core/ difference with XFA (for information only).

core/src/fpdftext/fpdf_text_int.cpp

 // Copyright 2014 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 <ctype.h>
 #include <algorithm>
 
 #include "../../../third_party/base/nonstd_unique_ptr.h"
@@ skipping 63 matching lines @@
   }
   return baseSpace;
 }
 
 }  // namespace
 
 CPDFText_ParseOptions::CPDFText_ParseOptions()
     : m_bGetCharCodeOnly(FALSE),
       m_bNormalizeObjs(TRUE),
       m_bOutputHyphen(FALSE) {}
+#ifndef PDF_ENABLE_XFA
 
+#else
+IPDF_TextPage* IPDF_TextPage::CreateTextPage(
+    const CPDF_Page* pPage,
+    CPDFText_ParseOptions ParserOptions) {
+  return new CPDF_TextPage(pPage, ParserOptions);
+}
+#endif
 IPDF_TextPage* IPDF_TextPage::CreateTextPage(const CPDF_Page* pPage,
                                              int flags) {
   return new CPDF_TextPage(pPage, flags);
 }
+#ifndef PDF_ENABLE_XFA
 
+#else
+IPDF_TextPage* IPDF_TextPage::CreateTextPage(const CPDF_PageObjects* pObjs,
+                                             int flags) {
+  return new CPDF_TextPage(pObjs, flags);
+}
+#endif
 IPDF_TextPageFind* IPDF_TextPageFind::CreatePageFind(
     const IPDF_TextPage* pTextPage) {
   return pTextPage ? new CPDF_TextPageFind(pTextPage) : nullptr;
 }
+#ifndef PDF_ENABLE_XFA
 
+#endif
 IPDF_LinkExtract* IPDF_LinkExtract::CreateLinkExtract() {
   return new CPDF_LinkExtract();
 }
+#ifndef PDF_ENABLE_XFA
 
+#endif
 #define TEXT_BLANK_CHAR L' '
 #define TEXT_LINEFEED_CHAR L'\n'
 #define TEXT_RETURN_CHAR L'\r'
 #define TEXT_EMPTY L""
 #define TEXT_BLANK L" "
 #define TEXT_RETURN_LINEFEED L"\r\n"
 #define TEXT_LINEFEED L"\n"
 #define TEXT_CHARRATIO_GAPDELTA 0.070
 
 CPDF_TextPage::CPDF_TextPage(const CPDF_Page* pPage, int flags)
     : m_pPage(pPage),
       m_charList(512),
       m_TempCharList(50),
       m_parserflag(flags),
       m_pPreTextObj(nullptr),
       m_bIsParsed(false),
       m_TextlineDir(-1),
       m_CurlineRect(0, 0, 0, 0) {
   m_TextBuf.EstimateSize(0, 10240);
   pPage->GetDisplayMatrix(m_DisplayMatrix, 0, 0, (int)pPage->GetPageWidth(),
                           (int)pPage->GetPageHeight(), 0);
 }
 
+#ifdef PDF_ENABLE_XFA
+CPDF_TextPage::CPDF_TextPage(const CPDF_Page* pPage,
+                             CPDFText_ParseOptions ParserOptions)
+    : m_ParseOptions(ParserOptions),
+      m_pPage(pPage),
+      m_charList(512),
+      m_TempCharList(50),
+      m_parserflag(0),
+      m_pPreTextObj(nullptr),
+      m_bIsParsed(false),
+      m_TextlineDir(-1),
+      m_CurlineRect(0, 0, 0, 0) {
+  m_TextBuf.EstimateSize(0, 10240);
+  pPage->GetDisplayMatrix(m_DisplayMatrix, 0, 0, (int)pPage->GetPageWidth(),
+                          (int)pPage->GetPageHeight(), 0);
+}
+
+CPDF_TextPage::CPDF_TextPage(const CPDF_PageObjects* pPage, int flags)
+    : m_pPage(pPage),
+      m_charList(512),
+      m_TempCharList(50),
+      m_parserflag(flags),
+      m_pPreTextObj(nullptr),
+      m_bIsParsed(false),
+      m_TextlineDir(-1),
+      m_CurlineRect(0, 0, 0, 0) {
+  m_TextBuf.EstimateSize(0, 10240);
+  CFX_FloatRect pageRect = pPage->CalcBoundingBox();
+  m_DisplayMatrix = CFX_AffineMatrix(1, 0, 0, -1, pageRect.right, pageRect.top);
+}
+#endif
 void CPDF_TextPage::NormalizeObjects(FX_BOOL bNormalize) {
   m_ParseOptions.m_bNormalizeObjs = bNormalize;
 }
 bool CPDF_TextPage::IsControlChar(const PAGECHAR_INFO& charInfo) {
   switch (charInfo.m_Unicode) {
     case 0x2:
     case 0x3:
     case 0x93:
     case 0x94:
     case 0x96:
@@ skipping 1217 matching lines @@
   int nContentMark = pMarkData->CountItems();
   if (nContentMark < 1) {
     return;
   }
   CFX_WideString actText;
   CPDF_Dictionary* pDict = NULL;
   int n = 0;
   for (n = 0; n < nContentMark; n++) {
     CPDF_ContentMarkItem& item = pMarkData->GetItem(n);
     CFX_ByteString tagStr = (CFX_ByteString)item.GetName();
+#ifdef PDF_ENABLE_XFA
+
+#endif
     pDict = ToDictionary(static_cast<CPDF_Object*>(item.GetParam()));
     CPDF_String* temp =
         ToString(pDict ? pDict->GetElement(FX_BSTRC("ActualText")) : nullptr);
     if (temp) {
       actText = temp->GetUnicodeText();
     }
   }
   FX_STRSIZE nItems = actText.GetLength();
   if (nItems < 1) {
     return;
@@ skipping 148 matching lines @@
     ProcessMarkedContent(Obj);
     m_pPreTextObj = pTextObj;
     m_perMatrix.Copy(formMatrix);
     return;
   }
   m_pPreTextObj = pTextObj;
   m_perMatrix.Copy(formMatrix);
   int nItems = pTextObj->CountItems();
   FX_FLOAT baseSpace = _CalculateBaseSpace(pTextObj, matrix);
 
+#ifndef PDF_ENABLE_XFA
   const FX_BOOL bR2L = IsRightToLeft(pTextObj, pFont, nItems);
   const FX_BOOL bIsBidiAndMirrorInverse =
+#else
+  FX_BOOL bIsBidiAndMirrosInverse = FALSE;

[Lei Zhang, 2015/10/30 06:15:30]

Weird that this file has this discrepancy.

+  CFX_BidiChar* BidiChar = new CFX_BidiChar;
+  int32_t nR2L = 0;
+  int32_t nL2R = 0;
+  int32_t start = 0, count = 0;
+  CPDF_TextObjectItem item;
+  for (int32_t i = 0; i < nItems; i++) {
+    pTextObj->GetItemInfo(i, &item);
+    if (item.m_CharCode == (FX_DWORD)-1) {
+      continue;
+    }
+    CFX_WideString wstrItem = pFont->UnicodeFromCharCode(item.m_CharCode);
+    FX_WCHAR wChar = wstrItem.GetAt(0);
+    if ((wstrItem.IsEmpty() || wChar == 0) && item.m_CharCode) {
+      wChar = (FX_WCHAR)item.m_CharCode;
+    }
+    if (!wChar) {
+      continue;
+    }
+    if (BidiChar && BidiChar->AppendChar(wChar)) {
+      CFX_BidiChar::Direction ret = BidiChar->GetBidiInfo(&start, &count);
+      if (ret == CFX_BidiChar::RIGHT) {
+        nR2L++;
+      } else if (ret == CFX_BidiChar::LEFT) {
+        nL2R++;
+      }
+    }
+  }
+  if (BidiChar && BidiChar->EndChar()) {
+    CFX_BidiChar::Direction ret = BidiChar->GetBidiInfo(&start, &count);
+    if (ret == CFX_BidiChar::RIGHT) {
+      nR2L++;
+    } else if (ret == CFX_BidiChar::LEFT) {
+      nL2R++;
+    }
+  }
+  FX_BOOL bR2L = FALSE;
+  if (nR2L > 0 && nR2L >= nL2R) {
+    bR2L = TRUE;
+  }
+  bIsBidiAndMirrosInverse =
+#endif
       bR2L && (matrix.a * matrix.d - matrix.b * matrix.c) < 0;
   int32_t iBufStartAppend = m_TempTextBuf.GetLength();
   int32_t iCharListStartAppend = m_TempCharList.GetSize();
 
   FX_FLOAT spacing = 0;
   for (int i = 0; i < nItems; i++) {
     CPDF_TextObjectItem item;
     PAGECHAR_INFO charinfo;
     charinfo.m_OriginX = 0;
     charinfo.m_OriginY = 0;
@@ skipping 138 matching lines @@
       } else if (i == 0) {
         CFX_WideString str = m_TempTextBuf.GetWideString();
         if (!str.IsEmpty() &&
             str.GetAt(str.GetLength() - 1) == TEXT_BLANK_CHAR) {
           m_TempTextBuf.Delete(m_TempTextBuf.GetLength() - 1, 1);
           m_TempCharList.Delete(m_TempCharList.GetSize() - 1);
         }
       }
     }
   }
+#ifndef PDF_ENABLE_XFA
   if (bIsBidiAndMirrorInverse) {
     SwapTempTextBuf(iCharListStartAppend, iBufStartAppend);
   }
 }
 void CPDF_TextPage::SwapTempTextBuf(int32_t iCharListStartAppend,
                                     int32_t iBufStartAppend) {
   int32_t i, j;
   i = iCharListStartAppend;
   j = m_TempCharList.GetSize() - 1;
   for (; i < j; i++, j--) {
@@ skipping 28 matching lines @@
     if (!wChar) {
       continue;
     }
     if (pBidiChar->AppendChar(wChar)) {
       CFX_BidiChar::Direction ret = pBidiChar->GetBidiInfo(&start, &count);
       if (ret == CFX_BidiChar::RIGHT) {
         nR2L++;
       } else if (ret == CFX_BidiChar::LEFT) {
         nL2R++;
       }
+#else
+  if (bIsBidiAndMirrosInverse) {
+    int32_t i, j;
+    i = iCharListStartAppend;
+    j = m_TempCharList.GetSize() - 1;
+    for (; i < j; i++, j--) {
+      std::swap(m_TempCharList[i], m_TempCharList[j]);
+      std::swap(m_TempCharList[i].m_Index, m_TempCharList[j].m_Index);
+#endif
     }
+#ifndef PDF_ENABLE_XFA
   }
   if (pBidiChar->EndChar()) {
     CFX_BidiChar::Direction ret = pBidiChar->GetBidiInfo(&start, &count);
     if (ret == CFX_BidiChar::RIGHT) {
       nR2L++;
     } else if (ret == CFX_BidiChar::LEFT) {
       nL2R++;
+#else
+    FX_WCHAR* pTempBuffer = m_TempTextBuf.GetBuffer();
+    i = iBufStartAppend;
+    j = m_TempTextBuf.GetLength() - 1;
+    for (; i < j; i++, j--) {
+      std::swap(pTempBuffer[i], pTempBuffer[j]);
+#endif
     }
   }
+#ifndef PDF_ENABLE_XFA
   return (nR2L > 0 && nR2L >= nL2R);
+#endif
 }
 int32_t CPDF_TextPage::GetTextObjectWritingMode(
     const CPDF_TextObject* pTextObj) {
   int32_t nChars = pTextObj->CountChars();
   if (nChars == 1) {
     return m_TextlineDir;
   }
   CPDF_TextObjectItem first, last;
   pTextObj->GetCharInfo(0, &first);
   pTextObj->GetCharInfo(nChars - 1, &last);
@@ skipping 970 matching lines @@
   if (!m_bIsParsed || index < 0 || index >= m_LinkList.GetSize()) {
     return;
   }
   CPDF_LinkExt* link = NULL;
   link = m_LinkList.GetAt(index);
   if (!link) {
     return;
   }
   m_pTextPage->GetRectArray(link->m_Start, link->m_Count, rects);
 }