| OLD | NEW |
| (Empty) |
| 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 <algorithm> | |
| 8 #include <cctype> | |
| 9 #include <cwctype> | |
| 10 #include <memory> | |
| 11 #include <utility> | |
| 12 #include <vector> | |
| 13 | |
| 14 #include "core/fpdfapi/fpdf_font/include/cpdf_font.h" | |
| 15 #include "core/fpdfapi/fpdf_page/include/cpdf_form.h" | |
| 16 #include "core/fpdfapi/fpdf_page/include/cpdf_formobject.h" | |
| 17 #include "core/fpdfapi/fpdf_page/include/cpdf_page.h" | |
| 18 #include "core/fpdfapi/fpdf_page/include/cpdf_pageobject.h" | |
| 19 #include "core/fpdfapi/fpdf_page/include/cpdf_textobject.h" | |
| 20 #include "core/fpdfapi/fpdf_parser/include/cpdf_dictionary.h" | |
| 21 #include "core/fpdfapi/fpdf_parser/include/cpdf_string.h" | |
| 22 #include "core/fpdftext/include/cpdf_linkextract.h" | |
| 23 #include "core/fpdftext/include/cpdf_textpage.h" | |
| 24 #include "core/fpdftext/include/cpdf_textpagefind.h" | |
| 25 #include "core/fpdftext/unicodenormalizationdata.h" | |
| 26 #include "core/fxcrt/fx_bidi.h" | |
| 27 #include "core/fxcrt/include/fx_ext.h" | |
| 28 #include "core/fxcrt/include/fx_ucd.h" | |
| 29 #include "third_party/base/stl_util.h" | |
| 30 | |
| 31 #define FPDFTEXT_MATCHCASE 0x00000001 | |
| 32 #define FPDFTEXT_MATCHWHOLEWORD 0x00000002 | |
| 33 #define FPDFTEXT_CONSECUTIVE 0x00000004 | |
| 34 | |
| 35 #define FPDFTEXT_CHAR_ERROR -1 | |
| 36 #define FPDFTEXT_CHAR_NORMAL 0 | |
| 37 #define FPDFTEXT_CHAR_GENERATED 1 | |
| 38 #define FPDFTEXT_CHAR_UNUNICODE 2 | |
| 39 #define FPDFTEXT_CHAR_HYPHEN 3 | |
| 40 #define FPDFTEXT_CHAR_PIECE 4 | |
| 41 | |
| 42 #define TEXT_SPACE_CHAR L' ' | |
| 43 #define TEXT_LINEFEED_CHAR L'\n' | |
| 44 #define TEXT_RETURN_CHAR L'\r' | |
| 45 #define TEXT_EMPTY L"" | |
| 46 #define TEXT_SPACE L" " | |
| 47 #define TEXT_RETURN_LINEFEED L"\r\n" | |
| 48 #define TEXT_LINEFEED L"\n" | |
| 49 #define TEXT_CHARRATIO_GAPDELTA 0.070 | |
| 50 | |
| 51 namespace { | |
| 52 | |
| 53 const FX_FLOAT kDefaultFontSize = 1.0f; | |
| 54 const uint16_t* const g_UnicodeData_Normalization_Maps[5] = { | |
| 55 nullptr, g_UnicodeData_Normalization_Map1, g_UnicodeData_Normalization_Map2, | |
| 56 g_UnicodeData_Normalization_Map3, g_UnicodeData_Normalization_Map4}; | |
| 57 | |
| 58 FX_BOOL IsIgnoreSpaceCharacter(FX_WCHAR curChar) { | |
| 59 if (curChar < 255) | |
| 60 return FALSE; | |
| 61 if ((curChar >= 0x0600 && curChar <= 0x06FF) || | |
| 62 (curChar >= 0xFE70 && curChar <= 0xFEFF) || | |
| 63 (curChar >= 0xFB50 && curChar <= 0xFDFF) || | |
| 64 (curChar >= 0x0400 && curChar <= 0x04FF) || | |
| 65 (curChar >= 0x0500 && curChar <= 0x052F) || | |
| 66 (curChar >= 0xA640 && curChar <= 0xA69F) || | |
| 67 (curChar >= 0x2DE0 && curChar <= 0x2DFF) || curChar == 8467 || | |
| 68 (curChar >= 0x2000 && curChar <= 0x206F)) { | |
| 69 return FALSE; | |
| 70 } | |
| 71 return TRUE; | |
| 72 } | |
| 73 | |
| 74 FX_FLOAT NormalizeThreshold(FX_FLOAT threshold) { | |
| 75 if (threshold < 300) | |
| 76 return threshold / 2.0f; | |
| 77 if (threshold < 500) | |
| 78 return threshold / 4.0f; | |
| 79 if (threshold < 700) | |
| 80 return threshold / 5.0f; | |
| 81 return threshold / 6.0f; | |
| 82 } | |
| 83 | |
| 84 FX_FLOAT CalculateBaseSpace(const CPDF_TextObject* pTextObj, | |
| 85 const CFX_Matrix& matrix) { | |
| 86 FX_FLOAT baseSpace = 0.0; | |
| 87 const int nItems = pTextObj->CountItems(); | |
| 88 if (pTextObj->m_TextState.GetObject()->m_CharSpace && nItems >= 3) { | |
| 89 bool bAllChar = true; | |
| 90 FX_FLOAT spacing = matrix.TransformDistance( | |
| 91 pTextObj->m_TextState.GetObject()->m_CharSpace); | |
| 92 baseSpace = spacing; | |
| 93 for (int i = 0; i < nItems; i++) { | |
| 94 CPDF_TextObjectItem item; | |
| 95 pTextObj->GetItemInfo(i, &item); | |
| 96 if (item.m_CharCode == static_cast<uint32_t>(-1)) { | |
| 97 FX_FLOAT fontsize_h = pTextObj->m_TextState.GetFontSizeH(); | |
| 98 FX_FLOAT kerning = -fontsize_h * item.m_OriginX / 1000; | |
| 99 baseSpace = std::min(baseSpace, kerning + spacing); | |
| 100 bAllChar = false; | |
| 101 } | |
| 102 } | |
| 103 if (baseSpace < 0.0 || (nItems == 3 && !bAllChar)) | |
| 104 baseSpace = 0.0; | |
| 105 } | |
| 106 return baseSpace; | |
| 107 } | |
| 108 | |
| 109 FX_STRSIZE Unicode_GetNormalization(FX_WCHAR wch, FX_WCHAR* pDst) { | |
| 110 wch = wch & 0xFFFF; | |
| 111 FX_WCHAR wFind = g_UnicodeData_Normalization[wch]; | |
| 112 if (!wFind) { | |
| 113 if (pDst) | |
| 114 *pDst = wch; | |
| 115 return 1; | |
| 116 } | |
| 117 if (wFind >= 0x8000) { | |
| 118 wch = wFind - 0x8000; | |
| 119 wFind = 1; | |
| 120 } else { | |
| 121 wch = wFind & 0x0FFF; | |
| 122 wFind >>= 12; | |
| 123 } | |
| 124 const uint16_t* pMap = g_UnicodeData_Normalization_Maps[wFind]; | |
| 125 if (pMap == g_UnicodeData_Normalization_Map4) { | |
| 126 pMap = g_UnicodeData_Normalization_Map4 + wch; | |
| 127 wFind = (FX_WCHAR)(*pMap++); | |
| 128 } else { | |
| 129 pMap += wch; | |
| 130 } | |
| 131 if (pDst) { | |
| 132 FX_WCHAR n = wFind; | |
| 133 while (n--) | |
| 134 *pDst++ = *pMap++; | |
| 135 } | |
| 136 return (FX_STRSIZE)wFind; | |
| 137 } | |
| 138 | |
| 139 float MaskPercentFilled(const std::vector<bool>& mask, | |
| 140 int32_t start, | |
| 141 int32_t end) { | |
| 142 if (start >= end) | |
| 143 return 0; | |
| 144 float count = std::count_if(mask.begin() + start, mask.begin() + end, | |
| 145 [](bool r) { return r; }); | |
| 146 return count / (end - start); | |
| 147 } | |
| 148 | |
| 149 } // namespace | |
| 150 | |
| 151 CPDF_TextPage::CPDF_TextPage(const CPDF_Page* pPage, FPDFText_Direction flags) | |
| 152 : m_pPage(pPage), | |
| 153 m_parserflag(flags), | |
| 154 m_pPreTextObj(nullptr), | |
| 155 m_bIsParsed(false), | |
| 156 m_TextlineDir(TextOrientation::Unknown) { | |
| 157 m_TextBuf.EstimateSize(0, 10240); | |
| 158 pPage->GetDisplayMatrix(m_DisplayMatrix, 0, 0, (int)pPage->GetPageWidth(), | |
| 159 (int)pPage->GetPageHeight(), 0); | |
| 160 } | |
| 161 | |
| 162 CPDF_TextPage::~CPDF_TextPage() {} | |
| 163 | |
| 164 bool CPDF_TextPage::IsControlChar(const PAGECHAR_INFO& charInfo) { | |
| 165 switch (charInfo.m_Unicode) { | |
| 166 case 0x2: | |
| 167 case 0x3: | |
| 168 case 0x93: | |
| 169 case 0x94: | |
| 170 case 0x96: | |
| 171 case 0x97: | |
| 172 case 0x98: | |
| 173 case 0xfffe: | |
| 174 return charInfo.m_Flag != FPDFTEXT_CHAR_HYPHEN; | |
| 175 default: | |
| 176 return false; | |
| 177 } | |
| 178 } | |
| 179 | |
| 180 void CPDF_TextPage::ParseTextPage() { | |
| 181 m_bIsParsed = false; | |
| 182 m_TextBuf.Clear(); | |
| 183 m_CharList.clear(); | |
| 184 m_pPreTextObj = nullptr; | |
| 185 ProcessObject(); | |
| 186 | |
| 187 m_bIsParsed = true; | |
| 188 m_CharIndex.clear(); | |
| 189 int nCount = pdfium::CollectionSize<int>(m_CharList); | |
| 190 if (nCount) | |
| 191 m_CharIndex.push_back(0); | |
| 192 | |
| 193 for (int i = 0; i < nCount; i++) { | |
| 194 int indexSize = pdfium::CollectionSize<int>(m_CharIndex); | |
| 195 const PAGECHAR_INFO& charinfo = m_CharList[i]; | |
| 196 if (charinfo.m_Flag == FPDFTEXT_CHAR_GENERATED || | |
| 197 (charinfo.m_Unicode != 0 && !IsControlChar(charinfo))) { | |
| 198 if (indexSize % 2) { | |
| 199 m_CharIndex.push_back(1); | |
| 200 } else { | |
| 201 if (indexSize <= 0) | |
| 202 continue; | |
| 203 m_CharIndex[indexSize - 1] += 1; | |
| 204 } | |
| 205 } else { | |
| 206 if (indexSize % 2) { | |
| 207 if (indexSize <= 0) | |
| 208 continue; | |
| 209 m_CharIndex[indexSize - 1] = i + 1; | |
| 210 } else { | |
| 211 m_CharIndex.push_back(i + 1); | |
| 212 } | |
| 213 } | |
| 214 } | |
| 215 int indexSize = pdfium::CollectionSize<int>(m_CharIndex); | |
| 216 if (indexSize % 2) | |
| 217 m_CharIndex.erase(m_CharIndex.begin() + indexSize - 1); | |
| 218 } | |
| 219 | |
| 220 int CPDF_TextPage::CountChars() const { | |
| 221 return pdfium::CollectionSize<int>(m_CharList); | |
| 222 } | |
| 223 | |
| 224 int CPDF_TextPage::CharIndexFromTextIndex(int TextIndex) const { | |
| 225 int indexSize = pdfium::CollectionSize<int>(m_CharIndex); | |
| 226 int count = 0; | |
| 227 for (int i = 0; i < indexSize; i += 2) { | |
| 228 count += m_CharIndex[i + 1]; | |
| 229 if (count > TextIndex) | |
| 230 return TextIndex - count + m_CharIndex[i + 1] + m_CharIndex[i]; | |
| 231 } | |
| 232 return -1; | |
| 233 } | |
| 234 | |
| 235 int CPDF_TextPage::TextIndexFromCharIndex(int CharIndex) const { | |
| 236 int indexSize = pdfium::CollectionSize<int>(m_CharIndex); | |
| 237 int count = 0; | |
| 238 for (int i = 0; i < indexSize; i += 2) { | |
| 239 count += m_CharIndex[i + 1]; | |
| 240 if (m_CharIndex[i + 1] + m_CharIndex[i] > CharIndex) { | |
| 241 if (CharIndex - m_CharIndex[i] < 0) | |
| 242 return -1; | |
| 243 | |
| 244 return CharIndex - m_CharIndex[i] + count - m_CharIndex[i + 1]; | |
| 245 } | |
| 246 } | |
| 247 return -1; | |
| 248 } | |
| 249 | |
| 250 std::vector<CFX_FloatRect> CPDF_TextPage::GetRectArray(int start, | |
| 251 int nCount) const { | |
| 252 if (start < 0 || nCount == 0 || !m_bIsParsed) | |
| 253 return std::vector<CFX_FloatRect>(); | |
| 254 | |
| 255 if (nCount + start > pdfium::CollectionSize<int>(m_CharList) || | |
| 256 nCount == -1) { | |
| 257 nCount = pdfium::CollectionSize<int>(m_CharList) - start; | |
| 258 } | |
| 259 | |
| 260 std::vector<CFX_FloatRect> rectArray; | |
| 261 CPDF_TextObject* pCurObj = nullptr; | |
| 262 CFX_FloatRect rect; | |
| 263 int curPos = start; | |
| 264 bool bFlagNewRect = true; | |
| 265 while (nCount--) { | |
| 266 PAGECHAR_INFO info_curchar = m_CharList[curPos++]; | |
| 267 if (info_curchar.m_Flag == FPDFTEXT_CHAR_GENERATED) | |
| 268 continue; | |
| 269 if (info_curchar.m_CharBox.Width() < 0.01 || | |
| 270 info_curchar.m_CharBox.Height() < 0.01) { | |
| 271 continue; | |
| 272 } | |
| 273 if (!pCurObj) | |
| 274 pCurObj = info_curchar.m_pTextObj; | |
| 275 if (pCurObj != info_curchar.m_pTextObj) { | |
| 276 rectArray.push_back(rect); | |
| 277 pCurObj = info_curchar.m_pTextObj; | |
| 278 bFlagNewRect = true; | |
| 279 } | |
| 280 if (bFlagNewRect) { | |
| 281 FX_FLOAT orgX = info_curchar.m_OriginX, orgY = info_curchar.m_OriginY; | |
| 282 CFX_Matrix matrix, matrix_reverse; | |
| 283 info_curchar.m_pTextObj->GetTextMatrix(&matrix); | |
| 284 matrix.Concat(info_curchar.m_Matrix); | |
| 285 matrix_reverse.SetReverse(matrix); | |
| 286 matrix_reverse.Transform(orgX, orgY); | |
| 287 rect.left = info_curchar.m_CharBox.left; | |
| 288 rect.right = info_curchar.m_CharBox.right; | |
| 289 if (pCurObj->GetFont()->GetTypeDescent()) { | |
| 290 rect.bottom = orgY + | |
| 291 pCurObj->GetFont()->GetTypeDescent() * | |
| 292 pCurObj->GetFontSize() / 1000; | |
| 293 FX_FLOAT xPosTemp = orgX; | |
| 294 matrix.Transform(xPosTemp, rect.bottom); | |
| 295 } else { | |
| 296 rect.bottom = info_curchar.m_CharBox.bottom; | |
| 297 } | |
| 298 if (pCurObj->GetFont()->GetTypeAscent()) { | |
| 299 rect.top = | |
| 300 orgY + | |
| 301 pCurObj->GetFont()->GetTypeAscent() * pCurObj->GetFontSize() / 1000; | |
| 302 FX_FLOAT xPosTemp = | |
| 303 orgX + | |
| 304 GetCharWidth(info_curchar.m_CharCode, pCurObj->GetFont()) * | |
| 305 pCurObj->GetFontSize() / 1000; | |
| 306 matrix.Transform(xPosTemp, rect.top); | |
| 307 } else { | |
| 308 rect.top = info_curchar.m_CharBox.top; | |
| 309 } | |
| 310 bFlagNewRect = false; | |
| 311 rect = info_curchar.m_CharBox; | |
| 312 rect.Normalize(); | |
| 313 } else { | |
| 314 info_curchar.m_CharBox.Normalize(); | |
| 315 rect.left = std::min(rect.left, info_curchar.m_CharBox.left); | |
| 316 rect.right = std::max(rect.right, info_curchar.m_CharBox.right); | |
| 317 rect.top = std::max(rect.top, info_curchar.m_CharBox.top); | |
| 318 rect.bottom = std::min(rect.bottom, info_curchar.m_CharBox.bottom); | |
| 319 } | |
| 320 } | |
| 321 rectArray.push_back(rect); | |
| 322 return rectArray; | |
| 323 } | |
| 324 | |
| 325 int CPDF_TextPage::GetIndexAtPos(CFX_FloatPoint point, | |
| 326 FX_FLOAT xTolerance, | |
| 327 FX_FLOAT yTolerance) const { | |
| 328 if (!m_bIsParsed) | |
| 329 return -3; | |
| 330 | |
| 331 int pos = 0; | |
| 332 int NearPos = -1; | |
| 333 double xdif = 5000; | |
| 334 double ydif = 5000; | |
| 335 while (pos < pdfium::CollectionSize<int>(m_CharList)) { | |
| 336 PAGECHAR_INFO charinfo = m_CharList[pos]; | |
| 337 CFX_FloatRect charrect = charinfo.m_CharBox; | |
| 338 if (charrect.Contains(point.x, point.y)) | |
| 339 break; | |
| 340 if (xTolerance > 0 || yTolerance > 0) { | |
| 341 CFX_FloatRect charRectExt; | |
| 342 charrect.Normalize(); | |
| 343 charRectExt.left = charrect.left - xTolerance / 2; | |
| 344 charRectExt.right = charrect.right + xTolerance / 2; | |
| 345 charRectExt.top = charrect.top + yTolerance / 2; | |
| 346 charRectExt.bottom = charrect.bottom - yTolerance / 2; | |
| 347 if (charRectExt.Contains(point.x, point.y)) { | |
| 348 double curXdif, curYdif; | |
| 349 curXdif = FXSYS_fabs(point.x - charrect.left) < | |
| 350 FXSYS_fabs(point.x - charrect.right) | |
| 351 ? FXSYS_fabs(point.x - charrect.left) | |
| 352 : FXSYS_fabs(point.x - charrect.right); | |
| 353 curYdif = FXSYS_fabs(point.y - charrect.bottom) < | |
| 354 FXSYS_fabs(point.y - charrect.top) | |
| 355 ? FXSYS_fabs(point.y - charrect.bottom) | |
| 356 : FXSYS_fabs(point.y - charrect.top); | |
| 357 if (curYdif + curXdif < xdif + ydif) { | |
| 358 ydif = curYdif; | |
| 359 xdif = curXdif; | |
| 360 NearPos = pos; | |
| 361 } | |
| 362 } | |
| 363 } | |
| 364 ++pos; | |
| 365 } | |
| 366 return pos < pdfium::CollectionSize<int>(m_CharList) ? pos : NearPos; | |
| 367 } | |
| 368 | |
| 369 CFX_WideString CPDF_TextPage::GetTextByRect(const CFX_FloatRect& rect) const { | |
| 370 if (!m_bIsParsed) | |
| 371 return CFX_WideString(); | |
| 372 | |
| 373 FX_FLOAT posy = 0; | |
| 374 bool IsContainPreChar = false; | |
| 375 bool IsAddLineFeed = false; | |
| 376 CFX_WideString strText; | |
| 377 for (const auto& charinfo : m_CharList) { | |
| 378 if (IsRectIntersect(rect, charinfo.m_CharBox)) { | |
| 379 if (FXSYS_fabs(posy - charinfo.m_OriginY) > 0 && !IsContainPreChar && | |
| 380 IsAddLineFeed) { | |
| 381 posy = charinfo.m_OriginY; | |
| 382 if (!strText.IsEmpty()) | |
| 383 strText += L"\r\n"; | |
| 384 } | |
| 385 IsContainPreChar = true; | |
| 386 IsAddLineFeed = false; | |
| 387 if (charinfo.m_Unicode) | |
| 388 strText += charinfo.m_Unicode; | |
| 389 } else if (charinfo.m_Unicode == 32) { | |
| 390 if (IsContainPreChar && charinfo.m_Unicode) { | |
| 391 strText += charinfo.m_Unicode; | |
| 392 IsContainPreChar = false; | |
| 393 IsAddLineFeed = false; | |
| 394 } | |
| 395 } else { | |
| 396 IsContainPreChar = false; | |
| 397 IsAddLineFeed = true; | |
| 398 } | |
| 399 } | |
| 400 return strText; | |
| 401 } | |
| 402 | |
| 403 int CPDF_TextPage::GetIndexAtPos(FX_FLOAT x, | |
| 404 FX_FLOAT y, | |
| 405 FX_FLOAT xTolerance, | |
| 406 FX_FLOAT yTolerance) const { | |
| 407 CFX_FloatPoint point(x, y); | |
| 408 return GetIndexAtPos(point, xTolerance, yTolerance); | |
| 409 } | |
| 410 | |
| 411 void CPDF_TextPage::GetCharInfo(int index, FPDF_CHAR_INFO* info) const { | |
| 412 if (!m_bIsParsed) | |
| 413 return; | |
| 414 | |
| 415 if (index < 0 || index >= pdfium::CollectionSize<int>(m_CharList)) | |
| 416 return; | |
| 417 | |
| 418 const PAGECHAR_INFO& charinfo = m_CharList[index]; | |
| 419 info->m_Charcode = charinfo.m_CharCode; | |
| 420 info->m_OriginX = charinfo.m_OriginX; | |
| 421 info->m_OriginY = charinfo.m_OriginY; | |
| 422 info->m_Unicode = charinfo.m_Unicode; | |
| 423 info->m_Flag = charinfo.m_Flag; | |
| 424 info->m_CharBox = charinfo.m_CharBox; | |
| 425 info->m_pTextObj = charinfo.m_pTextObj; | |
| 426 if (charinfo.m_pTextObj && charinfo.m_pTextObj->GetFont()) | |
| 427 info->m_FontSize = charinfo.m_pTextObj->GetFontSize(); | |
| 428 else | |
| 429 info->m_FontSize = kDefaultFontSize; | |
| 430 info->m_Matrix.Copy(charinfo.m_Matrix); | |
| 431 } | |
| 432 | |
| 433 void CPDF_TextPage::CheckMarkedContentObject(int32_t& start, | |
| 434 int32_t& nCount) const { | |
| 435 PAGECHAR_INFO charinfo = m_CharList[start]; | |
| 436 PAGECHAR_INFO charinfo2 = m_CharList[start + nCount - 1]; | |
| 437 if (FPDFTEXT_CHAR_PIECE != charinfo.m_Flag && | |
| 438 FPDFTEXT_CHAR_PIECE != charinfo2.m_Flag) { | |
| 439 return; | |
| 440 } | |
| 441 if (FPDFTEXT_CHAR_PIECE == charinfo.m_Flag) { | |
| 442 PAGECHAR_INFO charinfo1 = charinfo; | |
| 443 int startIndex = start; | |
| 444 while (FPDFTEXT_CHAR_PIECE == charinfo1.m_Flag && | |
| 445 charinfo1.m_Index == charinfo.m_Index) { | |
| 446 startIndex--; | |
| 447 if (startIndex < 0) | |
| 448 break; | |
| 449 charinfo1 = m_CharList[startIndex]; | |
| 450 } | |
| 451 startIndex++; | |
| 452 start = startIndex; | |
| 453 } | |
| 454 if (FPDFTEXT_CHAR_PIECE == charinfo2.m_Flag) { | |
| 455 PAGECHAR_INFO charinfo3 = charinfo2; | |
| 456 int endIndex = start + nCount - 1; | |
| 457 while (FPDFTEXT_CHAR_PIECE == charinfo3.m_Flag && | |
| 458 charinfo3.m_Index == charinfo2.m_Index) { | |
| 459 endIndex++; | |
| 460 if (endIndex >= pdfium::CollectionSize<int>(m_CharList)) | |
| 461 break; | |
| 462 charinfo3 = m_CharList[endIndex]; | |
| 463 } | |
| 464 endIndex--; | |
| 465 nCount = endIndex - start + 1; | |
| 466 } | |
| 467 } | |
| 468 | |
| 469 CFX_WideString CPDF_TextPage::GetPageText(int start, int nCount) const { | |
| 470 if (!m_bIsParsed || nCount == 0) | |
| 471 return L""; | |
| 472 | |
| 473 if (start < 0) | |
| 474 start = 0; | |
| 475 | |
| 476 if (nCount == -1) { | |
| 477 nCount = pdfium::CollectionSize<int>(m_CharList) - start; | |
| 478 return CFX_WideString( | |
| 479 m_TextBuf.AsStringC().Mid(start, m_TextBuf.AsStringC().GetLength())); | |
| 480 } | |
| 481 if (nCount <= 0 || m_CharList.empty()) | |
| 482 return L""; | |
| 483 if (nCount + start > pdfium::CollectionSize<int>(m_CharList) - 1) | |
| 484 nCount = pdfium::CollectionSize<int>(m_CharList) - start; | |
| 485 if (nCount <= 0) | |
| 486 return L""; | |
| 487 CheckMarkedContentObject(start, nCount); | |
| 488 int startindex = 0; | |
| 489 PAGECHAR_INFO charinfo = m_CharList[start]; | |
| 490 int startOffset = 0; | |
| 491 while (charinfo.m_Index == -1) { | |
| 492 startOffset++; | |
| 493 if (startOffset > nCount || | |
| 494 start + startOffset >= pdfium::CollectionSize<int>(m_CharList)) { | |
| 495 return L""; | |
| 496 } | |
| 497 charinfo = m_CharList[start + startOffset]; | |
| 498 } | |
| 499 startindex = charinfo.m_Index; | |
| 500 charinfo = m_CharList[start + nCount - 1]; | |
| 501 int nCountOffset = 0; | |
| 502 while (charinfo.m_Index == -1) { | |
| 503 nCountOffset++; | |
| 504 if (nCountOffset >= nCount) | |
| 505 return L""; | |
| 506 charinfo = m_CharList[start + nCount - nCountOffset - 1]; | |
| 507 } | |
| 508 nCount = start + nCount - nCountOffset - startindex; | |
| 509 if (nCount <= 0) | |
| 510 return L""; | |
| 511 return CFX_WideString(m_TextBuf.AsStringC().Mid(startindex, nCount)); | |
| 512 } | |
| 513 | |
| 514 int CPDF_TextPage::CountRects(int start, int nCount) { | |
| 515 if (!m_bIsParsed || start < 0) | |
| 516 return -1; | |
| 517 | |
| 518 if (nCount == -1 || | |
| 519 nCount + start > pdfium::CollectionSize<int>(m_CharList)) { | |
| 520 nCount = pdfium::CollectionSize<int>(m_CharList) - start; | |
| 521 } | |
| 522 m_SelRects = GetRectArray(start, nCount); | |
| 523 return pdfium::CollectionSize<int>(m_SelRects); | |
| 524 } | |
| 525 | |
| 526 void CPDF_TextPage::GetRect(int rectIndex, | |
| 527 FX_FLOAT& left, | |
| 528 FX_FLOAT& top, | |
| 529 FX_FLOAT& right, | |
| 530 FX_FLOAT& bottom) const { | |
| 531 if (!m_bIsParsed) | |
| 532 return; | |
| 533 | |
| 534 if (rectIndex < 0 || rectIndex >= pdfium::CollectionSize<int>(m_SelRects)) | |
| 535 return; | |
| 536 | |
| 537 left = m_SelRects[rectIndex].left; | |
| 538 top = m_SelRects[rectIndex].top; | |
| 539 right = m_SelRects[rectIndex].right; | |
| 540 bottom = m_SelRects[rectIndex].bottom; | |
| 541 } | |
| 542 | |
| 543 CPDF_TextPage::TextOrientation CPDF_TextPage::FindTextlineFlowOrientation() | |
| 544 const { | |
| 545 if (m_pPage->GetPageObjectList()->empty()) | |
| 546 return TextOrientation::Unknown; | |
| 547 | |
| 548 const int32_t nPageWidth = static_cast<int32_t>(m_pPage->GetPageWidth()); | |
| 549 const int32_t nPageHeight = static_cast<int32_t>(m_pPage->GetPageHeight()); | |
| 550 std::vector<bool> nHorizontalMask(nPageWidth); | |
| 551 std::vector<bool> nVerticalMask(nPageHeight); | |
| 552 FX_FLOAT fLineHeight = 0.0f; | |
| 553 int32_t nStartH = nPageWidth; | |
| 554 int32_t nEndH = 0; | |
| 555 int32_t nStartV = nPageHeight; | |
| 556 int32_t nEndV = 0; | |
| 557 for (const auto& pPageObj : *m_pPage->GetPageObjectList()) { | |
| 558 if (!pPageObj->IsText()) | |
| 559 continue; | |
| 560 | |
| 561 int32_t minH = std::max(static_cast<int32_t>(pPageObj->m_Left), 0); | |
| 562 int32_t maxH = | |
| 563 std::min(static_cast<int32_t>(pPageObj->m_Right), nPageWidth); | |
| 564 int32_t minV = std::max(static_cast<int32_t>(pPageObj->m_Bottom), 0); | |
| 565 int32_t maxV = std::min(static_cast<int32_t>(pPageObj->m_Top), nPageHeight); | |
| 566 if (minH >= maxH || minV >= maxV) | |
| 567 continue; | |
| 568 | |
| 569 for (int32_t i = minH; i < maxH; ++i) | |
| 570 nHorizontalMask[i] = true; | |
| 571 for (int32_t i = minV; i < maxV; ++i) | |
| 572 nVerticalMask[i] = true; | |
| 573 | |
| 574 nStartH = std::min(nStartH, minH); | |
| 575 nEndH = std::max(nEndH, maxH); | |
| 576 nStartV = std::min(nStartV, minV); | |
| 577 nEndV = std::max(nEndV, maxV); | |
| 578 | |
| 579 if (fLineHeight <= 0.0f) | |
| 580 fLineHeight = pPageObj->m_Top - pPageObj->m_Bottom; | |
| 581 } | |
| 582 const int32_t nDoubleLineHeight = 2 * fLineHeight; | |
| 583 if ((nEndV - nStartV) < nDoubleLineHeight) | |
| 584 return TextOrientation::Horizontal; | |
| 585 if ((nEndH - nStartH) < nDoubleLineHeight) | |
| 586 return TextOrientation::Vertical; | |
| 587 | |
| 588 const FX_FLOAT nSumH = MaskPercentFilled(nHorizontalMask, nStartH, nEndH); | |
| 589 if (nSumH > 0.8f) | |
| 590 return TextOrientation::Horizontal; | |
| 591 | |
| 592 const FX_FLOAT nSumV = MaskPercentFilled(nVerticalMask, nStartV, nEndV); | |
| 593 if (nSumH > nSumV) | |
| 594 return TextOrientation::Horizontal; | |
| 595 if (nSumH < nSumV) | |
| 596 return TextOrientation::Vertical; | |
| 597 return TextOrientation::Unknown; | |
| 598 } | |
| 599 | |
| 600 void CPDF_TextPage::AppendGeneratedCharacter(FX_WCHAR unicode, | |
| 601 const CFX_Matrix& formMatrix) { | |
| 602 PAGECHAR_INFO generateChar; | |
| 603 if (!GenerateCharInfo(unicode, generateChar)) | |
| 604 return; | |
| 605 | |
| 606 m_TextBuf.AppendChar(unicode); | |
| 607 if (!formMatrix.IsIdentity()) | |
| 608 generateChar.m_Matrix.Copy(formMatrix); | |
| 609 m_CharList.push_back(generateChar); | |
| 610 } | |
| 611 | |
| 612 void CPDF_TextPage::ProcessObject() { | |
| 613 if (m_pPage->GetPageObjectList()->empty()) | |
| 614 return; | |
| 615 | |
| 616 m_TextlineDir = FindTextlineFlowOrientation(); | |
| 617 const CPDF_PageObjectList* pObjList = m_pPage->GetPageObjectList(); | |
| 618 for (auto it = pObjList->begin(); it != pObjList->end(); ++it) { | |
| 619 if (CPDF_PageObject* pObj = it->get()) { | |
| 620 if (pObj->IsText()) { | |
| 621 CFX_Matrix matrix; | |
| 622 ProcessTextObject(pObj->AsText(), matrix, pObjList, it); | |
| 623 } else if (pObj->IsForm()) { | |
| 624 CFX_Matrix formMatrix(1, 0, 0, 1, 0, 0); | |
| 625 ProcessFormObject(pObj->AsForm(), formMatrix); | |
| 626 } | |
| 627 } | |
| 628 } | |
| 629 for (int i = 0; i < m_LineObj.GetSize(); i++) | |
| 630 ProcessTextObject(m_LineObj.GetAt(i)); | |
| 631 | |
| 632 m_LineObj.RemoveAll(); | |
| 633 CloseTempLine(); | |
| 634 } | |
| 635 | |
| 636 void CPDF_TextPage::ProcessFormObject(CPDF_FormObject* pFormObj, | |
| 637 const CFX_Matrix& formMatrix) { | |
| 638 CPDF_PageObjectList* pObjectList = pFormObj->m_pForm->GetPageObjectList(); | |
| 639 if (pObjectList->empty()) | |
| 640 return; | |
| 641 | |
| 642 CFX_Matrix curFormMatrix; | |
| 643 curFormMatrix.Copy(pFormObj->m_FormMatrix); | |
| 644 curFormMatrix.Concat(formMatrix); | |
| 645 | |
| 646 for (auto it = pObjectList->begin(); it != pObjectList->end(); ++it) { | |
| 647 if (CPDF_PageObject* pPageObj = it->get()) { | |
| 648 if (pPageObj->IsText()) | |
| 649 ProcessTextObject(pPageObj->AsText(), curFormMatrix, pObjectList, it); | |
| 650 else if (pPageObj->IsForm()) | |
| 651 ProcessFormObject(pPageObj->AsForm(), curFormMatrix); | |
| 652 } | |
| 653 } | |
| 654 } | |
| 655 | |
| 656 int CPDF_TextPage::GetCharWidth(uint32_t charCode, CPDF_Font* pFont) const { | |
| 657 if (charCode == CPDF_Font::kInvalidCharCode) | |
| 658 return 0; | |
| 659 | |
| 660 if (int w = pFont->GetCharWidthF(charCode)) | |
| 661 return w; | |
| 662 | |
| 663 CFX_ByteString str; | |
| 664 pFont->AppendChar(str, charCode); | |
| 665 if (int w = pFont->GetStringWidth(str.c_str(), 1)) | |
| 666 return w; | |
| 667 | |
| 668 return pFont->GetCharBBox(charCode).Width(); | |
| 669 } | |
| 670 | |
| 671 void CPDF_TextPage::AddCharInfoByLRDirection(FX_WCHAR wChar, | |
| 672 PAGECHAR_INFO info) { | |
| 673 if (IsControlChar(info)) { | |
| 674 info.m_Index = -1; | |
| 675 m_CharList.push_back(info); | |
| 676 return; | |
| 677 } | |
| 678 | |
| 679 info.m_Index = m_TextBuf.GetLength(); | |
| 680 if (wChar >= 0xFB00 && wChar <= 0xFB06) { | |
| 681 FX_WCHAR* pDst = nullptr; | |
| 682 FX_STRSIZE nCount = Unicode_GetNormalization(wChar, pDst); | |
| 683 if (nCount >= 1) { | |
| 684 pDst = FX_Alloc(FX_WCHAR, nCount); | |
| 685 Unicode_GetNormalization(wChar, pDst); | |
| 686 for (int nIndex = 0; nIndex < nCount; nIndex++) { | |
| 687 PAGECHAR_INFO info2 = info; | |
| 688 info2.m_Unicode = pDst[nIndex]; | |
| 689 info2.m_Flag = FPDFTEXT_CHAR_PIECE; | |
| 690 m_TextBuf.AppendChar(info2.m_Unicode); | |
| 691 m_CharList.push_back(info2); | |
| 692 } | |
| 693 FX_Free(pDst); | |
| 694 return; | |
| 695 } | |
| 696 } | |
| 697 m_TextBuf.AppendChar(wChar); | |
| 698 m_CharList.push_back(info); | |
| 699 } | |
| 700 | |
| 701 void CPDF_TextPage::AddCharInfoByRLDirection(FX_WCHAR wChar, | |
| 702 PAGECHAR_INFO info) { | |
| 703 if (IsControlChar(info)) { | |
| 704 info.m_Index = -1; | |
| 705 m_CharList.push_back(info); | |
| 706 return; | |
| 707 } | |
| 708 | |
| 709 info.m_Index = m_TextBuf.GetLength(); | |
| 710 wChar = FX_GetMirrorChar(wChar, TRUE, FALSE); | |
| 711 FX_WCHAR* pDst = nullptr; | |
| 712 FX_STRSIZE nCount = Unicode_GetNormalization(wChar, pDst); | |
| 713 if (nCount >= 1) { | |
| 714 pDst = FX_Alloc(FX_WCHAR, nCount); | |
| 715 Unicode_GetNormalization(wChar, pDst); | |
| 716 for (int nIndex = 0; nIndex < nCount; nIndex++) { | |
| 717 PAGECHAR_INFO info2 = info; | |
| 718 info2.m_Unicode = pDst[nIndex]; | |
| 719 info2.m_Flag = FPDFTEXT_CHAR_PIECE; | |
| 720 m_TextBuf.AppendChar(info2.m_Unicode); | |
| 721 m_CharList.push_back(info2); | |
| 722 } | |
| 723 FX_Free(pDst); | |
| 724 return; | |
| 725 } | |
| 726 info.m_Unicode = wChar; | |
| 727 m_TextBuf.AppendChar(info.m_Unicode); | |
| 728 m_CharList.push_back(info); | |
| 729 } | |
| 730 | |
| 731 void CPDF_TextPage::CloseTempLine() { | |
| 732 if (m_TempCharList.empty()) | |
| 733 return; | |
| 734 | |
| 735 CFX_WideString str = m_TempTextBuf.MakeString(); | |
| 736 bool bPrevSpace = false; | |
| 737 for (int i = 0; i < str.GetLength(); i++) { | |
| 738 if (str.GetAt(i) != ' ') { | |
| 739 bPrevSpace = false; | |
| 740 continue; | |
| 741 } | |
| 742 if (bPrevSpace) { | |
| 743 m_TempTextBuf.Delete(i, 1); | |
| 744 m_TempCharList.erase(m_TempCharList.begin() + i); | |
| 745 str.Delete(i); | |
| 746 i--; | |
| 747 } | |
| 748 bPrevSpace = true; | |
| 749 } | |
| 750 CFX_BidiString bidi(str); | |
| 751 if (m_parserflag == FPDFText_Direction::Right) | |
| 752 bidi.SetOverallDirectionRight(); | |
| 753 CFX_BidiChar::Direction eCurrentDirection = bidi.OverallDirection(); | |
| 754 for (const auto& segment : bidi) { | |
| 755 if (segment.direction == CFX_BidiChar::RIGHT || | |
| 756 (segment.direction == CFX_BidiChar::NEUTRAL && | |
| 757 eCurrentDirection == CFX_BidiChar::RIGHT)) { | |
| 758 eCurrentDirection = CFX_BidiChar::RIGHT; | |
| 759 for (int m = segment.start + segment.count; m > segment.start; --m) | |
| 760 AddCharInfoByRLDirection(bidi.CharAt(m - 1), m_TempCharList[m - 1]); | |
| 761 } else { | |
| 762 eCurrentDirection = CFX_BidiChar::LEFT; | |
| 763 for (int m = segment.start; m < segment.start + segment.count; m++) | |
| 764 AddCharInfoByLRDirection(bidi.CharAt(m), m_TempCharList[m]); | |
| 765 } | |
| 766 } | |
| 767 m_TempCharList.clear(); | |
| 768 m_TempTextBuf.Delete(0, m_TempTextBuf.GetLength()); | |
| 769 } | |
| 770 | |
| 771 void CPDF_TextPage::ProcessTextObject( | |
| 772 CPDF_TextObject* pTextObj, | |
| 773 const CFX_Matrix& formMatrix, | |
| 774 const CPDF_PageObjectList* pObjList, | |
| 775 CPDF_PageObjectList::const_iterator ObjPos) { | |
| 776 CFX_FloatRect re(pTextObj->m_Left, pTextObj->m_Bottom, pTextObj->m_Right, | |
| 777 pTextObj->m_Top); | |
| 778 if (FXSYS_fabs(pTextObj->m_Right - pTextObj->m_Left) < 0.01f) | |
| 779 return; | |
| 780 int count = m_LineObj.GetSize(); | |
| 781 PDFTEXT_Obj Obj; | |
| 782 Obj.m_pTextObj = pTextObj; | |
| 783 Obj.m_formMatrix = formMatrix; | |
| 784 if (count == 0) { | |
| 785 m_LineObj.Add(Obj); | |
| 786 return; | |
| 787 } | |
| 788 if (IsSameAsPreTextObject(pTextObj, pObjList, ObjPos)) | |
| 789 return; | |
| 790 PDFTEXT_Obj prev_Obj = m_LineObj.GetAt(count - 1); | |
| 791 CPDF_TextObjectItem item; | |
| 792 int nItem = prev_Obj.m_pTextObj->CountItems(); | |
| 793 prev_Obj.m_pTextObj->GetItemInfo(nItem - 1, &item); | |
| 794 FX_FLOAT prev_width = | |
| 795 GetCharWidth(item.m_CharCode, prev_Obj.m_pTextObj->GetFont()) * | |
| 796 prev_Obj.m_pTextObj->GetFontSize() / 1000; | |
| 797 CFX_Matrix prev_matrix; | |
| 798 prev_Obj.m_pTextObj->GetTextMatrix(&prev_matrix); | |
| 799 prev_width = FXSYS_fabs(prev_width); | |
| 800 prev_matrix.Concat(prev_Obj.m_formMatrix); | |
| 801 prev_width = prev_matrix.TransformDistance(prev_width); | |
| 802 pTextObj->GetItemInfo(0, &item); | |
| 803 FX_FLOAT this_width = GetCharWidth(item.m_CharCode, pTextObj->GetFont()) * | |
| 804 pTextObj->GetFontSize() / 1000; | |
| 805 this_width = FXSYS_fabs(this_width); | |
| 806 CFX_Matrix this_matrix; | |
| 807 pTextObj->GetTextMatrix(&this_matrix); | |
| 808 this_width = FXSYS_fabs(this_width); | |
| 809 this_matrix.Concat(formMatrix); | |
| 810 this_width = this_matrix.TransformDistance(this_width); | |
| 811 FX_FLOAT threshold = | |
| 812 prev_width > this_width ? prev_width / 4 : this_width / 4; | |
| 813 FX_FLOAT prev_x = prev_Obj.m_pTextObj->GetPosX(), | |
| 814 prev_y = prev_Obj.m_pTextObj->GetPosY(); | |
| 815 prev_Obj.m_formMatrix.Transform(prev_x, prev_y); | |
| 816 m_DisplayMatrix.Transform(prev_x, prev_y); | |
| 817 FX_FLOAT this_x = pTextObj->GetPosX(), this_y = pTextObj->GetPosY(); | |
| 818 formMatrix.Transform(this_x, this_y); | |
| 819 m_DisplayMatrix.Transform(this_x, this_y); | |
| 820 if (FXSYS_fabs(this_y - prev_y) > threshold * 2) { | |
| 821 for (int i = 0; i < count; i++) | |
| 822 ProcessTextObject(m_LineObj.GetAt(i)); | |
| 823 m_LineObj.RemoveAll(); | |
| 824 m_LineObj.Add(Obj); | |
| 825 return; | |
| 826 } | |
| 827 int i = 0; | |
| 828 for (i = count - 1; i >= 0; i--) { | |
| 829 PDFTEXT_Obj prev_text_obj = m_LineObj.GetAt(i); | |
| 830 FX_FLOAT Prev_x = prev_text_obj.m_pTextObj->GetPosX(), | |
| 831 Prev_y = prev_text_obj.m_pTextObj->GetPosY(); | |
| 832 prev_text_obj.m_formMatrix.Transform(Prev_x, Prev_y); | |
| 833 m_DisplayMatrix.Transform(Prev_x, Prev_y); | |
| 834 if (this_x >= Prev_x) { | |
| 835 if (i == count - 1) | |
| 836 m_LineObj.Add(Obj); | |
| 837 else | |
| 838 m_LineObj.InsertAt(i + 1, Obj); | |
| 839 break; | |
| 840 } | |
| 841 } | |
| 842 if (i < 0) | |
| 843 m_LineObj.InsertAt(0, Obj); | |
| 844 } | |
| 845 | |
| 846 FPDFText_MarkedContent CPDF_TextPage::PreMarkedContent(PDFTEXT_Obj Obj) { | |
| 847 CPDF_TextObject* pTextObj = Obj.m_pTextObj; | |
| 848 const CPDF_ContentMarkData* pMarkData = pTextObj->m_ContentMark.GetObject(); | |
| 849 if (!pMarkData) | |
| 850 return FPDFText_MarkedContent::Pass; | |
| 851 | |
| 852 int nContentMark = pMarkData->CountItems(); | |
| 853 if (nContentMark < 1) | |
| 854 return FPDFText_MarkedContent::Pass; | |
| 855 | |
| 856 CFX_WideString actText; | |
| 857 bool bExist = false; | |
| 858 CPDF_Dictionary* pDict = nullptr; | |
| 859 int n = 0; | |
| 860 for (n = 0; n < nContentMark; n++) { | |
| 861 const CPDF_ContentMarkItem& item = pMarkData->GetItem(n); | |
| 862 if (item.GetParamType() == CPDF_ContentMarkItem::ParamType::None) | |
| 863 continue; | |
| 864 pDict = item.GetParam(); | |
| 865 CPDF_String* temp = | |
| 866 ToString(pDict ? pDict->GetObjectBy("ActualText") : nullptr); | |
| 867 if (temp) { | |
| 868 bExist = true; | |
| 869 actText = temp->GetUnicodeText(); | |
| 870 } | |
| 871 } | |
| 872 if (!bExist) | |
| 873 return FPDFText_MarkedContent::Pass; | |
| 874 | |
| 875 if (m_pPreTextObj) { | |
| 876 const CPDF_ContentMarkData* pPreMarkData = | |
| 877 m_pPreTextObj->m_ContentMark.GetObject(); | |
| 878 if (pPreMarkData && pPreMarkData->CountItems() == n && | |
| 879 pDict == pPreMarkData->GetItem(n - 1).GetParam()) { | |
| 880 return FPDFText_MarkedContent::Done; | |
| 881 } | |
| 882 } | |
| 883 FX_STRSIZE nItems = actText.GetLength(); | |
| 884 if (nItems < 1) | |
| 885 return FPDFText_MarkedContent::Pass; | |
| 886 | |
| 887 CPDF_Font* pFont = pTextObj->GetFont(); | |
| 888 bExist = false; | |
| 889 for (FX_STRSIZE i = 0; i < nItems; i++) { | |
| 890 if (pFont->CharCodeFromUnicode(actText.GetAt(i)) != | |
| 891 CPDF_Font::kInvalidCharCode) { | |
| 892 bExist = true; | |
| 893 break; | |
| 894 } | |
| 895 } | |
| 896 if (!bExist) | |
| 897 return FPDFText_MarkedContent::Pass; | |
| 898 | |
| 899 bExist = false; | |
| 900 for (FX_STRSIZE i = 0; i < nItems; i++) { | |
| 901 FX_WCHAR wChar = actText.GetAt(i); | |
| 902 if ((wChar > 0x80 && wChar < 0xFFFD) || (wChar <= 0x80 && isprint(wChar))) { | |
| 903 bExist = true; | |
| 904 break; | |
| 905 } | |
| 906 } | |
| 907 if (!bExist) | |
| 908 return FPDFText_MarkedContent::Done; | |
| 909 | |
| 910 return FPDFText_MarkedContent::Delay; | |
| 911 } | |
| 912 | |
| 913 void CPDF_TextPage::ProcessMarkedContent(PDFTEXT_Obj Obj) { | |
| 914 CPDF_TextObject* pTextObj = Obj.m_pTextObj; | |
| 915 const CPDF_ContentMarkData* pMarkData = pTextObj->m_ContentMark.GetObject(); | |
| 916 if (!pMarkData) | |
| 917 return; | |
| 918 | |
| 919 int nContentMark = pMarkData->CountItems(); | |
| 920 if (nContentMark < 1) | |
| 921 return; | |
| 922 CFX_WideString actText; | |
| 923 CPDF_Dictionary* pDict = nullptr; | |
| 924 for (int n = 0; n < nContentMark; n++) { | |
| 925 const CPDF_ContentMarkItem& item = pMarkData->GetItem(n); | |
| 926 if (item.GetParamType() == CPDF_ContentMarkItem::ParamType::None) | |
| 927 continue; | |
| 928 pDict = item.GetParam(); | |
| 929 if (pDict) | |
| 930 actText = pDict->GetUnicodeTextBy("ActualText"); | |
| 931 } | |
| 932 FX_STRSIZE nItems = actText.GetLength(); | |
| 933 if (nItems < 1) | |
| 934 return; | |
| 935 | |
| 936 CPDF_Font* pFont = pTextObj->GetFont(); | |
| 937 CFX_Matrix formMatrix = Obj.m_formMatrix; | |
| 938 CFX_Matrix matrix; | |
| 939 pTextObj->GetTextMatrix(&matrix); | |
| 940 matrix.Concat(formMatrix); | |
| 941 FX_FLOAT fPosX = pTextObj->GetPosX(); | |
| 942 FX_FLOAT fPosY = pTextObj->GetPosY(); | |
| 943 int nCharInfoIndex = m_TextBuf.GetLength(); | |
| 944 CFX_FloatRect charBox; | |
| 945 charBox.top = pTextObj->m_Top; | |
| 946 charBox.left = pTextObj->m_Left; | |
| 947 charBox.right = pTextObj->m_Right; | |
| 948 charBox.bottom = pTextObj->m_Bottom; | |
| 949 for (FX_STRSIZE k = 0; k < nItems; k++) { | |
| 950 FX_WCHAR wChar = actText.GetAt(k); | |
| 951 if (wChar <= 0x80 && !isprint(wChar)) | |
| 952 wChar = 0x20; | |
| 953 if (wChar >= 0xFFFD) | |
| 954 continue; | |
| 955 PAGECHAR_INFO charinfo; | |
| 956 charinfo.m_OriginX = fPosX; | |
| 957 charinfo.m_OriginY = fPosY; | |
| 958 charinfo.m_Index = nCharInfoIndex; | |
| 959 charinfo.m_Unicode = wChar; | |
| 960 charinfo.m_CharCode = pFont->CharCodeFromUnicode(wChar); | |
| 961 charinfo.m_Flag = FPDFTEXT_CHAR_PIECE; | |
| 962 charinfo.m_pTextObj = pTextObj; | |
| 963 charinfo.m_CharBox.top = charBox.top; | |
| 964 charinfo.m_CharBox.left = charBox.left; | |
| 965 charinfo.m_CharBox.right = charBox.right; | |
| 966 charinfo.m_CharBox.bottom = charBox.bottom; | |
| 967 charinfo.m_Matrix.Copy(matrix); | |
| 968 m_TempTextBuf.AppendChar(wChar); | |
| 969 m_TempCharList.push_back(charinfo); | |
| 970 } | |
| 971 } | |
| 972 | |
| 973 void CPDF_TextPage::FindPreviousTextObject() { | |
| 974 if (m_TempCharList.empty() && m_CharList.empty()) | |
| 975 return; | |
| 976 | |
| 977 PAGECHAR_INFO preChar = | |
| 978 m_TempCharList.empty() ? m_CharList.back() : m_TempCharList.back(); | |
| 979 | |
| 980 if (preChar.m_pTextObj) | |
| 981 m_pPreTextObj = preChar.m_pTextObj; | |
| 982 } | |
| 983 | |
| 984 void CPDF_TextPage::SwapTempTextBuf(int32_t iCharListStartAppend, | |
| 985 int32_t iBufStartAppend) { | |
| 986 int32_t i = iCharListStartAppend; | |
| 987 int32_t j = pdfium::CollectionSize<int32_t>(m_TempCharList) - 1; | |
| 988 for (; i < j; i++, j--) { | |
| 989 std::swap(m_TempCharList[i], m_TempCharList[j]); | |
| 990 std::swap(m_TempCharList[i].m_Index, m_TempCharList[j].m_Index); | |
| 991 } | |
| 992 FX_WCHAR* pTempBuffer = m_TempTextBuf.GetBuffer(); | |
| 993 i = iBufStartAppend; | |
| 994 j = m_TempTextBuf.GetLength() - 1; | |
| 995 for (; i < j; i++, j--) | |
| 996 std::swap(pTempBuffer[i], pTempBuffer[j]); | |
| 997 } | |
| 998 | |
| 999 FX_BOOL CPDF_TextPage::IsRightToLeft(const CPDF_TextObject* pTextObj, | |
| 1000 const CPDF_Font* pFont, | |
| 1001 int nItems) const { | |
| 1002 CFX_WideString str; | |
| 1003 for (int32_t i = 0; i < nItems; i++) { | |
| 1004 CPDF_TextObjectItem item; | |
| 1005 pTextObj->GetItemInfo(i, &item); | |
| 1006 if (item.m_CharCode == static_cast<uint32_t>(-1)) | |
| 1007 continue; | |
| 1008 CFX_WideString wstrItem = pFont->UnicodeFromCharCode(item.m_CharCode); | |
| 1009 FX_WCHAR wChar = wstrItem.GetAt(0); | |
| 1010 if ((wstrItem.IsEmpty() || wChar == 0) && item.m_CharCode) | |
| 1011 wChar = (FX_WCHAR)item.m_CharCode; | |
| 1012 if (wChar) | |
| 1013 str += wChar; | |
| 1014 } | |
| 1015 return CFX_BidiString(str).OverallDirection() == CFX_BidiChar::RIGHT; | |
| 1016 } | |
| 1017 | |
| 1018 void CPDF_TextPage::ProcessTextObject(PDFTEXT_Obj Obj) { | |
| 1019 CPDF_TextObject* pTextObj = Obj.m_pTextObj; | |
| 1020 if (FXSYS_fabs(pTextObj->m_Right - pTextObj->m_Left) < 0.01f) | |
| 1021 return; | |
| 1022 CFX_Matrix formMatrix = Obj.m_formMatrix; | |
| 1023 CPDF_Font* pFont = pTextObj->GetFont(); | |
| 1024 CFX_Matrix matrix; | |
| 1025 pTextObj->GetTextMatrix(&matrix); | |
| 1026 matrix.Concat(formMatrix); | |
| 1027 FPDFText_MarkedContent ePreMKC = PreMarkedContent(Obj); | |
| 1028 if (ePreMKC == FPDFText_MarkedContent::Done) { | |
| 1029 m_pPreTextObj = pTextObj; | |
| 1030 m_perMatrix.Copy(formMatrix); | |
| 1031 return; | |
| 1032 } | |
| 1033 GenerateCharacter result = GenerateCharacter::None; | |
| 1034 if (m_pPreTextObj) { | |
| 1035 result = ProcessInsertObject(pTextObj, formMatrix); | |
| 1036 if (result == GenerateCharacter::LineBreak) { | |
| 1037 m_CurlineRect = | |
| 1038 CFX_FloatRect(Obj.m_pTextObj->m_Left, Obj.m_pTextObj->m_Bottom, | |
| 1039 Obj.m_pTextObj->m_Right, Obj.m_pTextObj->m_Top); | |
| 1040 } else { | |
| 1041 m_CurlineRect.Union( | |
| 1042 CFX_FloatRect(Obj.m_pTextObj->m_Left, Obj.m_pTextObj->m_Bottom, | |
| 1043 Obj.m_pTextObj->m_Right, Obj.m_pTextObj->m_Top)); | |
| 1044 } | |
| 1045 switch (result) { | |
| 1046 case GenerateCharacter::None: | |
| 1047 break; | |
| 1048 case GenerateCharacter::Space: { | |
| 1049 PAGECHAR_INFO generateChar; | |
| 1050 if (GenerateCharInfo(TEXT_SPACE_CHAR, generateChar)) { | |
| 1051 if (!formMatrix.IsIdentity()) | |
| 1052 generateChar.m_Matrix.Copy(formMatrix); | |
| 1053 m_TempTextBuf.AppendChar(TEXT_SPACE_CHAR); | |
| 1054 m_TempCharList.push_back(generateChar); | |
| 1055 } | |
| 1056 break; | |
| 1057 } | |
| 1058 case GenerateCharacter::LineBreak: | |
| 1059 CloseTempLine(); | |
| 1060 if (m_TextBuf.GetSize()) { | |
| 1061 AppendGeneratedCharacter(TEXT_RETURN_CHAR, formMatrix); | |
| 1062 AppendGeneratedCharacter(TEXT_LINEFEED_CHAR, formMatrix); | |
| 1063 } | |
| 1064 break; | |
| 1065 case GenerateCharacter::Hyphen: | |
| 1066 if (pTextObj->CountChars() == 1) { | |
| 1067 CPDF_TextObjectItem item; | |
| 1068 pTextObj->GetCharInfo(0, &item); | |
| 1069 CFX_WideString wstrItem = | |
| 1070 pTextObj->GetFont()->UnicodeFromCharCode(item.m_CharCode); | |
| 1071 if (wstrItem.IsEmpty()) | |
| 1072 wstrItem += (FX_WCHAR)item.m_CharCode; | |
| 1073 FX_WCHAR curChar = wstrItem.GetAt(0); | |
| 1074 if (curChar == 0x2D || curChar == 0xAD) | |
| 1075 return; | |
| 1076 } | |
| 1077 while (m_TempTextBuf.GetSize() > 0 && | |
| 1078 m_TempTextBuf.AsStringC().GetAt(m_TempTextBuf.GetLength() - 1) == | |
| 1079 0x20) { | |
| 1080 m_TempTextBuf.Delete(m_TempTextBuf.GetLength() - 1, 1); | |
| 1081 m_TempCharList.pop_back(); | |
| 1082 } | |
| 1083 PAGECHAR_INFO* charinfo = &m_TempCharList.back(); | |
| 1084 m_TempTextBuf.Delete(m_TempTextBuf.GetLength() - 1, 1); | |
| 1085 charinfo->m_Unicode = 0x2; | |
| 1086 charinfo->m_Flag = FPDFTEXT_CHAR_HYPHEN; | |
| 1087 m_TempTextBuf.AppendChar(0xfffe); | |
| 1088 break; | |
| 1089 } | |
| 1090 } else { | |
| 1091 m_CurlineRect = | |
| 1092 CFX_FloatRect(Obj.m_pTextObj->m_Left, Obj.m_pTextObj->m_Bottom, | |
| 1093 Obj.m_pTextObj->m_Right, Obj.m_pTextObj->m_Top); | |
| 1094 } | |
| 1095 if (ePreMKC == FPDFText_MarkedContent::Delay) { | |
| 1096 ProcessMarkedContent(Obj); | |
| 1097 m_pPreTextObj = pTextObj; | |
| 1098 m_perMatrix.Copy(formMatrix); | |
| 1099 return; | |
| 1100 } | |
| 1101 m_pPreTextObj = pTextObj; | |
| 1102 m_perMatrix.Copy(formMatrix); | |
| 1103 int nItems = pTextObj->CountItems(); | |
| 1104 FX_FLOAT baseSpace = CalculateBaseSpace(pTextObj, matrix); | |
| 1105 | |
| 1106 const FX_BOOL bR2L = IsRightToLeft(pTextObj, pFont, nItems); | |
| 1107 const FX_BOOL bIsBidiAndMirrorInverse = | |
| 1108 bR2L && (matrix.a * matrix.d - matrix.b * matrix.c) < 0; | |
| 1109 int32_t iBufStartAppend = m_TempTextBuf.GetLength(); | |
| 1110 int32_t iCharListStartAppend = | |
| 1111 pdfium::CollectionSize<int32_t>(m_TempCharList); | |
| 1112 | |
| 1113 FX_FLOAT spacing = 0; | |
| 1114 for (int i = 0; i < nItems; i++) { | |
| 1115 CPDF_TextObjectItem item; | |
| 1116 PAGECHAR_INFO charinfo; | |
| 1117 charinfo.m_OriginX = 0; | |
| 1118 charinfo.m_OriginY = 0; | |
| 1119 pTextObj->GetItemInfo(i, &item); | |
| 1120 if (item.m_CharCode == static_cast<uint32_t>(-1)) { | |
| 1121 CFX_WideString str = m_TempTextBuf.MakeString(); | |
| 1122 if (str.IsEmpty()) | |
| 1123 str = m_TextBuf.AsStringC(); | |
| 1124 if (str.IsEmpty() || str.GetAt(str.GetLength() - 1) == TEXT_SPACE_CHAR) | |
| 1125 continue; | |
| 1126 | |
| 1127 FX_FLOAT fontsize_h = pTextObj->m_TextState.GetFontSizeH(); | |
| 1128 spacing = -fontsize_h * item.m_OriginX / 1000; | |
| 1129 continue; | |
| 1130 } | |
| 1131 FX_FLOAT charSpace = pTextObj->m_TextState.GetObject()->m_CharSpace; | |
| 1132 if (charSpace > 0.001) | |
| 1133 spacing += matrix.TransformDistance(charSpace); | |
| 1134 else if (charSpace < -0.001) | |
| 1135 spacing -= matrix.TransformDistance(FXSYS_fabs(charSpace)); | |
| 1136 spacing -= baseSpace; | |
| 1137 if (spacing && i > 0) { | |
| 1138 int last_width = 0; | |
| 1139 FX_FLOAT fontsize_h = pTextObj->m_TextState.GetFontSizeH(); | |
| 1140 uint32_t space_charcode = pFont->CharCodeFromUnicode(' '); | |
| 1141 FX_FLOAT threshold = 0; | |
| 1142 if (space_charcode != CPDF_Font::kInvalidCharCode) | |
| 1143 threshold = fontsize_h * pFont->GetCharWidthF(space_charcode) / 1000; | |
| 1144 if (threshold > fontsize_h / 3) | |
| 1145 threshold = 0; | |
| 1146 else | |
| 1147 threshold /= 2; | |
| 1148 if (threshold == 0) { | |
| 1149 threshold = fontsize_h; | |
| 1150 int this_width = FXSYS_abs(GetCharWidth(item.m_CharCode, pFont)); | |
| 1151 threshold = this_width > last_width ? (FX_FLOAT)this_width | |
| 1152 : (FX_FLOAT)last_width; | |
| 1153 threshold = NormalizeThreshold(threshold); | |
| 1154 threshold = fontsize_h * threshold / 1000; | |
| 1155 } | |
| 1156 if (threshold && (spacing && spacing >= threshold)) { | |
| 1157 charinfo.m_Unicode = TEXT_SPACE_CHAR; | |
| 1158 charinfo.m_Flag = FPDFTEXT_CHAR_GENERATED; | |
| 1159 charinfo.m_pTextObj = pTextObj; | |
| 1160 charinfo.m_Index = m_TextBuf.GetLength(); | |
| 1161 m_TempTextBuf.AppendChar(TEXT_SPACE_CHAR); | |
| 1162 charinfo.m_CharCode = CPDF_Font::kInvalidCharCode; | |
| 1163 charinfo.m_Matrix.Copy(formMatrix); | |
| 1164 matrix.Transform(item.m_OriginX, item.m_OriginY, charinfo.m_OriginX, | |
| 1165 charinfo.m_OriginY); | |
| 1166 charinfo.m_CharBox = | |
| 1167 CFX_FloatRect(charinfo.m_OriginX, charinfo.m_OriginY, | |
| 1168 charinfo.m_OriginX, charinfo.m_OriginY); | |
| 1169 m_TempCharList.push_back(charinfo); | |
| 1170 } | |
| 1171 if (item.m_CharCode == CPDF_Font::kInvalidCharCode) | |
| 1172 continue; | |
| 1173 } | |
| 1174 spacing = 0; | |
| 1175 CFX_WideString wstrItem = pFont->UnicodeFromCharCode(item.m_CharCode); | |
| 1176 bool bNoUnicode = false; | |
| 1177 FX_WCHAR wChar = wstrItem.GetAt(0); | |
| 1178 if ((wstrItem.IsEmpty() || wChar == 0) && item.m_CharCode) { | |
| 1179 if (wstrItem.IsEmpty()) | |
| 1180 wstrItem += (FX_WCHAR)item.m_CharCode; | |
| 1181 else | |
| 1182 wstrItem.SetAt(0, (FX_WCHAR)item.m_CharCode); | |
| 1183 bNoUnicode = true; | |
| 1184 } | |
| 1185 charinfo.m_Index = -1; | |
| 1186 charinfo.m_CharCode = item.m_CharCode; | |
| 1187 if (bNoUnicode) | |
| 1188 charinfo.m_Flag = FPDFTEXT_CHAR_UNUNICODE; | |
| 1189 else | |
| 1190 charinfo.m_Flag = FPDFTEXT_CHAR_NORMAL; | |
| 1191 charinfo.m_pTextObj = pTextObj; | |
| 1192 charinfo.m_OriginX = 0, charinfo.m_OriginY = 0; | |
| 1193 matrix.Transform(item.m_OriginX, item.m_OriginY, charinfo.m_OriginX, | |
| 1194 charinfo.m_OriginY); | |
| 1195 FX_RECT rect = | |
| 1196 charinfo.m_pTextObj->GetFont()->GetCharBBox(charinfo.m_CharCode); | |
| 1197 charinfo.m_CharBox.top = | |
| 1198 rect.top * pTextObj->GetFontSize() / 1000 + item.m_OriginY; | |
| 1199 charinfo.m_CharBox.left = | |
| 1200 rect.left * pTextObj->GetFontSize() / 1000 + item.m_OriginX; | |
| 1201 charinfo.m_CharBox.right = | |
| 1202 rect.right * pTextObj->GetFontSize() / 1000 + item.m_OriginX; | |
| 1203 charinfo.m_CharBox.bottom = | |
| 1204 rect.bottom * pTextObj->GetFontSize() / 1000 + item.m_OriginY; | |
| 1205 if (fabsf(charinfo.m_CharBox.top - charinfo.m_CharBox.bottom) < 0.01f) { | |
| 1206 charinfo.m_CharBox.top = | |
| 1207 charinfo.m_CharBox.bottom + pTextObj->GetFontSize(); | |
| 1208 } | |
| 1209 if (fabsf(charinfo.m_CharBox.right - charinfo.m_CharBox.left) < 0.01f) { | |
| 1210 charinfo.m_CharBox.right = | |
| 1211 charinfo.m_CharBox.left + pTextObj->GetCharWidth(charinfo.m_CharCode); | |
| 1212 } | |
| 1213 matrix.TransformRect(charinfo.m_CharBox); | |
| 1214 charinfo.m_Matrix.Copy(matrix); | |
| 1215 if (wstrItem.IsEmpty()) { | |
| 1216 charinfo.m_Unicode = 0; | |
| 1217 m_TempCharList.push_back(charinfo); | |
| 1218 m_TempTextBuf.AppendChar(0xfffe); | |
| 1219 continue; | |
| 1220 } else { | |
| 1221 int nTotal = wstrItem.GetLength(); | |
| 1222 bool bDel = false; | |
| 1223 const int count = | |
| 1224 std::min(pdfium::CollectionSize<int>(m_TempCharList), 7); | |
| 1225 FX_FLOAT threshold = charinfo.m_Matrix.TransformXDistance( | |
| 1226 (FX_FLOAT)TEXT_CHARRATIO_GAPDELTA * pTextObj->GetFontSize()); | |
| 1227 for (int n = pdfium::CollectionSize<int>(m_TempCharList); | |
| 1228 n > pdfium::CollectionSize<int>(m_TempCharList) - count; n--) { | |
| 1229 const PAGECHAR_INFO& charinfo1 = m_TempCharList[n - 1]; | |
| 1230 if (charinfo1.m_CharCode == charinfo.m_CharCode && | |
| 1231 charinfo1.m_pTextObj->GetFont() == charinfo.m_pTextObj->GetFont() && | |
| 1232 FXSYS_fabs(charinfo1.m_OriginX - charinfo.m_OriginX) < threshold && | |
| 1233 FXSYS_fabs(charinfo1.m_OriginY - charinfo.m_OriginY) < threshold) { | |
| 1234 bDel = true; | |
| 1235 break; | |
| 1236 } | |
| 1237 } | |
| 1238 if (!bDel) { | |
| 1239 for (int nIndex = 0; nIndex < nTotal; nIndex++) { | |
| 1240 charinfo.m_Unicode = wstrItem.GetAt(nIndex); | |
| 1241 if (charinfo.m_Unicode) { | |
| 1242 charinfo.m_Index = m_TextBuf.GetLength(); | |
| 1243 m_TempTextBuf.AppendChar(charinfo.m_Unicode); | |
| 1244 } else { | |
| 1245 m_TempTextBuf.AppendChar(0xfffe); | |
| 1246 } | |
| 1247 m_TempCharList.push_back(charinfo); | |
| 1248 } | |
| 1249 } else if (i == 0) { | |
| 1250 CFX_WideString str = m_TempTextBuf.MakeString(); | |
| 1251 if (!str.IsEmpty() && | |
| 1252 str.GetAt(str.GetLength() - 1) == TEXT_SPACE_CHAR) { | |
| 1253 m_TempTextBuf.Delete(m_TempTextBuf.GetLength() - 1, 1); | |
| 1254 m_TempCharList.pop_back(); | |
| 1255 } | |
| 1256 } | |
| 1257 } | |
| 1258 } | |
| 1259 if (bIsBidiAndMirrorInverse) | |
| 1260 SwapTempTextBuf(iCharListStartAppend, iBufStartAppend); | |
| 1261 } | |
| 1262 | |
| 1263 CPDF_TextPage::TextOrientation CPDF_TextPage::GetTextObjectWritingMode( | |
| 1264 const CPDF_TextObject* pTextObj) const { | |
| 1265 int32_t nChars = pTextObj->CountChars(); | |
| 1266 if (nChars == 1) | |
| 1267 return m_TextlineDir; | |
| 1268 | |
| 1269 CPDF_TextObjectItem first, last; | |
| 1270 pTextObj->GetCharInfo(0, &first); | |
| 1271 pTextObj->GetCharInfo(nChars - 1, &last); | |
| 1272 CFX_Matrix textMatrix; | |
| 1273 pTextObj->GetTextMatrix(&textMatrix); | |
| 1274 textMatrix.TransformPoint(first.m_OriginX, first.m_OriginY); | |
| 1275 textMatrix.TransformPoint(last.m_OriginX, last.m_OriginY); | |
| 1276 FX_FLOAT dX = FXSYS_fabs(last.m_OriginX - first.m_OriginX); | |
| 1277 FX_FLOAT dY = FXSYS_fabs(last.m_OriginY - first.m_OriginY); | |
| 1278 if (dX <= 0.0001f && dY <= 0.0001f) | |
| 1279 return TextOrientation::Unknown; | |
| 1280 | |
| 1281 CFX_VectorF v(dX, dY); | |
| 1282 v.Normalize(); | |
| 1283 if (v.y <= 0.0872f) | |
| 1284 return v.x <= 0.0872f ? m_TextlineDir : TextOrientation::Horizontal; | |
| 1285 | |
| 1286 if (v.x <= 0.0872f) | |
| 1287 return TextOrientation::Vertical; | |
| 1288 | |
| 1289 return m_TextlineDir; | |
| 1290 } | |
| 1291 | |
| 1292 FX_BOOL CPDF_TextPage::IsHyphen(FX_WCHAR curChar) { | |
| 1293 CFX_WideString strCurText = m_TempTextBuf.MakeString(); | |
| 1294 if (strCurText.IsEmpty()) | |
| 1295 strCurText = m_TextBuf.AsStringC(); | |
| 1296 FX_STRSIZE nCount = strCurText.GetLength(); | |
| 1297 int nIndex = nCount - 1; | |
| 1298 FX_WCHAR wcTmp = strCurText.GetAt(nIndex); | |
| 1299 while (wcTmp == 0x20 && nIndex <= nCount - 1 && nIndex >= 0) | |
| 1300 wcTmp = strCurText.GetAt(--nIndex); | |
| 1301 if (0x2D == wcTmp || 0xAD == wcTmp) { | |
| 1302 if (--nIndex > 0) { | |
| 1303 FX_WCHAR preChar = strCurText.GetAt((nIndex)); | |
| 1304 if (((preChar >= L'A' && preChar <= L'Z') || | |
| 1305 (preChar >= L'a' && preChar <= L'z')) && | |
| 1306 ((curChar >= L'A' && curChar <= L'Z') || | |
| 1307 (curChar >= L'a' && curChar <= L'z'))) { | |
| 1308 return TRUE; | |
| 1309 } | |
| 1310 } | |
| 1311 const PAGECHAR_INFO* preInfo; | |
| 1312 if (!m_TempCharList.empty()) | |
| 1313 preInfo = &m_TempCharList.back(); | |
| 1314 else if (!m_CharList.empty()) | |
| 1315 preInfo = &m_CharList.back(); | |
| 1316 else | |
| 1317 return FALSE; | |
| 1318 if (FPDFTEXT_CHAR_PIECE == preInfo->m_Flag && | |
| 1319 (0xAD == preInfo->m_Unicode || 0x2D == preInfo->m_Unicode)) { | |
| 1320 return TRUE; | |
| 1321 } | |
| 1322 } | |
| 1323 return FALSE; | |
| 1324 } | |
| 1325 | |
| 1326 CPDF_TextPage::GenerateCharacter CPDF_TextPage::ProcessInsertObject( | |
| 1327 const CPDF_TextObject* pObj, | |
| 1328 const CFX_Matrix& formMatrix) { | |
| 1329 FindPreviousTextObject(); | |
| 1330 TextOrientation WritingMode = GetTextObjectWritingMode(pObj); | |
| 1331 if (WritingMode == TextOrientation::Unknown) | |
| 1332 WritingMode = GetTextObjectWritingMode(m_pPreTextObj); | |
| 1333 | |
| 1334 CFX_FloatRect this_rect(pObj->m_Left, pObj->m_Bottom, pObj->m_Right, | |
| 1335 pObj->m_Top); | |
| 1336 CFX_FloatRect prev_rect(m_pPreTextObj->m_Left, m_pPreTextObj->m_Bottom, | |
| 1337 m_pPreTextObj->m_Right, m_pPreTextObj->m_Top); | |
| 1338 CPDF_TextObjectItem PrevItem, item; | |
| 1339 int nItem = m_pPreTextObj->CountItems(); | |
| 1340 m_pPreTextObj->GetItemInfo(nItem - 1, &PrevItem); | |
| 1341 pObj->GetItemInfo(0, &item); | |
| 1342 CFX_WideString wstrItem = | |
| 1343 pObj->GetFont()->UnicodeFromCharCode(item.m_CharCode); | |
| 1344 if (wstrItem.IsEmpty()) | |
| 1345 wstrItem += static_cast<FX_WCHAR>(item.m_CharCode); | |
| 1346 FX_WCHAR curChar = wstrItem.GetAt(0); | |
| 1347 if (WritingMode == TextOrientation::Horizontal) { | |
| 1348 if (this_rect.Height() > 4.5 && prev_rect.Height() > 4.5) { | |
| 1349 FX_FLOAT top = | |
| 1350 this_rect.top < prev_rect.top ? this_rect.top : prev_rect.top; | |
| 1351 FX_FLOAT bottom = this_rect.bottom > prev_rect.bottom ? this_rect.bottom | |
| 1352 : prev_rect.bottom; | |
| 1353 if (bottom >= top) { | |
| 1354 return IsHyphen(curChar) ? GenerateCharacter::Hyphen | |
| 1355 : GenerateCharacter::LineBreak; | |
| 1356 } | |
| 1357 } | |
| 1358 } else if (WritingMode == TextOrientation::Vertical) { | |
| 1359 if (this_rect.Width() > pObj->GetFontSize() * 0.1f && | |
| 1360 prev_rect.Width() > m_pPreTextObj->GetFontSize() * 0.1f) { | |
| 1361 FX_FLOAT left = this_rect.left > m_CurlineRect.left ? this_rect.left | |
| 1362 : m_CurlineRect.left; | |
| 1363 FX_FLOAT right = this_rect.right < m_CurlineRect.right | |
| 1364 ? this_rect.right | |
| 1365 : m_CurlineRect.right; | |
| 1366 if (right <= left) { | |
| 1367 return IsHyphen(curChar) ? GenerateCharacter::Hyphen | |
| 1368 : GenerateCharacter::LineBreak; | |
| 1369 } | |
| 1370 } | |
| 1371 } | |
| 1372 FX_FLOAT last_pos = PrevItem.m_OriginX; | |
| 1373 int nLastWidth = GetCharWidth(PrevItem.m_CharCode, m_pPreTextObj->GetFont()); | |
| 1374 FX_FLOAT last_width = nLastWidth * m_pPreTextObj->GetFontSize() / 1000; | |
| 1375 last_width = FXSYS_fabs(last_width); | |
| 1376 int nThisWidth = GetCharWidth(item.m_CharCode, pObj->GetFont()); | |
| 1377 FX_FLOAT this_width = nThisWidth * pObj->GetFontSize() / 1000; | |
| 1378 this_width = FXSYS_fabs(this_width); | |
| 1379 FX_FLOAT threshold = | |
| 1380 last_width > this_width ? last_width / 4 : this_width / 4; | |
| 1381 CFX_Matrix prev_matrix, prev_reverse; | |
| 1382 m_pPreTextObj->GetTextMatrix(&prev_matrix); | |
| 1383 prev_matrix.Concat(m_perMatrix); | |
| 1384 prev_reverse.SetReverse(prev_matrix); | |
| 1385 FX_FLOAT x = pObj->GetPosX(); | |
| 1386 FX_FLOAT y = pObj->GetPosY(); | |
| 1387 formMatrix.Transform(x, y); | |
| 1388 prev_reverse.Transform(x, y); | |
| 1389 if (last_width < this_width) | |
| 1390 threshold = prev_reverse.TransformDistance(threshold); | |
| 1391 bool bNewline = false; | |
| 1392 if (WritingMode == TextOrientation::Horizontal) { | |
| 1393 CFX_FloatRect rect1(m_pPreTextObj->m_Left, pObj->m_Bottom, | |
| 1394 m_pPreTextObj->m_Right, pObj->m_Top); | |
| 1395 CFX_FloatRect rect2(m_pPreTextObj->m_Left, m_pPreTextObj->m_Bottom, | |
| 1396 m_pPreTextObj->m_Right, m_pPreTextObj->m_Top); | |
| 1397 CFX_FloatRect rect3 = rect1; | |
| 1398 rect1.Intersect(rect2); | |
| 1399 if ((rect1.IsEmpty() && rect2.Height() > 5 && rect3.Height() > 5) || | |
| 1400 ((y > threshold * 2 || y < threshold * -3) && | |
| 1401 (FXSYS_fabs(y) < 1 ? FXSYS_fabs(x) < FXSYS_fabs(y) : TRUE))) { | |
| 1402 bNewline = true; | |
| 1403 if (nItem > 1) { | |
| 1404 CPDF_TextObjectItem tempItem; | |
| 1405 m_pPreTextObj->GetItemInfo(0, &tempItem); | |
| 1406 CFX_Matrix m; | |
| 1407 m_pPreTextObj->GetTextMatrix(&m); | |
| 1408 if (PrevItem.m_OriginX > tempItem.m_OriginX && | |
| 1409 m_DisplayMatrix.a > 0.9 && m_DisplayMatrix.b < 0.1 && | |
| 1410 m_DisplayMatrix.c < 0.1 && m_DisplayMatrix.d < -0.9 && m.b < 0.1 && | |
| 1411 m.c < 0.1) { | |
| 1412 CFX_FloatRect re(0, m_pPreTextObj->m_Bottom, 1000, | |
| 1413 m_pPreTextObj->m_Top); | |
| 1414 if (re.Contains(pObj->GetPosX(), pObj->GetPosY())) { | |
| 1415 bNewline = false; | |
| 1416 } else { | |
| 1417 CFX_FloatRect rect(0, pObj->m_Bottom, 1000, pObj->m_Top); | |
| 1418 if (rect.Contains(m_pPreTextObj->GetPosX(), | |
| 1419 m_pPreTextObj->GetPosY())) { | |
| 1420 bNewline = false; | |
| 1421 } | |
| 1422 } | |
| 1423 } | |
| 1424 } | |
| 1425 } | |
| 1426 } | |
| 1427 if (bNewline) { | |
| 1428 return IsHyphen(curChar) ? GenerateCharacter::Hyphen | |
| 1429 : GenerateCharacter::LineBreak; | |
| 1430 } | |
| 1431 | |
| 1432 int32_t nChars = pObj->CountChars(); | |
| 1433 if (nChars == 1 && (0x2D == curChar || 0xAD == curChar) && | |
| 1434 IsHyphen(curChar)) { | |
| 1435 return GenerateCharacter::Hyphen; | |
| 1436 } | |
| 1437 CFX_WideString PrevStr = | |
| 1438 m_pPreTextObj->GetFont()->UnicodeFromCharCode(PrevItem.m_CharCode); | |
| 1439 FX_WCHAR preChar = PrevStr.GetAt(PrevStr.GetLength() - 1); | |
| 1440 CFX_Matrix matrix; | |
| 1441 pObj->GetTextMatrix(&matrix); | |
| 1442 matrix.Concat(formMatrix); | |
| 1443 threshold = (FX_FLOAT)(nLastWidth > nThisWidth ? nLastWidth : nThisWidth); | |
| 1444 threshold = threshold > 400 | |
| 1445 ? (threshold < 700 | |
| 1446 ? threshold / 4 | |
| 1447 : (threshold > 800 ? threshold / 6 : threshold / 5)) | |
| 1448 : (threshold / 2); | |
| 1449 if (nLastWidth >= nThisWidth) { | |
| 1450 threshold *= FXSYS_fabs(m_pPreTextObj->GetFontSize()); | |
| 1451 } else { | |
| 1452 threshold *= FXSYS_fabs(pObj->GetFontSize()); | |
| 1453 threshold = matrix.TransformDistance(threshold); | |
| 1454 threshold = prev_reverse.TransformDistance(threshold); | |
| 1455 } | |
| 1456 threshold /= 1000; | |
| 1457 if ((threshold < 1.4881 && threshold > 1.4879) || | |
| 1458 (threshold < 1.39001 && threshold > 1.38999)) { | |
| 1459 threshold *= 1.5; | |
| 1460 } | |
| 1461 if (FXSYS_fabs(last_pos + last_width - x) > threshold && curChar != L' ' && | |
| 1462 preChar != L' ') { | |
| 1463 if (curChar != L' ' && preChar != L' ') { | |
| 1464 if ((x - last_pos - last_width) > threshold || | |
| 1465 (last_pos - x - last_width) > threshold) { | |
| 1466 return GenerateCharacter::Space; | |
| 1467 } | |
| 1468 if (x < 0 && (last_pos - x - last_width) > threshold) | |
| 1469 return GenerateCharacter::Space; | |
| 1470 if ((x - last_pos - last_width) > this_width || | |
| 1471 (x - last_pos - this_width) > last_width) { | |
| 1472 return GenerateCharacter::Space; | |
| 1473 } | |
| 1474 } | |
| 1475 } | |
| 1476 return GenerateCharacter::None; | |
| 1477 } | |
| 1478 | |
| 1479 FX_BOOL CPDF_TextPage::IsSameTextObject(CPDF_TextObject* pTextObj1, | |
| 1480 CPDF_TextObject* pTextObj2) { | |
| 1481 if (!pTextObj1 || !pTextObj2) | |
| 1482 return FALSE; | |
| 1483 CFX_FloatRect rcPreObj(pTextObj2->m_Left, pTextObj2->m_Bottom, | |
| 1484 pTextObj2->m_Right, pTextObj2->m_Top); | |
| 1485 CFX_FloatRect rcCurObj(pTextObj1->m_Left, pTextObj1->m_Bottom, | |
| 1486 pTextObj1->m_Right, pTextObj1->m_Top); | |
| 1487 if (rcPreObj.IsEmpty() && rcCurObj.IsEmpty()) { | |
| 1488 FX_FLOAT dbXdif = FXSYS_fabs(rcPreObj.left - rcCurObj.left); | |
| 1489 size_t nCount = m_CharList.size(); | |
| 1490 if (nCount >= 2) { | |
| 1491 PAGECHAR_INFO perCharTemp = m_CharList[nCount - 2]; | |
| 1492 FX_FLOAT dbSpace = perCharTemp.m_CharBox.Width(); | |
| 1493 if (dbXdif > dbSpace) | |
| 1494 return FALSE; | |
| 1495 } | |
| 1496 } | |
| 1497 if (!rcPreObj.IsEmpty() || !rcCurObj.IsEmpty()) { | |
| 1498 rcPreObj.Intersect(rcCurObj); | |
| 1499 if (rcPreObj.IsEmpty()) | |
| 1500 return FALSE; | |
| 1501 if (FXSYS_fabs(rcPreObj.Width() - rcCurObj.Width()) > | |
| 1502 rcCurObj.Width() / 2) { | |
| 1503 return FALSE; | |
| 1504 } | |
| 1505 if (pTextObj2->GetFontSize() != pTextObj1->GetFontSize()) | |
| 1506 return FALSE; | |
| 1507 } | |
| 1508 int nPreCount = pTextObj2->CountItems(); | |
| 1509 int nCurCount = pTextObj1->CountItems(); | |
| 1510 if (nPreCount != nCurCount) | |
| 1511 return FALSE; | |
| 1512 // If both objects have no items, consider them same. | |
| 1513 if (!nPreCount) | |
| 1514 return TRUE; | |
| 1515 | |
| 1516 CPDF_TextObjectItem itemPer = {0, 0.0f, 0.0f}; | |
| 1517 CPDF_TextObjectItem itemCur = {0, 0.0f, 0.0f}; | |
| 1518 for (int i = 0; i < nPreCount; i++) { | |
| 1519 pTextObj2->GetItemInfo(i, &itemPer); | |
| 1520 pTextObj1->GetItemInfo(i, &itemCur); | |
| 1521 if (itemCur.m_CharCode != itemPer.m_CharCode) | |
| 1522 return FALSE; | |
| 1523 } | |
| 1524 if (FXSYS_fabs(pTextObj1->GetPosX() - pTextObj2->GetPosX()) > | |
| 1525 GetCharWidth(itemPer.m_CharCode, pTextObj2->GetFont()) * | |
| 1526 pTextObj2->GetFontSize() / 1000 * 0.9 || | |
| 1527 FXSYS_fabs(pTextObj1->GetPosY() - pTextObj2->GetPosY()) > | |
| 1528 std::max(std::max(rcPreObj.Height(), rcPreObj.Width()), | |
| 1529 pTextObj2->GetFontSize()) / | |
| 1530 8) { | |
| 1531 return FALSE; | |
| 1532 } | |
| 1533 return TRUE; | |
| 1534 } | |
| 1535 | |
| 1536 FX_BOOL CPDF_TextPage::IsSameAsPreTextObject( | |
| 1537 CPDF_TextObject* pTextObj, | |
| 1538 const CPDF_PageObjectList* pObjList, | |
| 1539 CPDF_PageObjectList::const_iterator iter) { | |
| 1540 int i = 0; | |
| 1541 while (i < 5 && iter != pObjList->begin()) { | |
| 1542 --iter; | |
| 1543 CPDF_PageObject* pOtherObj = iter->get(); | |
| 1544 if (pOtherObj == pTextObj || !pOtherObj->IsText()) | |
| 1545 continue; | |
| 1546 if (IsSameTextObject(pOtherObj->AsText(), pTextObj)) | |
| 1547 return TRUE; | |
| 1548 ++i; | |
| 1549 } | |
| 1550 return FALSE; | |
| 1551 } | |
| 1552 | |
| 1553 FX_BOOL CPDF_TextPage::GenerateCharInfo(FX_WCHAR unicode, PAGECHAR_INFO& info) { | |
| 1554 const PAGECHAR_INFO* preChar; | |
| 1555 if (!m_TempCharList.empty()) | |
| 1556 preChar = &m_TempCharList.back(); | |
| 1557 else if (!m_CharList.empty()) | |
| 1558 preChar = &m_CharList.back(); | |
| 1559 else | |
| 1560 return FALSE; | |
| 1561 info.m_Index = m_TextBuf.GetLength(); | |
| 1562 info.m_Unicode = unicode; | |
| 1563 info.m_pTextObj = nullptr; | |
| 1564 info.m_CharCode = CPDF_Font::kInvalidCharCode; | |
| 1565 info.m_Flag = FPDFTEXT_CHAR_GENERATED; | |
| 1566 int preWidth = 0; | |
| 1567 if (preChar->m_pTextObj && preChar->m_CharCode != -1) { | |
| 1568 preWidth = | |
| 1569 GetCharWidth(preChar->m_CharCode, preChar->m_pTextObj->GetFont()); | |
| 1570 } | |
| 1571 | |
| 1572 FX_FLOAT fFontSize = preChar->m_pTextObj ? preChar->m_pTextObj->GetFontSize() | |
| 1573 : preChar->m_CharBox.Height(); | |
| 1574 if (!fFontSize) | |
| 1575 fFontSize = kDefaultFontSize; | |
| 1576 | |
| 1577 info.m_OriginX = preChar->m_OriginX + preWidth * (fFontSize) / 1000; | |
| 1578 info.m_OriginY = preChar->m_OriginY; | |
| 1579 info.m_CharBox = CFX_FloatRect(info.m_OriginX, info.m_OriginY, info.m_OriginX, | |
| 1580 info.m_OriginY); | |
| 1581 return TRUE; | |
| 1582 } | |
| 1583 | |
| 1584 FX_BOOL CPDF_TextPage::IsRectIntersect(const CFX_FloatRect& rect1, | |
| 1585 const CFX_FloatRect& rect2) { | |
| 1586 CFX_FloatRect rect = rect1; | |
| 1587 rect.Intersect(rect2); | |
| 1588 return !rect.IsEmpty(); | |
| 1589 } | |
| 1590 | |
| 1591 CPDF_TextPageFind::CPDF_TextPageFind(const CPDF_TextPage* pTextPage) | |
| 1592 : m_pTextPage(pTextPage), | |
| 1593 m_flags(0), | |
| 1594 m_findNextStart(-1), | |
| 1595 m_findPreStart(-1), | |
| 1596 m_bMatchCase(FALSE), | |
| 1597 m_bMatchWholeWord(FALSE), | |
| 1598 m_resStart(0), | |
| 1599 m_resEnd(-1), | |
| 1600 m_IsFind(FALSE) { | |
| 1601 m_strText = m_pTextPage->GetPageText(); | |
| 1602 int nCount = pTextPage->CountChars(); | |
| 1603 if (nCount) { | |
| 1604 m_CharIndex.push_back(0); | |
| 1605 } | |
| 1606 for (int i = 0; i < nCount; i++) { | |
| 1607 FPDF_CHAR_INFO info; | |
| 1608 pTextPage->GetCharInfo(i, &info); | |
| 1609 int indexSize = pdfium::CollectionSize<int>(m_CharIndex); | |
| 1610 if (info.m_Flag == FPDFTEXT_CHAR_NORMAL || | |
| 1611 info.m_Flag == FPDFTEXT_CHAR_GENERATED) { | |
| 1612 if (indexSize % 2) { | |
| 1613 m_CharIndex.push_back(1); | |
| 1614 } else { | |
| 1615 if (indexSize <= 0) { | |
| 1616 continue; | |
| 1617 } | |
| 1618 m_CharIndex[indexSize - 1] += 1; | |
| 1619 } | |
| 1620 } else { | |
| 1621 if (indexSize % 2) { | |
| 1622 if (indexSize <= 0) { | |
| 1623 continue; | |
| 1624 } | |
| 1625 m_CharIndex[indexSize - 1] = i + 1; | |
| 1626 } else { | |
| 1627 m_CharIndex.push_back(i + 1); | |
| 1628 } | |
| 1629 } | |
| 1630 } | |
| 1631 int indexSize = pdfium::CollectionSize<int>(m_CharIndex); | |
| 1632 if (indexSize % 2) { | |
| 1633 m_CharIndex.erase(m_CharIndex.begin() + indexSize - 1); | |
| 1634 } | |
| 1635 } | |
| 1636 | |
| 1637 CPDF_TextPageFind::~CPDF_TextPageFind() {} | |
| 1638 | |
| 1639 int CPDF_TextPageFind::GetCharIndex(int index) const { | |
| 1640 return m_pTextPage->CharIndexFromTextIndex(index); | |
| 1641 } | |
| 1642 | |
| 1643 FX_BOOL CPDF_TextPageFind::FindFirst(const CFX_WideString& findwhat, | |
| 1644 int flags, | |
| 1645 int startPos) { | |
| 1646 if (!m_pTextPage) { | |
| 1647 return FALSE; | |
| 1648 } | |
| 1649 if (m_strText.IsEmpty() || m_bMatchCase != (flags & FPDFTEXT_MATCHCASE)) { | |
| 1650 m_strText = m_pTextPage->GetPageText(); | |
| 1651 } | |
| 1652 CFX_WideString findwhatStr = findwhat; | |
| 1653 m_findWhat = findwhatStr; | |
| 1654 m_flags = flags; | |
| 1655 m_bMatchCase = flags & FPDFTEXT_MATCHCASE; | |
| 1656 if (m_strText.IsEmpty()) { | |
| 1657 m_IsFind = FALSE; | |
| 1658 return TRUE; | |
| 1659 } | |
| 1660 FX_STRSIZE len = findwhatStr.GetLength(); | |
| 1661 if (!m_bMatchCase) { | |
| 1662 findwhatStr.MakeLower(); | |
| 1663 m_strText.MakeLower(); | |
| 1664 } | |
| 1665 m_bMatchWholeWord = flags & FPDFTEXT_MATCHWHOLEWORD; | |
| 1666 m_findNextStart = startPos; | |
| 1667 if (startPos == -1) { | |
| 1668 m_findPreStart = m_strText.GetLength() - 1; | |
| 1669 } else { | |
| 1670 m_findPreStart = startPos; | |
| 1671 } | |
| 1672 m_csFindWhatArray.clear(); | |
| 1673 int i = 0; | |
| 1674 while (i < len) { | |
| 1675 if (findwhatStr.GetAt(i) != ' ') { | |
| 1676 break; | |
| 1677 } | |
| 1678 i++; | |
| 1679 } | |
| 1680 if (i < len) { | |
| 1681 ExtractFindWhat(findwhatStr); | |
| 1682 } else { | |
| 1683 m_csFindWhatArray.push_back(findwhatStr); | |
| 1684 } | |
| 1685 if (m_csFindWhatArray.empty()) { | |
| 1686 return FALSE; | |
| 1687 } | |
| 1688 m_IsFind = TRUE; | |
| 1689 m_resStart = 0; | |
| 1690 m_resEnd = -1; | |
| 1691 return TRUE; | |
| 1692 } | |
| 1693 | |
| 1694 FX_BOOL CPDF_TextPageFind::FindNext() { | |
| 1695 if (!m_pTextPage) { | |
| 1696 return FALSE; | |
| 1697 } | |
| 1698 m_resArray.clear(); | |
| 1699 if (m_findNextStart == -1) { | |
| 1700 return FALSE; | |
| 1701 } | |
| 1702 if (m_strText.IsEmpty()) { | |
| 1703 m_IsFind = FALSE; | |
| 1704 return m_IsFind; | |
| 1705 } | |
| 1706 int strLen = m_strText.GetLength(); | |
| 1707 if (m_findNextStart > strLen - 1) { | |
| 1708 m_IsFind = FALSE; | |
| 1709 return m_IsFind; | |
| 1710 } | |
| 1711 int nCount = pdfium::CollectionSize<int>(m_csFindWhatArray); | |
| 1712 int nResultPos = 0; | |
| 1713 int nStartPos = 0; | |
| 1714 nStartPos = m_findNextStart; | |
| 1715 FX_BOOL bSpaceStart = FALSE; | |
| 1716 for (int iWord = 0; iWord < nCount; iWord++) { | |
| 1717 CFX_WideString csWord = m_csFindWhatArray[iWord]; | |
| 1718 if (csWord.IsEmpty()) { | |
| 1719 if (iWord == nCount - 1) { | |
| 1720 FX_WCHAR strInsert = m_strText.GetAt(nStartPos); | |
| 1721 if (strInsert == TEXT_LINEFEED_CHAR || strInsert == TEXT_SPACE_CHAR || | |
| 1722 strInsert == TEXT_RETURN_CHAR || strInsert == 160) { | |
| 1723 nResultPos = nStartPos + 1; | |
| 1724 break; | |
| 1725 } | |
| 1726 iWord = -1; | |
| 1727 } else if (iWord == 0) { | |
| 1728 bSpaceStart = TRUE; | |
| 1729 } | |
| 1730 continue; | |
| 1731 } | |
| 1732 int endIndex; | |
| 1733 nResultPos = m_strText.Find(csWord.c_str(), nStartPos); | |
| 1734 if (nResultPos == -1) { | |
| 1735 m_IsFind = FALSE; | |
| 1736 return m_IsFind; | |
| 1737 } | |
| 1738 endIndex = nResultPos + csWord.GetLength() - 1; | |
| 1739 if (iWord == 0) { | |
| 1740 m_resStart = nResultPos; | |
| 1741 } | |
| 1742 FX_BOOL bMatch = TRUE; | |
| 1743 if (iWord != 0 && !bSpaceStart) { | |
| 1744 int PreResEndPos = nStartPos; | |
| 1745 int curChar = csWord.GetAt(0); | |
| 1746 CFX_WideString lastWord = m_csFindWhatArray[iWord - 1]; | |
| 1747 int lastChar = lastWord.GetAt(lastWord.GetLength() - 1); | |
| 1748 if (nStartPos == nResultPos && | |
| 1749 !(IsIgnoreSpaceCharacter(lastChar) || | |
| 1750 IsIgnoreSpaceCharacter(curChar))) { | |
| 1751 bMatch = FALSE; | |
| 1752 } | |
| 1753 for (int d = PreResEndPos; d < nResultPos; d++) { | |
| 1754 FX_WCHAR strInsert = m_strText.GetAt(d); | |
| 1755 if (strInsert != TEXT_LINEFEED_CHAR && strInsert != TEXT_SPACE_CHAR && | |
| 1756 strInsert != TEXT_RETURN_CHAR && strInsert != 160) { | |
| 1757 bMatch = FALSE; | |
| 1758 break; | |
| 1759 } | |
| 1760 } | |
| 1761 } else if (bSpaceStart) { | |
| 1762 if (nResultPos > 0) { | |
| 1763 FX_WCHAR strInsert = m_strText.GetAt(nResultPos - 1); | |
| 1764 if (strInsert != TEXT_LINEFEED_CHAR && strInsert != TEXT_SPACE_CHAR && | |
| 1765 strInsert != TEXT_RETURN_CHAR && strInsert != 160) { | |
| 1766 bMatch = FALSE; | |
| 1767 m_resStart = nResultPos; | |
| 1768 } else { | |
| 1769 m_resStart = nResultPos - 1; | |
| 1770 } | |
| 1771 } | |
| 1772 } | |
| 1773 if (m_bMatchWholeWord && bMatch) { | |
| 1774 bMatch = IsMatchWholeWord(m_strText, nResultPos, endIndex); | |
| 1775 } | |
| 1776 nStartPos = endIndex + 1; | |
| 1777 if (!bMatch) { | |
| 1778 iWord = -1; | |
| 1779 if (bSpaceStart) { | |
| 1780 nStartPos = m_resStart + m_csFindWhatArray[1].GetLength(); | |
| 1781 } else { | |
| 1782 nStartPos = m_resStart + m_csFindWhatArray[0].GetLength(); | |
| 1783 } | |
| 1784 } | |
| 1785 } | |
| 1786 m_resEnd = nResultPos + m_csFindWhatArray.back().GetLength() - 1; | |
| 1787 m_IsFind = TRUE; | |
| 1788 int resStart = GetCharIndex(m_resStart); | |
| 1789 int resEnd = GetCharIndex(m_resEnd); | |
| 1790 m_resArray = m_pTextPage->GetRectArray(resStart, resEnd - resStart + 1); | |
| 1791 if (m_flags & FPDFTEXT_CONSECUTIVE) { | |
| 1792 m_findNextStart = m_resStart + 1; | |
| 1793 m_findPreStart = m_resEnd - 1; | |
| 1794 } else { | |
| 1795 m_findNextStart = m_resEnd + 1; | |
| 1796 m_findPreStart = m_resStart - 1; | |
| 1797 } | |
| 1798 return m_IsFind; | |
| 1799 } | |
| 1800 | |
| 1801 FX_BOOL CPDF_TextPageFind::FindPrev() { | |
| 1802 if (!m_pTextPage) { | |
| 1803 return FALSE; | |
| 1804 } | |
| 1805 m_resArray.clear(); | |
| 1806 if (m_strText.IsEmpty() || m_findPreStart < 0) { | |
| 1807 m_IsFind = FALSE; | |
| 1808 return m_IsFind; | |
| 1809 } | |
| 1810 CPDF_TextPageFind findEngine(m_pTextPage); | |
| 1811 FX_BOOL ret = findEngine.FindFirst(m_findWhat, m_flags); | |
| 1812 if (!ret) { | |
| 1813 m_IsFind = FALSE; | |
| 1814 return m_IsFind; | |
| 1815 } | |
| 1816 int order = -1, MatchedCount = 0; | |
| 1817 while (ret) { | |
| 1818 ret = findEngine.FindNext(); | |
| 1819 if (ret) { | |
| 1820 int order1 = findEngine.GetCurOrder(); | |
| 1821 int MatchedCount1 = findEngine.GetMatchedCount(); | |
| 1822 if (((order1 + MatchedCount1) - 1) > m_findPreStart) { | |
| 1823 break; | |
| 1824 } | |
| 1825 order = order1; | |
| 1826 MatchedCount = MatchedCount1; | |
| 1827 } | |
| 1828 } | |
| 1829 if (order == -1) { | |
| 1830 m_IsFind = FALSE; | |
| 1831 return m_IsFind; | |
| 1832 } | |
| 1833 m_resStart = m_pTextPage->TextIndexFromCharIndex(order); | |
| 1834 m_resEnd = m_pTextPage->TextIndexFromCharIndex(order + MatchedCount - 1); | |
| 1835 m_IsFind = TRUE; | |
| 1836 m_resArray = m_pTextPage->GetRectArray(order, MatchedCount); | |
| 1837 if (m_flags & FPDFTEXT_CONSECUTIVE) { | |
| 1838 m_findNextStart = m_resStart + 1; | |
| 1839 m_findPreStart = m_resEnd - 1; | |
| 1840 } else { | |
| 1841 m_findNextStart = m_resEnd + 1; | |
| 1842 m_findPreStart = m_resStart - 1; | |
| 1843 } | |
| 1844 return m_IsFind; | |
| 1845 } | |
| 1846 | |
| 1847 void CPDF_TextPageFind::ExtractFindWhat(const CFX_WideString& findwhat) { | |
| 1848 if (findwhat.IsEmpty()) { | |
| 1849 return; | |
| 1850 } | |
| 1851 int index = 0; | |
| 1852 while (1) { | |
| 1853 CFX_WideString csWord = TEXT_EMPTY; | |
| 1854 int ret = | |
| 1855 ExtractSubString(csWord, findwhat.c_str(), index, TEXT_SPACE_CHAR); | |
| 1856 if (csWord.IsEmpty()) { | |
| 1857 if (ret) { | |
| 1858 m_csFindWhatArray.push_back(L""); | |
| 1859 index++; | |
| 1860 continue; | |
| 1861 } else { | |
| 1862 break; | |
| 1863 } | |
| 1864 } | |
| 1865 int pos = 0; | |
| 1866 while (pos < csWord.GetLength()) { | |
| 1867 CFX_WideString curStr = csWord.Mid(pos, 1); | |
| 1868 FX_WCHAR curChar = csWord.GetAt(pos); | |
| 1869 if (IsIgnoreSpaceCharacter(curChar)) { | |
| 1870 if (pos > 0 && curChar == 0x2019) { | |
| 1871 pos++; | |
| 1872 continue; | |
| 1873 } | |
| 1874 if (pos > 0) { | |
| 1875 m_csFindWhatArray.push_back(csWord.Mid(0, pos)); | |
| 1876 } | |
| 1877 m_csFindWhatArray.push_back(curStr); | |
| 1878 if (pos == csWord.GetLength() - 1) { | |
| 1879 csWord.clear(); | |
| 1880 break; | |
| 1881 } | |
| 1882 csWord = csWord.Right(csWord.GetLength() - pos - 1); | |
| 1883 pos = 0; | |
| 1884 continue; | |
| 1885 } | |
| 1886 pos++; | |
| 1887 } | |
| 1888 if (!csWord.IsEmpty()) { | |
| 1889 m_csFindWhatArray.push_back(csWord); | |
| 1890 } | |
| 1891 index++; | |
| 1892 } | |
| 1893 } | |
| 1894 | |
| 1895 FX_BOOL CPDF_TextPageFind::IsMatchWholeWord(const CFX_WideString& csPageText, | |
| 1896 int startPos, | |
| 1897 int endPos) { | |
| 1898 FX_WCHAR char_left = 0; | |
| 1899 FX_WCHAR char_right = 0; | |
| 1900 int char_count = endPos - startPos + 1; | |
| 1901 if (char_count < 1) { | |
| 1902 return FALSE; | |
| 1903 } | |
| 1904 if (char_count == 1 && csPageText.GetAt(startPos) > 255) { | |
| 1905 return TRUE; | |
| 1906 } | |
| 1907 if (startPos - 1 >= 0) { | |
| 1908 char_left = csPageText.GetAt(startPos - 1); | |
| 1909 } | |
| 1910 if (startPos + char_count < csPageText.GetLength()) { | |
| 1911 char_right = csPageText.GetAt(startPos + char_count); | |
| 1912 } | |
| 1913 if ((char_left > 'A' && char_left < 'a') || | |
| 1914 (char_left > 'a' && char_left < 'z') || | |
| 1915 (char_left > 0xfb00 && char_left < 0xfb06) || std::iswdigit(char_left) || | |
| 1916 (char_right > 'A' && char_right < 'a') || | |
| 1917 (char_right > 'a' && char_right < 'z') || | |
| 1918 (char_right > 0xfb00 && char_right < 0xfb06) || | |
| 1919 std::iswdigit(char_right)) { | |
| 1920 return FALSE; | |
| 1921 } | |
| 1922 if (!(('A' > char_left || char_left > 'Z') && | |
| 1923 ('a' > char_left || char_left > 'z') && | |
| 1924 ('A' > char_right || char_right > 'Z') && | |
| 1925 ('a' > char_right || char_right > 'z'))) { | |
| 1926 return FALSE; | |
| 1927 } | |
| 1928 if (char_count > 0) { | |
| 1929 if (csPageText.GetAt(startPos) >= L'0' && | |
| 1930 csPageText.GetAt(startPos) <= L'9' && char_left >= L'0' && | |
| 1931 char_left <= L'9') { | |
| 1932 return FALSE; | |
| 1933 } | |
| 1934 if (csPageText.GetAt(endPos) >= L'0' && csPageText.GetAt(endPos) <= L'9' && | |
| 1935 char_right >= L'0' && char_right <= L'9') { | |
| 1936 return FALSE; | |
| 1937 } | |
| 1938 } | |
| 1939 return TRUE; | |
| 1940 } | |
| 1941 | |
| 1942 FX_BOOL CPDF_TextPageFind::ExtractSubString(CFX_WideString& rString, | |
| 1943 const FX_WCHAR* lpszFullString, | |
| 1944 int iSubString, | |
| 1945 FX_WCHAR chSep) { | |
| 1946 if (!lpszFullString) { | |
| 1947 return FALSE; | |
| 1948 } | |
| 1949 while (iSubString--) { | |
| 1950 lpszFullString = wcschr(lpszFullString, chSep); | |
| 1951 if (!lpszFullString) { | |
| 1952 rString.clear(); | |
| 1953 return FALSE; | |
| 1954 } | |
| 1955 lpszFullString++; | |
| 1956 while (*lpszFullString == chSep) { | |
| 1957 lpszFullString++; | |
| 1958 } | |
| 1959 } | |
| 1960 const FX_WCHAR* lpchEnd = wcschr(lpszFullString, chSep); | |
| 1961 int nLen = lpchEnd ? (int)(lpchEnd - lpszFullString) | |
| 1962 : (int)FXSYS_wcslen(lpszFullString); | |
| 1963 ASSERT(nLen >= 0); | |
| 1964 FXSYS_memcpy(rString.GetBuffer(nLen), lpszFullString, | |
| 1965 nLen * sizeof(FX_WCHAR)); | |
| 1966 rString.ReleaseBuffer(); | |
| 1967 return TRUE; | |
| 1968 } | |
| 1969 | |
| 1970 CFX_WideString CPDF_TextPageFind::MakeReverse(const CFX_WideString& str) { | |
| 1971 CFX_WideString str2; | |
| 1972 str2.clear(); | |
| 1973 int nlen = str.GetLength(); | |
| 1974 for (int i = nlen - 1; i >= 0; i--) { | |
| 1975 str2 += str.GetAt(i); | |
| 1976 } | |
| 1977 return str2; | |
| 1978 } | |
| 1979 | |
| 1980 int CPDF_TextPageFind::GetCurOrder() const { | |
| 1981 return GetCharIndex(m_resStart); | |
| 1982 } | |
| 1983 | |
| 1984 int CPDF_TextPageFind::GetMatchedCount() const { | |
| 1985 int resStart = GetCharIndex(m_resStart); | |
| 1986 int resEnd = GetCharIndex(m_resEnd); | |
| 1987 return resEnd - resStart + 1; | |
| 1988 } | |
| 1989 | |
| 1990 CPDF_LinkExtract::CPDF_LinkExtract(const CPDF_TextPage* pTextPage) | |
| 1991 : m_pTextPage(pTextPage) {} | |
| 1992 | |
| 1993 CPDF_LinkExtract::~CPDF_LinkExtract() { | |
| 1994 } | |
| 1995 | |
| 1996 void CPDF_LinkExtract::ExtractLinks() { | |
| 1997 m_LinkArray.clear(); | |
| 1998 if (!m_pTextPage->IsParsed()) | |
| 1999 return; | |
| 2000 | |
| 2001 m_strPageText = m_pTextPage->GetPageText(0, -1); | |
| 2002 if (m_strPageText.IsEmpty()) | |
| 2003 return; | |
| 2004 | |
| 2005 ParseLink(); | |
| 2006 } | |
| 2007 | |
| 2008 void CPDF_LinkExtract::ParseLink() { | |
| 2009 int start = 0, pos = 0; | |
| 2010 int TotalChar = m_pTextPage->CountChars(); | |
| 2011 while (pos < TotalChar) { | |
| 2012 FPDF_CHAR_INFO pageChar; | |
| 2013 m_pTextPage->GetCharInfo(pos, &pageChar); | |
| 2014 if (pageChar.m_Flag == FPDFTEXT_CHAR_GENERATED || | |
| 2015 pageChar.m_Unicode == 0x20 || pos == TotalChar - 1) { | |
| 2016 int nCount = pos - start; | |
| 2017 if (pos == TotalChar - 1) { | |
| 2018 nCount++; | |
| 2019 } | |
| 2020 CFX_WideString strBeCheck; | |
| 2021 strBeCheck = m_pTextPage->GetPageText(start, nCount); | |
| 2022 if (strBeCheck.GetLength() > 5) { | |
| 2023 while (strBeCheck.GetLength() > 0) { | |
| 2024 FX_WCHAR ch = strBeCheck.GetAt(strBeCheck.GetLength() - 1); | |
| 2025 if (ch == L')' || ch == L',' || ch == L'>' || ch == L'.') { | |
| 2026 strBeCheck = strBeCheck.Mid(0, strBeCheck.GetLength() - 1); | |
| 2027 nCount--; | |
| 2028 } else { | |
| 2029 break; | |
| 2030 } | |
| 2031 } | |
| 2032 if (nCount > 5 && | |
| 2033 (CheckWebLink(strBeCheck) || CheckMailLink(strBeCheck))) { | |
| 2034 m_LinkArray.push_back({start, nCount, strBeCheck}); | |
| 2035 } | |
| 2036 } | |
| 2037 start = ++pos; | |
| 2038 } else { | |
| 2039 pos++; | |
| 2040 } | |
| 2041 } | |
| 2042 } | |
| 2043 | |
| 2044 bool CPDF_LinkExtract::CheckWebLink(CFX_WideString& strBeCheck) { | |
| 2045 CFX_WideString str = strBeCheck; | |
| 2046 str.MakeLower(); | |
| 2047 if (str.Find(L"http://www.") != -1) { | |
| 2048 strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"http://www.")); | |
| 2049 return true; | |
| 2050 } | |
| 2051 if (str.Find(L"http://") != -1) { | |
| 2052 strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"http://")); | |
| 2053 return true; | |
| 2054 } | |
| 2055 if (str.Find(L"https://www.") != -1) { | |
| 2056 strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"https://www.")); | |
| 2057 return true; | |
| 2058 } | |
| 2059 if (str.Find(L"https://") != -1) { | |
| 2060 strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"https://")); | |
| 2061 return true; | |
| 2062 } | |
| 2063 if (str.Find(L"www.") != -1) { | |
| 2064 strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"www.")); | |
| 2065 strBeCheck = L"http://" + strBeCheck; | |
| 2066 return true; | |
| 2067 } | |
| 2068 return false; | |
| 2069 } | |
| 2070 | |
| 2071 bool CPDF_LinkExtract::CheckMailLink(CFX_WideString& str) { | |
| 2072 int aPos = str.Find(L'@'); | |
| 2073 // Invalid when no '@'. | |
| 2074 if (aPos < 1) | |
| 2075 return false; | |
| 2076 | |
| 2077 // Check the local part. | |
| 2078 int pPos = aPos; // Used to track the position of '@' or '.'. | |
| 2079 for (int i = aPos - 1; i >= 0; i--) { | |
| 2080 FX_WCHAR ch = str.GetAt(i); | |
| 2081 if (ch == L'_' || ch == L'-' || FXSYS_iswalnum(ch)) | |
| 2082 continue; | |
| 2083 | |
| 2084 if (ch != L'.' || i == pPos - 1 || i == 0) { | |
| 2085 if (i == aPos - 1) { | |
| 2086 // There is '.' or invalid char before '@'. | |
| 2087 return FALSE; | |
| 2088 } | |
| 2089 // End extracting for other invalid chars, '.' at the beginning, or | |
| 2090 // consecutive '.'. | |
| 2091 int removed_len = i == pPos - 1 ? i + 2 : i + 1; | |
| 2092 str = str.Right(str.GetLength() - removed_len); | |
| 2093 break; | |
| 2094 } | |
| 2095 // Found a valid '.'. | |
| 2096 pPos = i; | |
| 2097 } | |
| 2098 | |
| 2099 // Check the domain name part. | |
| 2100 aPos = str.Find(L'@'); | |
| 2101 if (aPos < 1) | |
| 2102 return false; | |
| 2103 | |
| 2104 str.TrimRight(L'.'); | |
| 2105 // At least one '.' in domain name, but not at the beginning. | |
| 2106 // TODO(weili): RFC5322 allows domain names to be a local name without '.'. | |
| 2107 // Check whether we should remove this check. | |
| 2108 int ePos = str.Find(L'.', aPos + 1); | |
| 2109 if (ePos == -1 || ePos == aPos + 1) | |
| 2110 return false; | |
| 2111 | |
| 2112 // Validate all other chars in domain name. | |
| 2113 int nLen = str.GetLength(); | |
| 2114 pPos = 0; // Used to track the position of '.'. | |
| 2115 for (int i = aPos + 1; i < nLen; i++) { | |
| 2116 FX_WCHAR wch = str.GetAt(i); | |
| 2117 if (wch == L'-' || FXSYS_iswalnum(wch)) | |
| 2118 continue; | |
| 2119 | |
| 2120 if (wch != L'.' || i == pPos + 1) { | |
| 2121 // Domain name should end before invalid char. | |
| 2122 int host_end = i == pPos + 1 ? i - 2 : i - 1; | |
| 2123 if (pPos > 0 && host_end - aPos >= 3) { | |
| 2124 // Trim the ending invalid chars if there is at least one '.' and name. | |
| 2125 str = str.Left(host_end + 1); | |
| 2126 break; | |
| 2127 } | |
| 2128 return false; | |
| 2129 } | |
| 2130 pPos = i; | |
| 2131 } | |
| 2132 | |
| 2133 if (str.Find(L"mailto:") == -1) | |
| 2134 str = L"mailto:" + str; | |
| 2135 | |
| 2136 return true; | |
| 2137 } | |
| 2138 | |
| 2139 CFX_WideString CPDF_LinkExtract::GetURL(size_t index) const { | |
| 2140 return index < m_LinkArray.size() ? m_LinkArray[index].m_strUrl : L""; | |
| 2141 } | |
| 2142 | |
| 2143 std::vector<CFX_FloatRect> CPDF_LinkExtract::GetRects(size_t index) const { | |
| 2144 if (index >= m_LinkArray.size()) | |
| 2145 return std::vector<CFX_FloatRect>(); | |
| 2146 | |
| 2147 return m_pTextPage->GetRectArray(m_LinkArray[index].m_Start, | |
| 2148 m_LinkArray[index].m_Count); | |
| 2149 } | |
| OLD | NEW |