1 // Copyright 2014 PDFium 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 // Original code copyright 2014 Foxit Software Inc. http://www.foxitsoftware.com
6
7 #include "core/fpdfapi/parser/fpdf_parser_decode.h"
8
9 #include <limits.h>
10
11 #include <algorithm>
12 #include <sstream>
13 #include <utility>
14 #include <vector>
15
16 #include "constants/stream_dict_common.h"
17 #include "core/fpdfapi/parser/cpdf_array.h"
18 #include "core/fpdfapi/parser/cpdf_dictionary.h"
19 #include "core/fpdfapi/parser/fpdf_parser_utility.h"
20 #include "core/fxcodec/fax/faxmodule.h"
21 #include "core/fxcodec/flate/flatemodule.h"
22 #include "core/fxcodec/fx_codec.h"
23 #include "core/fxcodec/scanlinedecoder.h"
24 #include "core/fxcrt/fx_extension.h"
25 #include "core/fxcrt/fx_safe_types.h"
26 #include "third_party/base/numerics/safe_math.h"
27 #include "third_party/base/stl_util.h"
28
29 namespace {
30
31 const uint32_t kMaxStreamSize = 20 * 1024 * 1024;
32
GetUnicodeFromBigEndianBytes(const uint8_t * bytes)33 uint16_t GetUnicodeFromBigEndianBytes(const uint8_t* bytes) {
34 return bytes[0] << 8 | bytes[1];
35 }
36
GetUnicodeFromLittleEndianBytes(const uint8_t * bytes)37 uint16_t GetUnicodeFromLittleEndianBytes(const uint8_t* bytes) {
38 return bytes[1] << 8 | bytes[0];
39 }
40
CheckFlateDecodeParams(int Colors,int BitsPerComponent,int Columns)41 bool CheckFlateDecodeParams(int Colors, int BitsPerComponent, int Columns) {
42 if (Colors < 0 || BitsPerComponent < 0 || Columns < 0)
43 return false;
44
45 pdfium::base::CheckedNumeric<int> check = Columns;
46 check *= Colors;
47 check *= BitsPerComponent;
48 if (!check.IsValid())
49 return false;
50
51 return check.ValueOrDie() <= INT_MAX - 7;
52 }
53
GetA85Result(uint32_t res,size_t i)54 uint8_t GetA85Result(uint32_t res, size_t i) {
55 return static_cast<uint8_t>(res >> (3 - i) * 8);
56 }
57
58 } // namespace
59
60 const uint16_t PDFDocEncoding[256] = {
61 0x0000, 0x0001, 0x0002, 0x0003, 0x0004, 0x0005, 0x0006, 0x0007, 0x0008,
62 0x0009, 0x000a, 0x000b, 0x000c, 0x000d, 0x000e, 0x000f, 0x0010, 0x0011,
63 0x0012, 0x0013, 0x0014, 0x0015, 0x0016, 0x0017, 0x02d8, 0x02c7, 0x02c6,
64 0x02d9, 0x02dd, 0x02db, 0x02da, 0x02dc, 0x0020, 0x0021, 0x0022, 0x0023,
65 0x0024, 0x0025, 0x0026, 0x0027, 0x0028, 0x0029, 0x002a, 0x002b, 0x002c,
66 0x002d, 0x002e, 0x002f, 0x0030, 0x0031, 0x0032, 0x0033, 0x0034, 0x0035,
67 0x0036, 0x0037, 0x0038, 0x0039, 0x003a, 0x003b, 0x003c, 0x003d, 0x003e,
68 0x003f, 0x0040, 0x0041, 0x0042, 0x0043, 0x0044, 0x0045, 0x0046, 0x0047,
69 0x0048, 0x0049, 0x004a, 0x004b, 0x004c, 0x004d, 0x004e, 0x004f, 0x0050,
70 0x0051, 0x0052, 0x0053, 0x0054, 0x0055, 0x0056, 0x0057, 0x0058, 0x0059,
71 0x005a, 0x005b, 0x005c, 0x005d, 0x005e, 0x005f, 0x0060, 0x0061, 0x0062,
72 0x0063, 0x0064, 0x0065, 0x0066, 0x0067, 0x0068, 0x0069, 0x006a, 0x006b,
73 0x006c, 0x006d, 0x006e, 0x006f, 0x0070, 0x0071, 0x0072, 0x0073, 0x0074,
74 0x0075, 0x0076, 0x0077, 0x0078, 0x0079, 0x007a, 0x007b, 0x007c, 0x007d,
75 0x007e, 0x0000, 0x2022, 0x2020, 0x2021, 0x2026, 0x2014, 0x2013, 0x0192,
76 0x2044, 0x2039, 0x203a, 0x2212, 0x2030, 0x201e, 0x201c, 0x201d, 0x2018,
77 0x2019, 0x201a, 0x2122, 0xfb01, 0xfb02, 0x0141, 0x0152, 0x0160, 0x0178,
78 0x017d, 0x0131, 0x0142, 0x0153, 0x0161, 0x017e, 0x0000, 0x20ac, 0x00a1,
79 0x00a2, 0x00a3, 0x00a4, 0x00a5, 0x00a6, 0x00a7, 0x00a8, 0x00a9, 0x00aa,
80 0x00ab, 0x00ac, 0x0000, 0x00ae, 0x00af, 0x00b0, 0x00b1, 0x00b2, 0x00b3,
81 0x00b4, 0x00b5, 0x00b6, 0x00b7, 0x00b8, 0x00b9, 0x00ba, 0x00bb, 0x00bc,
82 0x00bd, 0x00be, 0x00bf, 0x00c0, 0x00c1, 0x00c2, 0x00c3, 0x00c4, 0x00c5,
83 0x00c6, 0x00c7, 0x00c8, 0x00c9, 0x00ca, 0x00cb, 0x00cc, 0x00cd, 0x00ce,
84 0x00cf, 0x00d0, 0x00d1, 0x00d2, 0x00d3, 0x00d4, 0x00d5, 0x00d6, 0x00d7,
85 0x00d8, 0x00d9, 0x00da, 0x00db, 0x00dc, 0x00dd, 0x00de, 0x00df, 0x00e0,
86 0x00e1, 0x00e2, 0x00e3, 0x00e4, 0x00e5, 0x00e6, 0x00e7, 0x00e8, 0x00e9,
87 0x00ea, 0x00eb, 0x00ec, 0x00ed, 0x00ee, 0x00ef, 0x00f0, 0x00f1, 0x00f2,
88 0x00f3, 0x00f4, 0x00f5, 0x00f6, 0x00f7, 0x00f8, 0x00f9, 0x00fa, 0x00fb,
89 0x00fc, 0x00fd, 0x00fe, 0x00ff};
90
ValidateDecoderPipeline(const CPDF_Array * pDecoders)91 bool ValidateDecoderPipeline(const CPDF_Array* pDecoders) {
92 size_t count = pDecoders->size();
93 if (count == 0)
94 return true;
95
96 for (size_t i = 0; i < count; ++i) {
97 if (!pDecoders->GetObjectAt(i)->IsName())
98 return false;
99 }
100
101 if (count == 1)
102 return true;
103
104 // TODO(thestig): Consolidate all the places that use these filter names.
105 static const char kValidDecoders[][16] = {
106 "FlateDecode", "Fl", "LZWDecode", "LZW", "ASCII85Decode", "A85",
107 "ASCIIHexDecode", "AHx", "RunLengthDecode", "RL"};
108 for (size_t i = 0; i < count - 1; ++i) {
109 if (!pdfium::ContainsValue(kValidDecoders, pDecoders->GetStringAt(i)))
110 return false;
111 }
112 return true;
113 }
114
A85Decode(pdfium::span<const uint8_t> src_span,std::unique_ptr<uint8_t,FxFreeDeleter> * dest_buf,uint32_t * dest_size)115 uint32_t A85Decode(pdfium::span<const uint8_t> src_span,
116 std::unique_ptr<uint8_t, FxFreeDeleter>* dest_buf,
117 uint32_t* dest_size) {
118 *dest_size = 0;
119 if (src_span.empty()) {
120 dest_buf->reset();
121 return 0;
122 }
123
124 // Count legal characters and zeros.
125 uint32_t zcount = 0;
126 uint32_t pos = 0;
127 while (pos < src_span.size()) {
128 uint8_t ch = src_span[pos];
129 if (ch == 'z') {
130 zcount++;
131 } else if ((ch < '!' || ch > 'u') && !PDFCharIsLineEnding(ch) &&
132 ch != ' ' && ch != '\t') {
133 break;
134 }
135 pos++;
136 }
137 // No content to decode.
138 if (pos == 0)
139 return 0;
140
141 // Count the space needed to contain non-zero characters. The encoding ratio
142 // of Ascii85 is 4:5.
143 uint32_t space_for_non_zeroes = (pos - zcount) / 5 * 4 + 4;
144 FX_SAFE_UINT32 size = zcount;
145 size *= 4;
146 size += space_for_non_zeroes;
147 if (!size.IsValid())
148 return FX_INVALID_OFFSET;
149
150 dest_buf->reset(FX_Alloc(uint8_t, size.ValueOrDie()));
151 uint8_t* dest_buf_ptr = dest_buf->get();
152 size_t state = 0;
153 uint32_t res = 0;
154 pos = 0;
155 while (pos < src_span.size()) {
156 uint8_t ch = src_span[pos++];
157 if (PDFCharIsLineEnding(ch) || ch == ' ' || ch == '\t')
158 continue;
159
160 if (ch == 'z') {
161 memset(dest_buf_ptr + *dest_size, 0, 4);
162 state = 0;
163 res = 0;
164 *dest_size += 4;
165 continue;
166 }
167
168 // Check for the end or illegal character.
169 if (ch < '!' || ch > 'u')
170 break;
171
172 res = res * 85 + ch - 33;
173 if (state < 4) {
174 ++state;
175 continue;
176 }
177
178 for (size_t i = 0; i < 4; ++i) {
179 dest_buf_ptr[(*dest_size)++] = GetA85Result(res, i);
180 }
181 state = 0;
182 res = 0;
183 }
184 // Handle partial group.
185 if (state) {
186 for (size_t i = state; i < 5; ++i)
187 res = res * 85 + 84;
188 for (size_t i = 0; i < state - 1; ++i)
189 dest_buf_ptr[(*dest_size)++] = GetA85Result(res, i);
190 }
191 if (pos < src_span.size() && src_span[pos] == '>')
192 ++pos;
193 return pos;
194 }
195
HexDecode(pdfium::span<const uint8_t> src_span,std::unique_ptr<uint8_t,FxFreeDeleter> * dest_buf,uint32_t * dest_size)196 uint32_t HexDecode(pdfium::span<const uint8_t> src_span,
197 std::unique_ptr<uint8_t, FxFreeDeleter>* dest_buf,
198 uint32_t* dest_size) {
199 *dest_size = 0;
200 if (src_span.empty()) {
201 dest_buf->reset();
202 return 0;
203 }
204
205 uint32_t i = 0;
206 // Find the end of data.
207 while (i < src_span.size() && src_span[i] != '>')
208 ++i;
209
210 dest_buf->reset(FX_Alloc(uint8_t, i / 2 + 1));
211 uint8_t* dest_buf_ptr = dest_buf->get();
212 bool bFirst = true;
213 for (i = 0; i < src_span.size(); ++i) {
214 uint8_t ch = src_span[i];
215 if (PDFCharIsLineEnding(ch) || ch == ' ' || ch == '\t')
216 continue;
217
218 if (ch == '>') {
219 ++i;
220 break;
221 }
222 if (!std::isxdigit(ch))
223 continue;
224
225 int digit = FXSYS_HexCharToInt(ch);
226 if (bFirst)
227 dest_buf_ptr[*dest_size] = digit * 16;
228 else
229 dest_buf_ptr[(*dest_size)++] += digit;
230 bFirst = !bFirst;
231 }
232 if (!bFirst)
233 ++(*dest_size);
234 return i;
235 }
236
RunLengthDecode(pdfium::span<const uint8_t> src_span,std::unique_ptr<uint8_t,FxFreeDeleter> * dest_buf,uint32_t * dest_size)237 uint32_t RunLengthDecode(pdfium::span<const uint8_t> src_span,
238 std::unique_ptr<uint8_t, FxFreeDeleter>* dest_buf,
239 uint32_t* dest_size) {
240 size_t i = 0;
241 *dest_size = 0;
242 while (i < src_span.size()) {
243 if (src_span[i] == 128)
244 break;
245
246 uint32_t old = *dest_size;
247 if (src_span[i] < 128) {
248 *dest_size += src_span[i] + 1;
249 if (*dest_size < old)
250 return FX_INVALID_OFFSET;
251 i += src_span[i] + 2;
252 } else {
253 *dest_size += 257 - src_span[i];
254 if (*dest_size < old)
255 return FX_INVALID_OFFSET;
256 i += 2;
257 }
258 }
259 if (*dest_size >= kMaxStreamSize)
260 return FX_INVALID_OFFSET;
261
262 dest_buf->reset(FX_Alloc(uint8_t, *dest_size));
263 pdfium::span<uint8_t> dest_span(dest_buf->get(), *dest_size);
264 i = 0;
265 int dest_count = 0;
266 while (i < src_span.size()) {
267 if (src_span[i] == 128)
268 break;
269
270 if (src_span[i] < 128) {
271 uint32_t copy_len = src_span[i] + 1;
272 uint32_t buf_left = src_span.size() - i - 1;
273 if (buf_left < copy_len) {
274 uint32_t delta = copy_len - buf_left;
275 copy_len = buf_left;
276 memset(&dest_span[dest_count + copy_len], '\0', delta);
277 }
278 auto copy_span = src_span.subspan(i + 1, copy_len);
279 memcpy(&dest_span[dest_count], copy_span.data(), copy_span.size());
280 dest_count += src_span[i] + 1;
281 i += src_span[i] + 2;
282 } else {
283 int fill = 0;
284 if (i < src_span.size() - 1)
285 fill = src_span[i + 1];
286 memset(&dest_span[dest_count], fill, 257 - src_span[i]);
287 dest_count += 257 - src_span[i];
288 i += 2;
289 }
290 }
291 return std::min(i + 1, src_span.size());
292 }
293
CreateFaxDecoder(pdfium::span<const uint8_t> src_span,int width,int height,const CPDF_Dictionary * pParams)294 std::unique_ptr<ScanlineDecoder> CreateFaxDecoder(
295 pdfium::span<const uint8_t> src_span,
296 int width,
297 int height,
298 const CPDF_Dictionary* pParams) {
299 int K = 0;
300 bool EndOfLine = false;
301 bool ByteAlign = false;
302 bool BlackIs1 = false;
303 int Columns = 1728;
304 int Rows = 0;
305 if (pParams) {
306 K = pParams->GetIntegerFor("K");
307 EndOfLine = !!pParams->GetIntegerFor("EndOfLine");
308 ByteAlign = !!pParams->GetIntegerFor("EncodedByteAlign");
309 BlackIs1 = !!pParams->GetIntegerFor("BlackIs1");
310 Columns = pParams->GetIntegerFor("Columns", 1728);
311 Rows = pParams->GetIntegerFor("Rows");
312 if (Rows > USHRT_MAX)
313 Rows = 0;
314 }
315 return FaxModule::CreateDecoder(src_span, width, height, K, EndOfLine,
316 ByteAlign, BlackIs1, Columns, Rows);
317 }
318
CreateFlateDecoder(pdfium::span<const uint8_t> src_span,int width,int height,int nComps,int bpc,const CPDF_Dictionary * pParams)319 std::unique_ptr<ScanlineDecoder> CreateFlateDecoder(
320 pdfium::span<const uint8_t> src_span,
321 int width,
322 int height,
323 int nComps,
324 int bpc,
325 const CPDF_Dictionary* pParams) {
326 int predictor = 0;
327 int Colors = 0;
328 int BitsPerComponent = 0;
329 int Columns = 0;
330 if (pParams) {
331 predictor = pParams->GetIntegerFor("Predictor");
332 Colors = pParams->GetIntegerFor("Colors", 1);
333 BitsPerComponent = pParams->GetIntegerFor("BitsPerComponent", 8);
334 Columns = pParams->GetIntegerFor("Columns", 1);
335 if (!CheckFlateDecodeParams(Colors, BitsPerComponent, Columns))
336 return nullptr;
337 }
338 return FlateModule::CreateDecoder(src_span, width, height, nComps, bpc,
339 predictor, Colors, BitsPerComponent,
340 Columns);
341 }
342
FlateOrLZWDecode(bool bLZW,pdfium::span<const uint8_t> src_span,const CPDF_Dictionary * pParams,uint32_t estimated_size,std::unique_ptr<uint8_t,FxFreeDeleter> * dest_buf,uint32_t * dest_size)343 uint32_t FlateOrLZWDecode(bool bLZW,
344 pdfium::span<const uint8_t> src_span,
345 const CPDF_Dictionary* pParams,
346 uint32_t estimated_size,
347 std::unique_ptr<uint8_t, FxFreeDeleter>* dest_buf,
348 uint32_t* dest_size) {
349 int predictor = 0;
350 int Colors = 0;
351 int BitsPerComponent = 0;
352 int Columns = 0;
353 bool bEarlyChange = true;
354 if (pParams) {
355 predictor = pParams->GetIntegerFor("Predictor");
356 bEarlyChange = !!pParams->GetIntegerFor("EarlyChange", 1);
357 Colors = pParams->GetIntegerFor("Colors", 1);
358 BitsPerComponent = pParams->GetIntegerFor("BitsPerComponent", 8);
359 Columns = pParams->GetIntegerFor("Columns", 1);
360 if (!CheckFlateDecodeParams(Colors, BitsPerComponent, Columns))
361 return FX_INVALID_OFFSET;
362 }
363 return FlateModule::FlateOrLZWDecode(bLZW, src_span, bEarlyChange, predictor,
364 Colors, BitsPerComponent, Columns,
365 estimated_size, dest_buf, dest_size);
366 }
367
368 Optional<std::vector<std::pair<ByteString, const CPDF_Object*>>>
GetDecoderArray(const CPDF_Dictionary * pDict)369 GetDecoderArray(const CPDF_Dictionary* pDict) {
370 const CPDF_Object* pDecoder = pDict->GetDirectObjectFor("Filter");
371 if (!pDecoder || (!pDecoder->IsArray() && !pDecoder->IsName()))
372 return {};
373
374 const CPDF_Object* pParams =
375 pDict->GetDirectObjectFor(pdfium::stream::kDecodeParms);
376
377 std::vector<std::pair<ByteString, const CPDF_Object*>> decoder_array;
378 if (const CPDF_Array* pDecoders = pDecoder->AsArray()) {
379 if (!ValidateDecoderPipeline(pDecoders))
380 return {};
381
382 const CPDF_Array* pParamsArray = ToArray(pParams);
383 for (size_t i = 0; i < pDecoders->size(); ++i) {
384 decoder_array.push_back(
385 {pDecoders->GetStringAt(i),
386 pParamsArray ? pParamsArray->GetDictAt(i) : nullptr});
387 }
388 } else {
389 decoder_array.push_back(
390 {pDecoder->GetString(), pParams ? pParams->GetDict() : nullptr});
391 }
392
393 return decoder_array;
394 }
395
PDF_DataDecode(pdfium::span<const uint8_t> src_span,uint32_t last_estimated_size,bool bImageAcc,const std::vector<std::pair<ByteString,const CPDF_Object * >> & decoder_array,std::unique_ptr<uint8_t,FxFreeDeleter> * dest_buf,uint32_t * dest_size,ByteString * ImageEncoding,RetainPtr<const CPDF_Dictionary> * pImageParams)396 bool PDF_DataDecode(
397 pdfium::span<const uint8_t> src_span,
398 uint32_t last_estimated_size,
399 bool bImageAcc,
400 const std::vector<std::pair<ByteString, const CPDF_Object*>>& decoder_array,
401 std::unique_ptr<uint8_t, FxFreeDeleter>* dest_buf,
402 uint32_t* dest_size,
403 ByteString* ImageEncoding,
404 RetainPtr<const CPDF_Dictionary>* pImageParams) {
405 std::unique_ptr<uint8_t, FxFreeDeleter> result;
406 // May be changed to point to |result| in the for-loop below. So put it below
407 // |result| and let it get destroyed first.
408 pdfium::span<const uint8_t> last_span = src_span;
409 size_t nSize = decoder_array.size();
410 for (size_t i = 0; i < nSize; ++i) {
411 int estimated_size = i == nSize - 1 ? last_estimated_size : 0;
412 ByteString decoder = decoder_array[i].first;
413 const CPDF_Dictionary* pParam = ToDictionary(decoder_array[i].second);
414 std::unique_ptr<uint8_t, FxFreeDeleter> new_buf;
415 uint32_t new_size = 0xFFFFFFFF;
416 uint32_t offset = FX_INVALID_OFFSET;
417 if (decoder == "Crypt")
418 continue;
419 if (decoder == "FlateDecode" || decoder == "Fl") {
420 if (bImageAcc && i == nSize - 1) {
421 *ImageEncoding = "FlateDecode";
422 *dest_buf = std::move(result);
423 *dest_size = last_span.size();
424 pImageParams->Reset(pParam);
425 return true;
426 }
427 offset = FlateOrLZWDecode(false, last_span, pParam, estimated_size,
428 &new_buf, &new_size);
429 } else if (decoder == "LZWDecode" || decoder == "LZW") {
430 offset = FlateOrLZWDecode(true, last_span, pParam, estimated_size,
431 &new_buf, &new_size);
432 } else if (decoder == "ASCII85Decode" || decoder == "A85") {
433 offset = A85Decode(last_span, &new_buf, &new_size);
434 } else if (decoder == "ASCIIHexDecode" || decoder == "AHx") {
435 offset = HexDecode(last_span, &new_buf, &new_size);
436 } else if (decoder == "RunLengthDecode" || decoder == "RL") {
437 if (bImageAcc && i == nSize - 1) {
438 *ImageEncoding = "RunLengthDecode";
439 *dest_buf = std::move(result);
440 *dest_size = last_span.size();
441 pImageParams->Reset(pParam);
442 return true;
443 }
444 offset = RunLengthDecode(last_span, &new_buf, &new_size);
445 } else {
446 // If we get here, assume it's an image decoder.
447 if (decoder == "DCT")
448 decoder = "DCTDecode";
449 else if (decoder == "CCF")
450 decoder = "CCITTFaxDecode";
451 *ImageEncoding = std::move(decoder);
452 pImageParams->Reset(pParam);
453 *dest_buf = std::move(result);
454 *dest_size = last_span.size();
455 return true;
456 }
457 if (offset == FX_INVALID_OFFSET)
458 return false;
459
460 last_span = {new_buf.get(), new_size};
461 result = std::move(new_buf);
462 }
463 ImageEncoding->clear();
464 *pImageParams = nullptr;
465 *dest_buf = std::move(result);
466 *dest_size = last_span.size();
467 return true;
468 }
469
PDF_DecodeText(pdfium::span<const uint8_t> span)470 WideString PDF_DecodeText(pdfium::span<const uint8_t> span) {
471 int dest_pos = 0;
472 WideString result;
473 if (span.size() >= 2 && ((span[0] == 0xfe && span[1] == 0xff) ||
474 (span[0] == 0xff && span[1] == 0xfe))) {
475 size_t max_chars = (span.size() - 2) / 2;
476 if (!max_chars)
477 return result;
478
479 pdfium::span<wchar_t> dest_buf = result.GetBuffer(max_chars);
480 uint16_t (*GetUnicodeFromBytes)(const uint8_t*) =
481 span[0] == 0xfe ? GetUnicodeFromBigEndianBytes
482 : GetUnicodeFromLittleEndianBytes;
483 const uint8_t* unicode_str = &span[2];
484 for (size_t i = 0; i < max_chars * 2; i += 2) {
485 uint16_t unicode = GetUnicodeFromBytes(unicode_str + i);
486
487 // 0x001B is a begin/end marker for language metadata region that
488 // should not be in the decoded text.
489 if (unicode == 0x001B) {
490 i += 2;
491 for (; i < max_chars * 2; i += 2) {
492 unicode = GetUnicodeFromBytes(unicode_str + i);
493 if (unicode == 0x001B) {
494 i += 2;
495 if (i < max_chars * 2)
496 unicode = GetUnicodeFromBytes(unicode_str + i);
497 break;
498 }
499 }
500 if (i >= max_chars * 2)
501 break;
502 }
503
504 dest_buf[dest_pos++] = unicode;
505 }
506 } else {
507 pdfium::span<wchar_t> dest_buf = result.GetBuffer(span.size());
508 for (size_t i = 0; i < span.size(); ++i)
509 dest_buf[i] = PDFDocEncoding[span[i]];
510 dest_pos = span.size();
511 }
512 result.ReleaseBuffer(dest_pos);
513 return result;
514 }
515
PDF_EncodeText(const WideString & str)516 ByteString PDF_EncodeText(const WideString& str) {
517 size_t i = 0;
518 size_t len = str.GetLength();
519 ByteString result;
520 {
521 pdfium::span<char> dest_buf = result.GetBuffer(len);
522 for (i = 0; i < len; ++i) {
523 int code;
524 for (code = 0; code < 256; ++code) {
525 if (PDFDocEncoding[code] == str[i])
526 break;
527 }
528 if (code == 256)
529 break;
530
531 dest_buf[i] = code;
532 }
533 }
534 result.ReleaseBuffer(i);
535 if (i == len)
536 return result;
537
538 if (len > INT_MAX / 2 - 1) {
539 result.ReleaseBuffer(0);
540 return result;
541 }
542
543 size_t dest_index = 0;
544 size_t encLen = len * 2 + 2;
545 {
546 pdfium::span<uint8_t> dest_buf =
547 pdfium::as_writable_bytes(result.GetBuffer(encLen));
548 dest_buf[dest_index++] = 0xfe;
549 dest_buf[dest_index++] = 0xff;
550 for (size_t j = 0; j < len; ++j) {
551 dest_buf[dest_index++] = str[j] >> 8;
552 dest_buf[dest_index++] = static_cast<uint8_t>(str[j]);
553 }
554 }
555 result.ReleaseBuffer(encLen);
556 return result;
557 }
558
PDF_EncodeString(const ByteString & src,bool bHex)559 ByteString PDF_EncodeString(const ByteString& src, bool bHex) {
560 std::ostringstream result;
561 int srclen = src.GetLength();
562 if (bHex) {
563 result << '<';
564 for (int i = 0; i < srclen; ++i) {
565 char buf[2];
566 FXSYS_IntToTwoHexChars(src[i], buf);
567 result << buf[0];
568 result << buf[1];
569 }
570 result << '>';
571 return ByteString(result);
572 }
573 result << '(';
574 for (int i = 0; i < srclen; ++i) {
575 uint8_t ch = src[i];
576 if (ch == 0x0a) {
577 result << "\\n";
578 continue;
579 }
580 if (ch == 0x0d) {
581 result << "\\r";
582 continue;
583 }
584 if (ch == ')' || ch == '\\' || ch == '(')
585 result << '\\';
586 result << static_cast<char>(ch);
587 }
588 result << ')';
589 return ByteString(result);
590 }
591
FlateEncode(pdfium::span<const uint8_t> src_span,std::unique_ptr<uint8_t,FxFreeDeleter> * dest_buf,uint32_t * dest_size)592 bool FlateEncode(pdfium::span<const uint8_t> src_span,
593 std::unique_ptr<uint8_t, FxFreeDeleter>* dest_buf,
594 uint32_t* dest_size) {
595 return FlateModule::Encode(src_span.data(), src_span.size(), dest_buf,
596 dest_size);
597 }
598
FlateDecode(pdfium::span<const uint8_t> src_span,std::unique_ptr<uint8_t,FxFreeDeleter> * dest_buf,uint32_t * dest_size)599 uint32_t FlateDecode(pdfium::span<const uint8_t> src_span,
600 std::unique_ptr<uint8_t, FxFreeDeleter>* dest_buf,
601 uint32_t* dest_size) {
602 return FlateModule::FlateOrLZWDecode(false, src_span, false, 0, 0, 0, 0, 0,
603 dest_buf, dest_size);
604 }
605