diff options
| -rw-r--r-- | BUILD.gn | 4 | ||||
| -rw-r--r-- | core/fpdftext/cpdf_linkextract.cpp | 173 | ||||
| -rw-r--r-- | core/fpdftext/cpdf_textpage.cpp (renamed from core/fpdftext/fpdf_text_int.cpp) | 604 | ||||
| -rw-r--r-- | core/fpdftext/cpdf_textpagefind.cpp | 407 | ||||
| -rw-r--r-- | core/fpdftext/include/cpdf_textpage.h | 20 |
5 files changed, 605 insertions, 603 deletions
@@ -538,7 +538,9 @@ static_library("fpdfapi") { static_library("fpdftext") { sources = [ - "core/fpdftext/fpdf_text_int.cpp", + "core/fpdftext/cpdf_linkextract.cpp", + "core/fpdftext/cpdf_textpage.cpp", + "core/fpdftext/cpdf_textpagefind.cpp", "core/fpdftext/include/cpdf_linkextract.h", "core/fpdftext/include/cpdf_textpage.h", "core/fpdftext/include/cpdf_textpagefind.h", diff --git a/core/fpdftext/cpdf_linkextract.cpp b/core/fpdftext/cpdf_linkextract.cpp new file mode 100644 index 0000000000..1677b67b55 --- /dev/null +++ b/core/fpdftext/cpdf_linkextract.cpp @@ -0,0 +1,173 @@ +// Copyright 2016 PDFium Authors. All rights reserved. +// Use of this source code is governed by a BSD-style license that can be +// found in the LICENSE file. + +// Original code copyright 2014 Foxit Software Inc. http://www.foxitsoftware.com + +#include "core/fpdftext/include/cpdf_linkextract.h" + +#include <vector> + +#include "core/fpdftext/include/cpdf_textpage.h" +#include "core/fxcrt/include/fx_ext.h" +#include "core/fxcrt/include/fx_string.h" +#include "core/fxcrt/include/fx_system.h" + +CPDF_LinkExtract::CPDF_LinkExtract(const CPDF_TextPage* pTextPage) + : m_pTextPage(pTextPage) {} + +CPDF_LinkExtract::~CPDF_LinkExtract() {} + +void CPDF_LinkExtract::ExtractLinks() { + m_LinkArray.clear(); + if (!m_pTextPage->IsParsed()) + return; + + m_strPageText = m_pTextPage->GetPageText(0, -1); + if (m_strPageText.IsEmpty()) + return; + + ParseLink(); +} + +void CPDF_LinkExtract::ParseLink() { + int start = 0, pos = 0; + int TotalChar = m_pTextPage->CountChars(); + while (pos < TotalChar) { + FPDF_CHAR_INFO pageChar; + m_pTextPage->GetCharInfo(pos, &pageChar); + if (pageChar.m_Flag == FPDFTEXT_CHAR_GENERATED || + pageChar.m_Unicode == 0x20 || pos == TotalChar - 1) { + int nCount = pos - start; + if (pos == TotalChar - 1) + nCount++; + CFX_WideString strBeCheck; + strBeCheck = m_pTextPage->GetPageText(start, nCount); + if (strBeCheck.GetLength() > 5) { + while (strBeCheck.GetLength() > 0) { + FX_WCHAR ch = strBeCheck.GetAt(strBeCheck.GetLength() - 1); + if (ch == L')' || ch == L',' || ch == L'>' || ch == L'.') { + strBeCheck = strBeCheck.Mid(0, strBeCheck.GetLength() - 1); + nCount--; + } else { + break; + } + } + if (nCount > 5 && + (CheckWebLink(strBeCheck) || CheckMailLink(strBeCheck))) { + m_LinkArray.push_back({start, nCount, strBeCheck}); + } + } + start = ++pos; + } else { + pos++; + } + } +} + +bool CPDF_LinkExtract::CheckWebLink(CFX_WideString& strBeCheck) { + CFX_WideString str = strBeCheck; + str.MakeLower(); + if (str.Find(L"http://www.") != -1) { + strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"http://www.")); + return true; + } + if (str.Find(L"http://") != -1) { + strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"http://")); + return true; + } + if (str.Find(L"https://www.") != -1) { + strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"https://www.")); + return true; + } + if (str.Find(L"https://") != -1) { + strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"https://")); + return true; + } + if (str.Find(L"www.") != -1) { + strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"www.")); + strBeCheck = L"http://" + strBeCheck; + return true; + } + return false; +} + +bool CPDF_LinkExtract::CheckMailLink(CFX_WideString& str) { + int aPos = str.Find(L'@'); + // Invalid when no '@'. + if (aPos < 1) + return false; + + // Check the local part. + int pPos = aPos; // Used to track the position of '@' or '.'. + for (int i = aPos - 1; i >= 0; i--) { + FX_WCHAR ch = str.GetAt(i); + if (ch == L'_' || ch == L'-' || FXSYS_iswalnum(ch)) + continue; + + if (ch != L'.' || i == pPos - 1 || i == 0) { + if (i == aPos - 1) { + // There is '.' or invalid char before '@'. + return FALSE; + } + // End extracting for other invalid chars, '.' at the beginning, or + // consecutive '.'. + int removed_len = i == pPos - 1 ? i + 2 : i + 1; + str = str.Right(str.GetLength() - removed_len); + break; + } + // Found a valid '.'. + pPos = i; + } + + // Check the domain name part. + aPos = str.Find(L'@'); + if (aPos < 1) + return false; + + str.TrimRight(L'.'); + // At least one '.' in domain name, but not at the beginning. + // TODO(weili): RFC5322 allows domain names to be a local name without '.'. + // Check whether we should remove this check. + int ePos = str.Find(L'.', aPos + 1); + if (ePos == -1 || ePos == aPos + 1) + return false; + + // Validate all other chars in domain name. + int nLen = str.GetLength(); + pPos = 0; // Used to track the position of '.'. + for (int i = aPos + 1; i < nLen; i++) { + FX_WCHAR wch = str.GetAt(i); + if (wch == L'-' || FXSYS_iswalnum(wch)) + continue; + + if (wch != L'.' || i == pPos + 1) { + // Domain name should end before invalid char. + int host_end = i == pPos + 1 ? i - 2 : i - 1; + if (pPos > 0 && host_end - aPos >= 3) { + // Trim the ending invalid chars if there is at least one '.' and name. + str = str.Left(host_end + 1); + break; + } + return false; + } + pPos = i; + } + + if (str.Find(L"mailto:") == -1) + str = L"mailto:" + str; + + return true; +} + +CFX_WideString CPDF_LinkExtract::GetURL(size_t index) const { + return index < m_LinkArray.size() ? m_LinkArray[index].m_strUrl : L""; +} + +std::vector<CFX_FloatRect> CPDF_LinkExtract::GetRects(size_t index) const { + if (index >= m_LinkArray.size()) + return std::vector<CFX_FloatRect>(); + + return m_pTextPage->GetRectArray(m_LinkArray[index].m_Start, + m_LinkArray[index].m_Count); +} diff --git a/core/fpdftext/fpdf_text_int.cpp b/core/fpdftext/cpdf_textpage.cpp index fbd9c9c8c1..3981cfee40 100644 --- a/core/fpdftext/fpdf_text_int.cpp +++ b/core/fpdftext/cpdf_textpage.cpp @@ -4,10 +4,9 @@ // Original code copyright 2014 Foxit Software Inc. http://www.foxitsoftware.com +#include "core/fpdftext/include/cpdf_textpage.h" + #include <algorithm> -#include <cctype> -#include <cwctype> -#include <memory> #include <utility> #include <vector> @@ -19,35 +18,12 @@ #include "core/fpdfapi/fpdf_page/include/cpdf_textobject.h" #include "core/fpdfapi/fpdf_parser/include/cpdf_dictionary.h" #include "core/fpdfapi/fpdf_parser/include/cpdf_string.h" -#include "core/fpdftext/include/cpdf_linkextract.h" -#include "core/fpdftext/include/cpdf_textpage.h" -#include "core/fpdftext/include/cpdf_textpagefind.h" #include "core/fpdftext/unicodenormalizationdata.h" #include "core/fxcrt/fx_bidi.h" #include "core/fxcrt/include/fx_ext.h" #include "core/fxcrt/include/fx_ucd.h" #include "third_party/base/stl_util.h" -#define FPDFTEXT_MATCHCASE 0x00000001 -#define FPDFTEXT_MATCHWHOLEWORD 0x00000002 -#define FPDFTEXT_CONSECUTIVE 0x00000004 - -#define FPDFTEXT_CHAR_ERROR -1 -#define FPDFTEXT_CHAR_NORMAL 0 -#define FPDFTEXT_CHAR_GENERATED 1 -#define FPDFTEXT_CHAR_UNUNICODE 2 -#define FPDFTEXT_CHAR_HYPHEN 3 -#define FPDFTEXT_CHAR_PIECE 4 - -#define TEXT_SPACE_CHAR L' ' -#define TEXT_LINEFEED_CHAR L'\n' -#define TEXT_RETURN_CHAR L'\r' -#define TEXT_EMPTY L"" -#define TEXT_SPACE L" " -#define TEXT_RETURN_LINEFEED L"\r\n" -#define TEXT_LINEFEED L"\n" -#define TEXT_CHARRATIO_GAPDELTA 0.070 - namespace { const FX_FLOAT kDefaultFontSize = 1.0f; @@ -55,22 +31,6 @@ const uint16_t* const g_UnicodeData_Normalization_Maps[5] = { nullptr, g_UnicodeData_Normalization_Map1, g_UnicodeData_Normalization_Map2, g_UnicodeData_Normalization_Map3, g_UnicodeData_Normalization_Map4}; -FX_BOOL IsIgnoreSpaceCharacter(FX_WCHAR curChar) { - if (curChar < 255) - return FALSE; - if ((curChar >= 0x0600 && curChar <= 0x06FF) || - (curChar >= 0xFE70 && curChar <= 0xFEFF) || - (curChar >= 0xFB50 && curChar <= 0xFDFF) || - (curChar >= 0x0400 && curChar <= 0x04FF) || - (curChar >= 0x0500 && curChar <= 0x052F) || - (curChar >= 0xA640 && curChar <= 0xA69F) || - (curChar >= 0x2DE0 && curChar <= 0x2DFF) || curChar == 8467 || - (curChar >= 0x2000 && curChar <= 0x206F)) { - return FALSE; - } - return TRUE; -} - FX_FLOAT NormalizeThreshold(FX_FLOAT threshold) { if (threshold < 300) return threshold / 2.0f; @@ -1587,563 +1547,3 @@ FX_BOOL CPDF_TextPage::IsRectIntersect(const CFX_FloatRect& rect1, rect.Intersect(rect2); return !rect.IsEmpty(); } - -CPDF_TextPageFind::CPDF_TextPageFind(const CPDF_TextPage* pTextPage) - : m_pTextPage(pTextPage), - m_flags(0), - m_findNextStart(-1), - m_findPreStart(-1), - m_bMatchCase(FALSE), - m_bMatchWholeWord(FALSE), - m_resStart(0), - m_resEnd(-1), - m_IsFind(FALSE) { - m_strText = m_pTextPage->GetPageText(); - int nCount = pTextPage->CountChars(); - if (nCount) { - m_CharIndex.push_back(0); - } - for (int i = 0; i < nCount; i++) { - FPDF_CHAR_INFO info; - pTextPage->GetCharInfo(i, &info); - int indexSize = pdfium::CollectionSize<int>(m_CharIndex); - if (info.m_Flag == FPDFTEXT_CHAR_NORMAL || - info.m_Flag == FPDFTEXT_CHAR_GENERATED) { - if (indexSize % 2) { - m_CharIndex.push_back(1); - } else { - if (indexSize <= 0) { - continue; - } - m_CharIndex[indexSize - 1] += 1; - } - } else { - if (indexSize % 2) { - if (indexSize <= 0) { - continue; - } - m_CharIndex[indexSize - 1] = i + 1; - } else { - m_CharIndex.push_back(i + 1); - } - } - } - int indexSize = pdfium::CollectionSize<int>(m_CharIndex); - if (indexSize % 2) { - m_CharIndex.erase(m_CharIndex.begin() + indexSize - 1); - } -} - -CPDF_TextPageFind::~CPDF_TextPageFind() {} - -int CPDF_TextPageFind::GetCharIndex(int index) const { - return m_pTextPage->CharIndexFromTextIndex(index); -} - -FX_BOOL CPDF_TextPageFind::FindFirst(const CFX_WideString& findwhat, - int flags, - int startPos) { - if (!m_pTextPage) { - return FALSE; - } - if (m_strText.IsEmpty() || m_bMatchCase != (flags & FPDFTEXT_MATCHCASE)) { - m_strText = m_pTextPage->GetPageText(); - } - CFX_WideString findwhatStr = findwhat; - m_findWhat = findwhatStr; - m_flags = flags; - m_bMatchCase = flags & FPDFTEXT_MATCHCASE; - if (m_strText.IsEmpty()) { - m_IsFind = FALSE; - return TRUE; - } - FX_STRSIZE len = findwhatStr.GetLength(); - if (!m_bMatchCase) { - findwhatStr.MakeLower(); - m_strText.MakeLower(); - } - m_bMatchWholeWord = flags & FPDFTEXT_MATCHWHOLEWORD; - m_findNextStart = startPos; - if (startPos == -1) { - m_findPreStart = m_strText.GetLength() - 1; - } else { - m_findPreStart = startPos; - } - m_csFindWhatArray.clear(); - int i = 0; - while (i < len) { - if (findwhatStr.GetAt(i) != ' ') { - break; - } - i++; - } - if (i < len) { - ExtractFindWhat(findwhatStr); - } else { - m_csFindWhatArray.push_back(findwhatStr); - } - if (m_csFindWhatArray.empty()) { - return FALSE; - } - m_IsFind = TRUE; - m_resStart = 0; - m_resEnd = -1; - return TRUE; -} - -FX_BOOL CPDF_TextPageFind::FindNext() { - if (!m_pTextPage) { - return FALSE; - } - m_resArray.clear(); - if (m_findNextStart == -1) { - return FALSE; - } - if (m_strText.IsEmpty()) { - m_IsFind = FALSE; - return m_IsFind; - } - int strLen = m_strText.GetLength(); - if (m_findNextStart > strLen - 1) { - m_IsFind = FALSE; - return m_IsFind; - } - int nCount = pdfium::CollectionSize<int>(m_csFindWhatArray); - int nResultPos = 0; - int nStartPos = 0; - nStartPos = m_findNextStart; - FX_BOOL bSpaceStart = FALSE; - for (int iWord = 0; iWord < nCount; iWord++) { - CFX_WideString csWord = m_csFindWhatArray[iWord]; - if (csWord.IsEmpty()) { - if (iWord == nCount - 1) { - FX_WCHAR strInsert = m_strText.GetAt(nStartPos); - if (strInsert == TEXT_LINEFEED_CHAR || strInsert == TEXT_SPACE_CHAR || - strInsert == TEXT_RETURN_CHAR || strInsert == 160) { - nResultPos = nStartPos + 1; - break; - } - iWord = -1; - } else if (iWord == 0) { - bSpaceStart = TRUE; - } - continue; - } - int endIndex; - nResultPos = m_strText.Find(csWord.c_str(), nStartPos); - if (nResultPos == -1) { - m_IsFind = FALSE; - return m_IsFind; - } - endIndex = nResultPos + csWord.GetLength() - 1; - if (iWord == 0) { - m_resStart = nResultPos; - } - FX_BOOL bMatch = TRUE; - if (iWord != 0 && !bSpaceStart) { - int PreResEndPos = nStartPos; - int curChar = csWord.GetAt(0); - CFX_WideString lastWord = m_csFindWhatArray[iWord - 1]; - int lastChar = lastWord.GetAt(lastWord.GetLength() - 1); - if (nStartPos == nResultPos && - !(IsIgnoreSpaceCharacter(lastChar) || - IsIgnoreSpaceCharacter(curChar))) { - bMatch = FALSE; - } - for (int d = PreResEndPos; d < nResultPos; d++) { - FX_WCHAR strInsert = m_strText.GetAt(d); - if (strInsert != TEXT_LINEFEED_CHAR && strInsert != TEXT_SPACE_CHAR && - strInsert != TEXT_RETURN_CHAR && strInsert != 160) { - bMatch = FALSE; - break; - } - } - } else if (bSpaceStart) { - if (nResultPos > 0) { - FX_WCHAR strInsert = m_strText.GetAt(nResultPos - 1); - if (strInsert != TEXT_LINEFEED_CHAR && strInsert != TEXT_SPACE_CHAR && - strInsert != TEXT_RETURN_CHAR && strInsert != 160) { - bMatch = FALSE; - m_resStart = nResultPos; - } else { - m_resStart = nResultPos - 1; - } - } - } - if (m_bMatchWholeWord && bMatch) { - bMatch = IsMatchWholeWord(m_strText, nResultPos, endIndex); - } - nStartPos = endIndex + 1; - if (!bMatch) { - iWord = -1; - if (bSpaceStart) { - nStartPos = m_resStart + m_csFindWhatArray[1].GetLength(); - } else { - nStartPos = m_resStart + m_csFindWhatArray[0].GetLength(); - } - } - } - m_resEnd = nResultPos + m_csFindWhatArray.back().GetLength() - 1; - m_IsFind = TRUE; - int resStart = GetCharIndex(m_resStart); - int resEnd = GetCharIndex(m_resEnd); - m_resArray = m_pTextPage->GetRectArray(resStart, resEnd - resStart + 1); - if (m_flags & FPDFTEXT_CONSECUTIVE) { - m_findNextStart = m_resStart + 1; - m_findPreStart = m_resEnd - 1; - } else { - m_findNextStart = m_resEnd + 1; - m_findPreStart = m_resStart - 1; - } - return m_IsFind; -} - -FX_BOOL CPDF_TextPageFind::FindPrev() { - if (!m_pTextPage) { - return FALSE; - } - m_resArray.clear(); - if (m_strText.IsEmpty() || m_findPreStart < 0) { - m_IsFind = FALSE; - return m_IsFind; - } - CPDF_TextPageFind findEngine(m_pTextPage); - FX_BOOL ret = findEngine.FindFirst(m_findWhat, m_flags); - if (!ret) { - m_IsFind = FALSE; - return m_IsFind; - } - int order = -1, MatchedCount = 0; - while (ret) { - ret = findEngine.FindNext(); - if (ret) { - int order1 = findEngine.GetCurOrder(); - int MatchedCount1 = findEngine.GetMatchedCount(); - if (((order1 + MatchedCount1) - 1) > m_findPreStart) { - break; - } - order = order1; - MatchedCount = MatchedCount1; - } - } - if (order == -1) { - m_IsFind = FALSE; - return m_IsFind; - } - m_resStart = m_pTextPage->TextIndexFromCharIndex(order); - m_resEnd = m_pTextPage->TextIndexFromCharIndex(order + MatchedCount - 1); - m_IsFind = TRUE; - m_resArray = m_pTextPage->GetRectArray(order, MatchedCount); - if (m_flags & FPDFTEXT_CONSECUTIVE) { - m_findNextStart = m_resStart + 1; - m_findPreStart = m_resEnd - 1; - } else { - m_findNextStart = m_resEnd + 1; - m_findPreStart = m_resStart - 1; - } - return m_IsFind; -} - -void CPDF_TextPageFind::ExtractFindWhat(const CFX_WideString& findwhat) { - if (findwhat.IsEmpty()) { - return; - } - int index = 0; - while (1) { - CFX_WideString csWord = TEXT_EMPTY; - int ret = - ExtractSubString(csWord, findwhat.c_str(), index, TEXT_SPACE_CHAR); - if (csWord.IsEmpty()) { - if (ret) { - m_csFindWhatArray.push_back(L""); - index++; - continue; - } else { - break; - } - } - int pos = 0; - while (pos < csWord.GetLength()) { - CFX_WideString curStr = csWord.Mid(pos, 1); - FX_WCHAR curChar = csWord.GetAt(pos); - if (IsIgnoreSpaceCharacter(curChar)) { - if (pos > 0 && curChar == 0x2019) { - pos++; - continue; - } - if (pos > 0) { - m_csFindWhatArray.push_back(csWord.Mid(0, pos)); - } - m_csFindWhatArray.push_back(curStr); - if (pos == csWord.GetLength() - 1) { - csWord.clear(); - break; - } - csWord = csWord.Right(csWord.GetLength() - pos - 1); - pos = 0; - continue; - } - pos++; - } - if (!csWord.IsEmpty()) { - m_csFindWhatArray.push_back(csWord); - } - index++; - } -} - -FX_BOOL CPDF_TextPageFind::IsMatchWholeWord(const CFX_WideString& csPageText, - int startPos, - int endPos) { - FX_WCHAR char_left = 0; - FX_WCHAR char_right = 0; - int char_count = endPos - startPos + 1; - if (char_count < 1) { - return FALSE; - } - if (char_count == 1 && csPageText.GetAt(startPos) > 255) { - return TRUE; - } - if (startPos - 1 >= 0) { - char_left = csPageText.GetAt(startPos - 1); - } - if (startPos + char_count < csPageText.GetLength()) { - char_right = csPageText.GetAt(startPos + char_count); - } - if ((char_left > 'A' && char_left < 'a') || - (char_left > 'a' && char_left < 'z') || - (char_left > 0xfb00 && char_left < 0xfb06) || std::iswdigit(char_left) || - (char_right > 'A' && char_right < 'a') || - (char_right > 'a' && char_right < 'z') || - (char_right > 0xfb00 && char_right < 0xfb06) || - std::iswdigit(char_right)) { - return FALSE; - } - if (!(('A' > char_left || char_left > 'Z') && - ('a' > char_left || char_left > 'z') && - ('A' > char_right || char_right > 'Z') && - ('a' > char_right || char_right > 'z'))) { - return FALSE; - } - if (char_count > 0) { - if (csPageText.GetAt(startPos) >= L'0' && - csPageText.GetAt(startPos) <= L'9' && char_left >= L'0' && - char_left <= L'9') { - return FALSE; - } - if (csPageText.GetAt(endPos) >= L'0' && csPageText.GetAt(endPos) <= L'9' && - char_right >= L'0' && char_right <= L'9') { - return FALSE; - } - } - return TRUE; -} - -FX_BOOL CPDF_TextPageFind::ExtractSubString(CFX_WideString& rString, - const FX_WCHAR* lpszFullString, - int iSubString, - FX_WCHAR chSep) { - if (!lpszFullString) { - return FALSE; - } - while (iSubString--) { - lpszFullString = wcschr(lpszFullString, chSep); - if (!lpszFullString) { - rString.clear(); - return FALSE; - } - lpszFullString++; - while (*lpszFullString == chSep) { - lpszFullString++; - } - } - const FX_WCHAR* lpchEnd = wcschr(lpszFullString, chSep); - int nLen = lpchEnd ? (int)(lpchEnd - lpszFullString) - : (int)FXSYS_wcslen(lpszFullString); - ASSERT(nLen >= 0); - FXSYS_memcpy(rString.GetBuffer(nLen), lpszFullString, - nLen * sizeof(FX_WCHAR)); - rString.ReleaseBuffer(); - return TRUE; -} - -CFX_WideString CPDF_TextPageFind::MakeReverse(const CFX_WideString& str) { - CFX_WideString str2; - str2.clear(); - int nlen = str.GetLength(); - for (int i = nlen - 1; i >= 0; i--) { - str2 += str.GetAt(i); - } - return str2; -} - -int CPDF_TextPageFind::GetCurOrder() const { - return GetCharIndex(m_resStart); -} - -int CPDF_TextPageFind::GetMatchedCount() const { - int resStart = GetCharIndex(m_resStart); - int resEnd = GetCharIndex(m_resEnd); - return resEnd - resStart + 1; -} - -CPDF_LinkExtract::CPDF_LinkExtract(const CPDF_TextPage* pTextPage) - : m_pTextPage(pTextPage) {} - -CPDF_LinkExtract::~CPDF_LinkExtract() { -} - -void CPDF_LinkExtract::ExtractLinks() { - m_LinkArray.clear(); - if (!m_pTextPage->IsParsed()) - return; - - m_strPageText = m_pTextPage->GetPageText(0, -1); - if (m_strPageText.IsEmpty()) - return; - - ParseLink(); -} - -void CPDF_LinkExtract::ParseLink() { - int start = 0, pos = 0; - int TotalChar = m_pTextPage->CountChars(); - while (pos < TotalChar) { - FPDF_CHAR_INFO pageChar; - m_pTextPage->GetCharInfo(pos, &pageChar); - if (pageChar.m_Flag == FPDFTEXT_CHAR_GENERATED || - pageChar.m_Unicode == 0x20 || pos == TotalChar - 1) { - int nCount = pos - start; - if (pos == TotalChar - 1) { - nCount++; - } - CFX_WideString strBeCheck; - strBeCheck = m_pTextPage->GetPageText(start, nCount); - if (strBeCheck.GetLength() > 5) { - while (strBeCheck.GetLength() > 0) { - FX_WCHAR ch = strBeCheck.GetAt(strBeCheck.GetLength() - 1); - if (ch == L')' || ch == L',' || ch == L'>' || ch == L'.') { - strBeCheck = strBeCheck.Mid(0, strBeCheck.GetLength() - 1); - nCount--; - } else { - break; - } - } - if (nCount > 5 && - (CheckWebLink(strBeCheck) || CheckMailLink(strBeCheck))) { - m_LinkArray.push_back({start, nCount, strBeCheck}); - } - } - start = ++pos; - } else { - pos++; - } - } -} - -bool CPDF_LinkExtract::CheckWebLink(CFX_WideString& strBeCheck) { - CFX_WideString str = strBeCheck; - str.MakeLower(); - if (str.Find(L"http://www.") != -1) { - strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"http://www.")); - return true; - } - if (str.Find(L"http://") != -1) { - strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"http://")); - return true; - } - if (str.Find(L"https://www.") != -1) { - strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"https://www.")); - return true; - } - if (str.Find(L"https://") != -1) { - strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"https://")); - return true; - } - if (str.Find(L"www.") != -1) { - strBeCheck = strBeCheck.Right(str.GetLength() - str.Find(L"www.")); - strBeCheck = L"http://" + strBeCheck; - return true; - } - return false; -} - -bool CPDF_LinkExtract::CheckMailLink(CFX_WideString& str) { - int aPos = str.Find(L'@'); - // Invalid when no '@'. - if (aPos < 1) - return false; - - // Check the local part. - int pPos = aPos; // Used to track the position of '@' or '.'. - for (int i = aPos - 1; i >= 0; i--) { - FX_WCHAR ch = str.GetAt(i); - if (ch == L'_' || ch == L'-' || FXSYS_iswalnum(ch)) - continue; - - if (ch != L'.' || i == pPos - 1 || i == 0) { - if (i == aPos - 1) { - // There is '.' or invalid char before '@'. - return FALSE; - } - // End extracting for other invalid chars, '.' at the beginning, or - // consecutive '.'. - int removed_len = i == pPos - 1 ? i + 2 : i + 1; - str = str.Right(str.GetLength() - removed_len); - break; - } - // Found a valid '.'. - pPos = i; - } - - // Check the domain name part. - aPos = str.Find(L'@'); - if (aPos < 1) - return false; - - str.TrimRight(L'.'); - // At least one '.' in domain name, but not at the beginning. - // TODO(weili): RFC5322 allows domain names to be a local name without '.'. - // Check whether we should remove this check. - int ePos = str.Find(L'.', aPos + 1); - if (ePos == -1 || ePos == aPos + 1) - return false; - - // Validate all other chars in domain name. - int nLen = str.GetLength(); - pPos = 0; // Used to track the position of '.'. - for (int i = aPos + 1; i < nLen; i++) { - FX_WCHAR wch = str.GetAt(i); - if (wch == L'-' || FXSYS_iswalnum(wch)) - continue; - - if (wch != L'.' || i == pPos + 1) { - // Domain name should end before invalid char. - int host_end = i == pPos + 1 ? i - 2 : i - 1; - if (pPos > 0 && host_end - aPos >= 3) { - // Trim the ending invalid chars if there is at least one '.' and name. - str = str.Left(host_end + 1); - break; - } - return false; - } - pPos = i; - } - - if (str.Find(L"mailto:") == -1) - str = L"mailto:" + str; - - return true; -} - -CFX_WideString CPDF_LinkExtract::GetURL(size_t index) const { - return index < m_LinkArray.size() ? m_LinkArray[index].m_strUrl : L""; -} - -std::vector<CFX_FloatRect> CPDF_LinkExtract::GetRects(size_t index) const { - if (index >= m_LinkArray.size()) - return std::vector<CFX_FloatRect>(); - - return m_pTextPage->GetRectArray(m_LinkArray[index].m_Start, - m_LinkArray[index].m_Count); -} diff --git a/core/fpdftext/cpdf_textpagefind.cpp b/core/fpdftext/cpdf_textpagefind.cpp new file mode 100644 index 0000000000..a67bdf15d5 --- /dev/null +++ b/core/fpdftext/cpdf_textpagefind.cpp @@ -0,0 +1,407 @@ +// Copyright 2016 PDFium Authors. All rights reserved. +// Use of this source code is governed by a BSD-style license that can be +// found in the LICENSE file. + +// Original code copyright 2014 Foxit Software Inc. http://www.foxitsoftware.com + +#include "core/fpdftext/include/cpdf_textpagefind.h" + +#include <cwchar> +#include <cwctype> +#include <vector> + +#include "core/fpdftext/include/cpdf_textpage.h" +#include "core/fxcrt/include/fx_string.h" +#include "core/fxcrt/include/fx_system.h" +#include "third_party/base/stl_util.h" + +namespace { + +FX_BOOL IsIgnoreSpaceCharacter(FX_WCHAR curChar) { + if (curChar < 255 || (curChar >= 0x0600 && curChar <= 0x06FF) || + (curChar >= 0xFE70 && curChar <= 0xFEFF) || + (curChar >= 0xFB50 && curChar <= 0xFDFF) || + (curChar >= 0x0400 && curChar <= 0x04FF) || + (curChar >= 0x0500 && curChar <= 0x052F) || + (curChar >= 0xA640 && curChar <= 0xA69F) || + (curChar >= 0x2DE0 && curChar <= 0x2DFF) || curChar == 8467 || + (curChar >= 0x2000 && curChar <= 0x206F)) { + return FALSE; + } + return TRUE; +} + +} // namespace + +CPDF_TextPageFind::CPDF_TextPageFind(const CPDF_TextPage* pTextPage) + : m_pTextPage(pTextPage), + m_flags(0), + m_findNextStart(-1), + m_findPreStart(-1), + m_bMatchCase(FALSE), + m_bMatchWholeWord(FALSE), + m_resStart(0), + m_resEnd(-1), + m_IsFind(FALSE) { + m_strText = m_pTextPage->GetPageText(); + int nCount = pTextPage->CountChars(); + if (nCount) + m_CharIndex.push_back(0); + for (int i = 0; i < nCount; i++) { + FPDF_CHAR_INFO info; + pTextPage->GetCharInfo(i, &info); + int indexSize = pdfium::CollectionSize<int>(m_CharIndex); + if (info.m_Flag == FPDFTEXT_CHAR_NORMAL || + info.m_Flag == FPDFTEXT_CHAR_GENERATED) { + if (indexSize % 2) { + m_CharIndex.push_back(1); + } else { + if (indexSize <= 0) + continue; + m_CharIndex[indexSize - 1] += 1; + } + } else { + if (indexSize % 2) { + if (indexSize <= 0) + continue; + m_CharIndex[indexSize - 1] = i + 1; + } else { + m_CharIndex.push_back(i + 1); + } + } + } + int indexSize = pdfium::CollectionSize<int>(m_CharIndex); + if (indexSize % 2) + m_CharIndex.erase(m_CharIndex.begin() + indexSize - 1); +} + +CPDF_TextPageFind::~CPDF_TextPageFind() {} + +int CPDF_TextPageFind::GetCharIndex(int index) const { + return m_pTextPage->CharIndexFromTextIndex(index); +} + +FX_BOOL CPDF_TextPageFind::FindFirst(const CFX_WideString& findwhat, + int flags, + int startPos) { + if (!m_pTextPage) + return FALSE; + if (m_strText.IsEmpty() || m_bMatchCase != (flags & FPDFTEXT_MATCHCASE)) + m_strText = m_pTextPage->GetPageText(); + CFX_WideString findwhatStr = findwhat; + m_findWhat = findwhatStr; + m_flags = flags; + m_bMatchCase = flags & FPDFTEXT_MATCHCASE; + if (m_strText.IsEmpty()) { + m_IsFind = FALSE; + return TRUE; + } + FX_STRSIZE len = findwhatStr.GetLength(); + if (!m_bMatchCase) { + findwhatStr.MakeLower(); + m_strText.MakeLower(); + } + m_bMatchWholeWord = flags & FPDFTEXT_MATCHWHOLEWORD; + m_findNextStart = startPos; + if (startPos == -1) + m_findPreStart = m_strText.GetLength() - 1; + else + m_findPreStart = startPos; + m_csFindWhatArray.clear(); + int i = 0; + while (i < len) { + if (findwhatStr.GetAt(i) != ' ') + break; + i++; + } + if (i < len) + ExtractFindWhat(findwhatStr); + else + m_csFindWhatArray.push_back(findwhatStr); + if (m_csFindWhatArray.empty()) + return FALSE; + m_IsFind = TRUE; + m_resStart = 0; + m_resEnd = -1; + return TRUE; +} + +FX_BOOL CPDF_TextPageFind::FindNext() { + if (!m_pTextPage) + return FALSE; + m_resArray.clear(); + if (m_findNextStart == -1) + return FALSE; + if (m_strText.IsEmpty()) { + m_IsFind = FALSE; + return m_IsFind; + } + int strLen = m_strText.GetLength(); + if (m_findNextStart > strLen - 1) { + m_IsFind = FALSE; + return m_IsFind; + } + int nCount = pdfium::CollectionSize<int>(m_csFindWhatArray); + int nResultPos = 0; + int nStartPos = 0; + nStartPos = m_findNextStart; + bool bSpaceStart = false; + for (int iWord = 0; iWord < nCount; iWord++) { + CFX_WideString csWord = m_csFindWhatArray[iWord]; + if (csWord.IsEmpty()) { + if (iWord == nCount - 1) { + FX_WCHAR strInsert = m_strText.GetAt(nStartPos); + if (strInsert == TEXT_LINEFEED_CHAR || strInsert == TEXT_SPACE_CHAR || + strInsert == TEXT_RETURN_CHAR || strInsert == 160) { + nResultPos = nStartPos + 1; + break; + } + iWord = -1; + } else if (iWord == 0) { + bSpaceStart = true; + } + continue; + } + int endIndex; + nResultPos = m_strText.Find(csWord.c_str(), nStartPos); + if (nResultPos == -1) { + m_IsFind = FALSE; + return m_IsFind; + } + endIndex = nResultPos + csWord.GetLength() - 1; + if (iWord == 0) + m_resStart = nResultPos; + FX_BOOL bMatch = TRUE; + if (iWord != 0 && !bSpaceStart) { + int PreResEndPos = nStartPos; + int curChar = csWord.GetAt(0); + CFX_WideString lastWord = m_csFindWhatArray[iWord - 1]; + int lastChar = lastWord.GetAt(lastWord.GetLength() - 1); + if (nStartPos == nResultPos && + !(IsIgnoreSpaceCharacter(lastChar) || + IsIgnoreSpaceCharacter(curChar))) { + bMatch = FALSE; + } + for (int d = PreResEndPos; d < nResultPos; d++) { + FX_WCHAR strInsert = m_strText.GetAt(d); + if (strInsert != TEXT_LINEFEED_CHAR && strInsert != TEXT_SPACE_CHAR && + strInsert != TEXT_RETURN_CHAR && strInsert != 160) { + bMatch = FALSE; + break; + } + } + } else if (bSpaceStart) { + if (nResultPos > 0) { + FX_WCHAR strInsert = m_strText.GetAt(nResultPos - 1); + if (strInsert != TEXT_LINEFEED_CHAR && strInsert != TEXT_SPACE_CHAR && + strInsert != TEXT_RETURN_CHAR && strInsert != 160) { + bMatch = FALSE; + m_resStart = nResultPos; + } else { + m_resStart = nResultPos - 1; + } + } + } + if (m_bMatchWholeWord && bMatch) { + bMatch = IsMatchWholeWord(m_strText, nResultPos, endIndex); + } + nStartPos = endIndex + 1; + if (!bMatch) { + iWord = -1; + if (bSpaceStart) + nStartPos = m_resStart + m_csFindWhatArray[1].GetLength(); + else + nStartPos = m_resStart + m_csFindWhatArray[0].GetLength(); + } + } + m_resEnd = nResultPos + m_csFindWhatArray.back().GetLength() - 1; + m_IsFind = TRUE; + int resStart = GetCharIndex(m_resStart); + int resEnd = GetCharIndex(m_resEnd); + m_resArray = m_pTextPage->GetRectArray(resStart, resEnd - resStart + 1); + if (m_flags & FPDFTEXT_CONSECUTIVE) { + m_findNextStart = m_resStart + 1; + m_findPreStart = m_resEnd - 1; + } else { + m_findNextStart = m_resEnd + 1; + m_findPreStart = m_resStart - 1; + } + return m_IsFind; +} + +FX_BOOL CPDF_TextPageFind::FindPrev() { + if (!m_pTextPage) + return FALSE; + m_resArray.clear(); + if (m_strText.IsEmpty() || m_findPreStart < 0) { + m_IsFind = FALSE; + return m_IsFind; + } + CPDF_TextPageFind findEngine(m_pTextPage); + FX_BOOL ret = findEngine.FindFirst(m_findWhat, m_flags); + if (!ret) { + m_IsFind = FALSE; + return m_IsFind; + } + int order = -1, MatchedCount = 0; + while (ret) { + ret = findEngine.FindNext(); + if (ret) { + int order1 = findEngine.GetCurOrder(); + int MatchedCount1 = findEngine.GetMatchedCount(); + if (((order1 + MatchedCount1) - 1) > m_findPreStart) + break; + order = order1; + MatchedCount = MatchedCount1; + } + } + if (order == -1) { + m_IsFind = FALSE; + return m_IsFind; + } + m_resStart = m_pTextPage->TextIndexFromCharIndex(order); + m_resEnd = m_pTextPage->TextIndexFromCharIndex(order + MatchedCount - 1); + m_IsFind = TRUE; + m_resArray = m_pTextPage->GetRectArray(order, MatchedCount); + if (m_flags & FPDFTEXT_CONSECUTIVE) { + m_findNextStart = m_resStart + 1; + m_findPreStart = m_resEnd - 1; + } else { + m_findNextStart = m_resEnd + 1; + m_findPreStart = m_resStart - 1; + } + return m_IsFind; +} + +void CPDF_TextPageFind::ExtractFindWhat(const CFX_WideString& findwhat) { + if (findwhat.IsEmpty()) + return; + int index = 0; + while (1) { + CFX_WideString csWord = TEXT_EMPTY; + int ret = + ExtractSubString(csWord, findwhat.c_str(), index, TEXT_SPACE_CHAR); + if (csWord.IsEmpty()) { + if (ret) { + m_csFindWhatArray.push_back(L""); + index++; + continue; + } else { + break; + } + } + int pos = 0; + while (pos < csWord.GetLength()) { + CFX_WideString curStr = csWord.Mid(pos, 1); + FX_WCHAR curChar = csWord.GetAt(pos); + if (IsIgnoreSpaceCharacter(curChar)) { + if (pos > 0 && curChar == 0x2019) { + pos++; + continue; + } + if (pos > 0) + m_csFindWhatArray.push_back(csWord.Mid(0, pos)); + m_csFindWhatArray.push_back(curStr); + if (pos == csWord.GetLength() - 1) { + csWord.clear(); + break; + } + csWord = csWord.Right(csWord.GetLength() - pos - 1); + pos = 0; + continue; + } + pos++; + } + if (!csWord.IsEmpty()) + m_csFindWhatArray.push_back(csWord); + index++; + } +} + +FX_BOOL CPDF_TextPageFind::IsMatchWholeWord(const CFX_WideString& csPageText, + int startPos, + int endPos) { + FX_WCHAR char_left = 0; + FX_WCHAR char_right = 0; + int char_count = endPos - startPos + 1; + if (char_count < 1) + return FALSE; + if (char_count == 1 && csPageText.GetAt(startPos) > 255) + return TRUE; + if (startPos - 1 >= 0) + char_left = csPageText.GetAt(startPos - 1); + if (startPos + char_count < csPageText.GetLength()) + char_right = csPageText.GetAt(startPos + char_count); + if ((char_left > 'A' && char_left < 'a') || + (char_left > 'a' && char_left < 'z') || + (char_left > 0xfb00 && char_left < 0xfb06) || std::iswdigit(char_left) || + (char_right > 'A' && char_right < 'a') || + (char_right > 'a' && char_right < 'z') || + (char_right > 0xfb00 && char_right < 0xfb06) || + std::iswdigit(char_right)) { + return FALSE; + } + if (!(('A' > char_left || char_left > 'Z') && + ('a' > char_left || char_left > 'z') && + ('A' > char_right || char_right > 'Z') && + ('a' > char_right || char_right > 'z'))) { + return FALSE; + } + if (char_count > 0) { + if (csPageText.GetAt(startPos) >= L'0' && + csPageText.GetAt(startPos) <= L'9' && char_left >= L'0' && + char_left <= L'9') { + return FALSE; + } + if (csPageText.GetAt(endPos) >= L'0' && csPageText.GetAt(endPos) <= L'9' && + char_right >= L'0' && char_right <= L'9') { + return FALSE; + } + } + return TRUE; +} + +FX_BOOL CPDF_TextPageFind::ExtractSubString(CFX_WideString& rString, + const FX_WCHAR* lpszFullString, + int iSubString, + FX_WCHAR chSep) { + if (!lpszFullString) + return FALSE; + while (iSubString--) { + lpszFullString = std::wcschr(lpszFullString, chSep); + if (!lpszFullString) { + rString.clear(); + return FALSE; + } + lpszFullString++; + while (*lpszFullString == chSep) + lpszFullString++; + } + const FX_WCHAR* lpchEnd = std::wcschr(lpszFullString, chSep); + int nLen = lpchEnd ? (int)(lpchEnd - lpszFullString) + : (int)FXSYS_wcslen(lpszFullString); + ASSERT(nLen >= 0); + FXSYS_memcpy(rString.GetBuffer(nLen), lpszFullString, + nLen * sizeof(FX_WCHAR)); + rString.ReleaseBuffer(); + return TRUE; +} + +CFX_WideString CPDF_TextPageFind::MakeReverse(const CFX_WideString& str) { + CFX_WideString str2; + str2.clear(); + int nlen = str.GetLength(); + for (int i = nlen - 1; i >= 0; i--) + str2 += str.GetAt(i); + return str2; +} + +int CPDF_TextPageFind::GetCurOrder() const { + return GetCharIndex(m_resStart); +} + +int CPDF_TextPageFind::GetMatchedCount() const { + int resStart = GetCharIndex(m_resStart); + int resEnd = GetCharIndex(m_resEnd); + return resEnd - resStart + 1; +} diff --git a/core/fpdftext/include/cpdf_textpage.h b/core/fpdftext/include/cpdf_textpage.h index 7f25fd7cf5..64b1613e7e 100644 --- a/core/fpdftext/include/cpdf_textpage.h +++ b/core/fpdftext/include/cpdf_textpage.h @@ -20,6 +20,26 @@ class CPDF_FormObject; class CPDF_Page; class CPDF_TextObject; +#define FPDFTEXT_MATCHCASE 0x00000001 +#define FPDFTEXT_MATCHWHOLEWORD 0x00000002 +#define FPDFTEXT_CONSECUTIVE 0x00000004 + +#define FPDFTEXT_CHAR_ERROR -1 +#define FPDFTEXT_CHAR_NORMAL 0 +#define FPDFTEXT_CHAR_GENERATED 1 +#define FPDFTEXT_CHAR_UNUNICODE 2 +#define FPDFTEXT_CHAR_HYPHEN 3 +#define FPDFTEXT_CHAR_PIECE 4 + +#define TEXT_SPACE_CHAR L' ' +#define TEXT_LINEFEED_CHAR L'\n' +#define TEXT_RETURN_CHAR L'\r' +#define TEXT_EMPTY L"" +#define TEXT_SPACE L" " +#define TEXT_RETURN_LINEFEED L"\r\n" +#define TEXT_LINEFEED L"\n" +#define TEXT_CHARRATIO_GAPDELTA 0.070 + enum class FPDFText_MarkedContent { Pass = 0, Done, Delay }; enum class FPDFText_Direction { Left = -1, Right = 1 }; |