Revert "Fix the issue that pdfium swallows 'fi' or 'ff' in some pdf files"

This reverts commit 9bd18183ba8210c91d71c3060146235750a4c71c.

3 files changed, 36 insertions, 2 deletions

diff --git a/core/src/fpdfapi/fpdf_page/fpdf_page_parser.cpp b/core/src/fpdfapi/fpdf_page/fpdf_page_parser.cpp
index 4e5ef1c898..059dd4c2a6 100644
--- a/core/src/fpdfapi/fpdf_page/fpdf_page_parser.cpp
+++ b/core/src/fpdfapi/fpdf_page/fpdf_page_parser.cpp
@@ -11,6 +11,7 @@
 
 #define REQUIRE_PARAMS(count)  \
   if (m_ParamCount != count) { \
+    m_bAbort = TRUE;           \
     return;                    \
   }
 
@@ -33,6 +34,7 @@ CPDF_StreamContentParser::CPDF_StreamContentParser(
       m_Level(level),
       m_ParamStartPos(0),
       m_ParamCount(0),
+      m_bAbort(FALSE),
       m_pCurStates(new CPDF_AllStates),
       m_pLastTextObject(nullptr),
       m_DefFontSize(0),
diff --git a/core/src/fpdfapi/fpdf_page/fpdf_page_parser_old.cpp b/core/src/fpdfapi/fpdf_page/fpdf_page_parser_old.cpp
index c9bcff6db6..48e9b98d3a 100644
--- a/core/src/fpdfapi/fpdf_page/fpdf_page_parser_old.cpp
+++ b/core/src/fpdfapi/fpdf_page/fpdf_page_parser_old.cpp
@@ -9,7 +9,27 @@
 #include "../../../include/fxcodec/fx_codec.h"
 #include "pageint.h"
 #include <limits.h>
-
+const FX_CHAR* const _PDF_OpCharType =
+    "IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII"
+    "IIVIIIIVIIVIIIIIVVIIIIIIIIIIIIII"
+    "IIVVVVVVIVVVVVVIVVVVVIIVVIIIIIII"
+    "IIVVVVVVVVVVVVVVIVVVIIVVIVVIIIII"
+    "IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII"
+    "IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII"
+    "IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII"
+    "IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII";
+FX_BOOL _PDF_HasInvalidOpChar(const FX_CHAR* op) {
+  if (!op) {
+    return FALSE;
+  }
+  uint8_t ch;
+  while ((ch = *op++)) {
+    if (_PDF_OpCharType[ch] == 'I') {
+      return TRUE;
+    }
+  }
+  return FALSE;
+}
 class CPDF_StreamParserAutoClearer {
  public:
   CPDF_StreamParserAutoClearer(CPDF_StreamParser** scoped_variable,
@@ -41,7 +61,13 @@ FX_DWORD CPDF_StreamContentParser::Parse(const uint8_t* pData,
       case CPDF_StreamParser::EndOfData:
         return m_pSyntax->GetPos();
       case CPDF_StreamParser::Keyword:
-        OnOperator((char*)syntax.GetWordBuf());
+        if (!OnOperator((char*)syntax.GetWordBuf()) &&
+            _PDF_HasInvalidOpChar((char*)syntax.GetWordBuf())) {
+          m_bAbort = TRUE;
+        }
+        if (m_bAbort) {
+          return m_pSyntax->GetPos();
+        }
         ClearAllParams();
         break;
       case CPDF_StreamParser::Number:
@@ -1100,6 +1126,10 @@ void CPDF_ContentParser::Continue(IFX_Pause* pPause) {
         m_CurrentOffset +=
             m_pParser->Parse(m_pData + m_CurrentOffset,
                              m_Size - m_CurrentOffset, PARSE_STEP_LIMIT);
+        if (m_pParser->ShouldAbort()) {
+          m_InternalStage = PAGEPARSE_STAGE_CHECKCLIP;
+          continue;
+        }
       }
     }
     if (m_InternalStage == PAGEPARSE_STAGE_CHECKCLIP) {
diff --git a/core/src/fpdfapi/fpdf_page/pageint.h b/core/src/fpdfapi/fpdf_page/pageint.h
index c85523b833..6bec07268c 100644
--- a/core/src/fpdfapi/fpdf_page/pageint.h
+++ b/core/src/fpdfapi/fpdf_page/pageint.h
@@ -162,6 +162,7 @@ class CPDF_StreamContentParser {
                            int level);
   ~CPDF_StreamContentParser();
 
+  FX_BOOL ShouldAbort() const { return m_bAbort; }
   CPDF_PageObjects* GetObjectList() const { return m_pObjectList; }
   CPDF_AllStates* GetCurStates() const { return m_pCurStates.get(); }
   FX_BOOL IsColored() const { return m_bColored; }
@@ -304,6 +305,7 @@ class CPDF_StreamContentParser {
   _ContentParam m_ParamBuf1[PARAM_BUF_SIZE];
   FX_DWORD m_ParamStartPos;
   FX_DWORD m_ParamCount;
+  FX_BOOL m_bAbort;
   CPDF_StreamParser* m_pSyntax;
   nonstd::unique_ptr<CPDF_AllStates> m_pCurStates;
   CPDF_ContentMark m_CurContentMark;