More define cleanup.

core/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 <algorithm>
 #include <cctype>
 #include <cwctype>
 #include <memory>
@@ skipping 10 matching lines @@
 #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_RLTB 1
-#define FPDFTEXT_LEFT -1
-#define FPDFTEXT_RIGHT 1

[dsinclair, 2016/05/02 20:59:55]

RLTB and RIGHT were both the same value so I removed RLTB.

[Tom Sepez, 2016/05/02 21:45:39]

Acknowledged.

-
 #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 FPDFTEXT_MC_PASS 0
-#define FPDFTEXT_MC_DONE 1
-#define FPDFTEXT_MC_DELAY 2
+
+#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
 
 namespace {
 
 const FX_FLOAT kDefaultFontSize = 1.0f;
 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) {
@@ skipping 78 matching lines @@
     FX_WCHAR n = wFind;
     while (n--) {
       *pDst++ = *pMap++;
     }
   }
   return (FX_STRSIZE)wFind;
 }
 
 }  // namespace
 
-#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)
+CPDF_TextPage::CPDF_TextPage(const CPDF_Page* pPage, FPDFText_Direction flags)
     : m_pPage(pPage),
       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);
 }
@@ skipping 554 matching lines @@
   }
   if (segmentStatus == 1) {
     segmentStatus = 2;
     m_Segments.Add(segment);
     segment.m_Start = 0;
     segment.m_nCount = 0;
   }
   return m_Segments.GetSize();
 }
 
-int CPDF_TextPage::GetWordBreak(int index, int direction) const {
+int CPDF_TextPage::GetWordBreak(int index, FPDFText_Direction direction) const {
   if (!m_bIsParsed)
     return -1;
 
-  if (direction != FPDFTEXT_LEFT && direction != FPDFTEXT_RIGHT)

[Tom Sepez, 2016/05/02 21:45:39]

given values of -1 and 1, is it possible that they were also silently using 0 as
a "dont't know"? 

[dsinclair, 2016/05/03 13:58:16]

Turns out, GetWordBreak is never called. Removed.

Now, the only time it's checked is to compare it to ::Right. The setter methods
would have previously set 1 or 0, which means ::Left was actually impossible. It
will now set ::Left or ::Right depending on the value set in the viewer
preferences.

-    return -1;
-
   if (index < 0 || index >= pdfium::CollectionSize<int>(m_CharList))
     return -1;
 
   const PAGECHAR_INFO& charinfo = m_CharList[index];
   if (charinfo.m_Index == -1 || charinfo.m_Flag == FPDFTEXT_CHAR_GENERATED) {
     return index;
   }
   if (!IsLetter(charinfo.m_Unicode)) {
     return index;
   }
   int breakPos = index;
-  if (direction == FPDFTEXT_LEFT) {
+  if (direction == FPDFText_Direction::Left) {
     while (--breakPos > 0) {
       if (!IsLetter(m_CharList[breakPos].m_Unicode))
         break;
     }
-  } else if (direction == FPDFTEXT_RIGHT) {
+  } else if (direction == FPDFText_Direction::Right) {
     while (++breakPos < pdfium::CollectionSize<int>(m_CharList)) {
       if (!IsLetter(m_CharList[breakPos].m_Unicode))
         break;
     }
   }
   return breakPos;
 }
 
 int32_t CPDF_TextPage::FindTextlineFlowDirection() {
   const int32_t nPageWidth = static_cast<int32_t>(m_pPage->GetPageWidth());
@@ skipping 225 matching lines @@
     }
     if (bPrevSpace) {
       m_TempTextBuf.Delete(i, 1);
       m_TempCharList.erase(m_TempCharList.begin() + i);
       str.Delete(i);
       i--;
     }
     bPrevSpace = TRUE;
   }
   CFX_BidiString bidi(str);
-  if (m_parserflag == FPDFTEXT_RLTB)
+  if (m_parserflag == FPDFText_Direction::Right)
     bidi.SetOverallDirectionRight();
   CFX_BidiChar::Direction eCurrentDirection = bidi.OverallDirection();
   for (const auto& segment : bidi) {
     if (segment.direction == CFX_BidiChar::RIGHT ||
         (segment.direction == CFX_BidiChar::NEUTRAL &&
          eCurrentDirection == CFX_BidiChar::RIGHT)) {
       eCurrentDirection = CFX_BidiChar::RIGHT;
       for (int m = segment.start + segment.count; m > segment.start; --m)
         AddCharInfoByRLDirection(bidi.CharAt(m - 1), m_TempCharList[m - 1]);
     } else {
@@ skipping 81 matching lines @@
         m_LineObj.InsertAt(i + 1, Obj);
       }
       break;
     }
   }
   if (i < 0) {
     m_LineObj.InsertAt(0, Obj);
   }
 }
 
-int32_t CPDF_TextPage::PreMarkedContent(PDFTEXT_Obj Obj) {
+FPDFText_MarkedContent CPDF_TextPage::PreMarkedContent(PDFTEXT_Obj Obj) {
   CPDF_TextObject* pTextObj = Obj.m_pTextObj;
   CPDF_ContentMarkData* pMarkData =
       (CPDF_ContentMarkData*)pTextObj->m_ContentMark.GetObject();
   if (!pMarkData)
-    return FPDFTEXT_MC_PASS;
+    return FPDFText_MarkedContent::Pass;
 
   int nContentMark = pMarkData->CountItems();
   if (nContentMark < 1)
-    return FPDFTEXT_MC_PASS;
+    return FPDFText_MarkedContent::Pass;
+
   CFX_WideString actText;
   FX_BOOL bExist = FALSE;
   CPDF_Dictionary* pDict = NULL;
   int n = 0;
   for (n = 0; n < nContentMark; n++) {
     CPDF_ContentMarkItem& item = pMarkData->GetItem(n);
     if (item.GetParamType() == CPDF_ContentMarkItem::ParamType::None)
       continue;
     pDict = item.GetParam();
     CPDF_String* temp =
         ToString(pDict ? pDict->GetObjectBy("ActualText") : nullptr);
     if (temp) {
       bExist = TRUE;
       actText = temp->GetUnicodeText();
     }
   }
   if (!bExist)
-    return FPDFTEXT_MC_PASS;
+    return FPDFText_MarkedContent::Pass;
 
   if (m_pPreTextObj) {
     CPDF_ContentMarkData* pPreMarkData =
         (CPDF_ContentMarkData*)m_pPreTextObj->m_ContentMark.GetObject();
     if (pPreMarkData && pPreMarkData->CountItems() == n &&
         pDict == pPreMarkData->GetItem(n - 1).GetParam()) {
-      return FPDFTEXT_MC_DONE;
+      return FPDFText_MarkedContent::Done;
     }
   }
   FX_STRSIZE nItems = actText.GetLength();
   if (nItems < 1)
-    return FPDFTEXT_MC_PASS;
+    return FPDFText_MarkedContent::Pass;
 
   CPDF_Font* pFont = pTextObj->GetFont();
   bExist = FALSE;
   for (FX_STRSIZE i = 0; i < nItems; i++) {
     if (pFont->CharCodeFromUnicode(actText.GetAt(i)) !=
         CPDF_Font::kInvalidCharCode) {
       bExist = TRUE;
       break;
     }
   }
   if (!bExist)
-    return FPDFTEXT_MC_PASS;
+    return FPDFText_MarkedContent::Pass;
 
   bExist = FALSE;
   for (FX_STRSIZE i = 0; i < nItems; i++) {
     FX_WCHAR wChar = actText.GetAt(i);
     if ((wChar > 0x80 && wChar < 0xFFFD) || (wChar <= 0x80 && isprint(wChar))) {
       bExist = TRUE;
       break;
     }
   }
   if (!bExist)
-    return FPDFTEXT_MC_DONE;
+    return FPDFText_MarkedContent::Done;
 
-  return FPDFTEXT_MC_DELAY;
+  return FPDFText_MarkedContent::Delay;
 }
 
 void CPDF_TextPage::ProcessMarkedContent(PDFTEXT_Obj Obj) {
   CPDF_TextObject* pTextObj = Obj.m_pTextObj;
   CPDF_ContentMarkData* pMarkData =
       (CPDF_ContentMarkData*)pTextObj->m_ContentMark.GetObject();
   if (!pMarkData)
     return;
 
   int nContentMark = pMarkData->CountItems();
@@ skipping 102 matching lines @@
 void CPDF_TextPage::ProcessTextObject(PDFTEXT_Obj Obj) {
   CPDF_TextObject* pTextObj = Obj.m_pTextObj;
   if (FXSYS_fabs(pTextObj->m_Right - pTextObj->m_Left) < 0.01f) {
     return;
   }
   CFX_Matrix formMatrix = Obj.m_formMatrix;
   CPDF_Font* pFont = pTextObj->GetFont();
   CFX_Matrix matrix;
   pTextObj->GetTextMatrix(&matrix);
   matrix.Concat(formMatrix);
-  int32_t bPreMKC = PreMarkedContent(Obj);
-  if (FPDFTEXT_MC_DONE == bPreMKC) {
+  FPDFText_MarkedContent bPreMKC = PreMarkedContent(Obj);

[Tom Sepez, 2016/05/02 21:45:39]

prolly shoujld be ePreMKC since its an enum not a bool.

[dsinclair, 2016/05/03 13:58:16]

Sigh, hungarian notation, so verbose.

[Tom Sepez, 2016/05/04 20:31:13]

Actually, it points out a problem that when the type of a variable changes, so
must its name, which makes generic programming difficult and why its a bad idea
in general.

+  if (bPreMKC == FPDFText_MarkedContent::Done) {
     m_pPreTextObj = pTextObj;
     m_perMatrix.Copy(formMatrix);
     return;
   }
   int result = 0;
   if (m_pPreTextObj) {
     result = ProcessInsertObject(pTextObj, formMatrix);
     if (2 == result) {
       m_CurlineRect =
           CFX_FloatRect(Obj.m_pTextObj->m_Left, Obj.m_pTextObj->m_Bottom,
@@ skipping 55 matching lines @@
       m_TempTextBuf.Delete(m_TempTextBuf.GetLength() - 1, 1);
       charinfo->m_Unicode = 0x2;
       charinfo->m_Flag = FPDFTEXT_CHAR_HYPHEN;
       m_TempTextBuf.AppendChar(0xfffe);
     }
   } else {
     m_CurlineRect =
         CFX_FloatRect(Obj.m_pTextObj->m_Left, Obj.m_pTextObj->m_Bottom,
                       Obj.m_pTextObj->m_Right, Obj.m_pTextObj->m_Top);
   }
-  if (FPDFTEXT_MC_DELAY == bPreMKC) {
+  if (bPreMKC == FPDFText_MarkedContent::Delay) {
     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);
 
@@ skipping 1066 matching lines @@
   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);
 }