBuf, mUnicharData.Elements() + mUnicharDataOffset,
readCount * sizeof(char16_t));
mUnicharDataOffset += readCount;
*aReadCount = readCount;
return NS_OK;
}
NS_IMETHODIMP
nsConverterInputStream::ReadSegments(nsWriteUnicharSegmentFun aWriter,
void* aClosure,
uint32_t aCount, uint32_t *aReadCount)
{
NS_ASSERTION(mUnicharDataLength >= mUnicharDataOffset, "unsigned madness");
uint32_t bytesToWrite = mUnicharDataLength - mUnicharDataOffset;
nsresult rv;
if (0 == bytesToWrite) {
// Fill the unichar buffer
bytesToWrite = Fill(&rv);
if (bytesToWrite <= 0) {
*aReadCount = 0;
return rv;
}
}
if (bytesToWrite > aCount)
bytesToWrite = aCount;
uint32_t bytesWritten;
uint32_t totalBytesWritten = 0;
while (bytesToWrite) {
rv = aWriter(this, aClosure,
mUnicharData.Elements() + mUnicharDataOffset,
totalBytesWritten, bytesToWrite, &bytesWritten);
if (NS_FAILED(rv)) {
// don't propagate errors to the caller
break;
}
bytesToWrite -= bytesWritten;
totalBytesWritten += bytesWritten;
mUnicharDataOffset += bytesWritten;
}
*aReadCount = totalBytesWritten;
return NS_OK;
}
NS_IMETHODIMP
nsConverterInputStream::ReadString(uint32_t aCount, nsAString& aString,
uint32_t* aReadCount)
{
NS_ASSERTION(mUnicharDataLength >= mUnicharDataOffset, "unsigned madness");
uint32_t readCount = mUnicharDataLength - mUnicharDataOffset;
if (0 == readCount) {
// Fill the unichar buffer
readCount = Fill(&mLastErrorCode);
if (readCount == 0) {
*aReadCount = 0;
return mLastErrorCode;
}
}
if (readCount > aCount) {
readCount = aCount;
}
const char16_t* buf = mUnicharData.Elements() + mUnicharDataOffset;
aString.Assign(buf, readCount);
mUnicharDataOffset += readCount;
*aReadCount = readCount;
return NS_OK;
}
uint32_t
nsConverterInputStream::Fill(nsresult * aErrorCode)
{
if (nullptr == mInput) {
// We already closed the stream!
*aErrorCode = NS_BASE_STREAM_CLOSED;
return 0;
}
if (NS_FAILED(mLastErrorCode)) {
// We failed to completely convert last time, and error-recovery
// is disabled. We will fare no better this time, so...
*aErrorCode = mLastErrorCode;
return 0;
}
// We assume a many to one conversion and are using equal sizes for
// the two buffers. However if an error happens at the very start
// of a byte buffer we may end up in a situation where n bytes lead
// to n+1 unicode chars. Thus we need to keep track of the leftover
// bytes as we convert.
uint32_t nb;
*aErrorCode = NS_FillArray(mByteData, mInput, mLeftOverBytes, &nb);
if (nb == 0 && mLeftOverBytes == 0) {
// No more data
*aErrorCode = NS_OK;
return 0;
}
NS_ASSERTION(uint32_t(nb) + mLeftOverBytes == mByteData.Length(),
"mByteData is lying to us somewhere");
// Now convert as much of the byte buffer to unicode as possible
mUnicharDataOffset = 0;
mUnicharDataLength = 0;
uint32_t srcConsumed = 0;
do {
int32_t srcLen = mByteData.Length() - srcConsumed;
int32_t dstLen = mUnicharData.Capacity() - mUnicharDataLength;
*aErrorCode = mConverter->Convert(mByteData.Elements()+srcConsumed,
&srcLen,
mUnicharData.Elements()+mUnicharDataLength,
&dstLen);
mUnicharDataLength += dstLen;
// XXX if srcLen is negative, we want to drop the _first_ byte in
// the erroneous byte sequence and try again. This is not quite
// possible right now -- see bug 160784
srcConsumed += srcLen;
if (NS_FAILED(*aErrorCode) && mReplacementChar) {
NS_ASSERTION(0 < mUnicharData.Capacity() - mUnicharDataLength,
"Decoder returned an error but filled the output buffer! "
"Should not happen.");
mUnicharData.Elements()[mUnicharDataLength++] = mReplacementChar;
++srcConsumed;
// XXX this is needed to make sure we don't underrun our buffer;
// bug 160784 again
srcConsumed = std::max<uint32_t>(srcConsumed, 0);
mConverter->Reset();
}
NS_ASSERTION(srcConsumed <= mByteData.Length(),
"Whoa. The converter should have returned NS_OK_UDEC_MOREINPUT before this point!");
} while (mReplacementChar &&
NS_FAILED(*aErrorCode) &&
mUnicharData.Capacity() > mUnicharDataLength);
mLeftOverBytes = mByteData.Length() - srcConsumed;
return mUnicharDataLength;
}
NS_IMETHODIMP
nsConverterInputStream::ReadLine(nsAString& aLine, bool* aResult)
{
if (!mLineBuffer) {
mLineBuffer = new nsLineBuffer<char16_t>;
}
return NS_ReadLine(this, mLineBuffer.get(), aLine, aResult);
}
|
