Use string pools in some dictionaries

core/fpdfapi/fpdf_parser/cpdf_syntax_parser.cpp

 // 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/fpdfapi/fpdf_parser/cpdf_syntax_parser.h"
 
 #include <vector>
 
@@ skipping 19 matching lines @@
   CFX_ByteStringC m_bsTag;
   FX_STRSIZE m_Offset;
 };
 
 }  // namespace
 
 // static
 int CPDF_SyntaxParser::s_CurrentRecursionDepth = 0;
 
 CPDF_SyntaxParser::CPDF_SyntaxParser()
     : m_MetadataObjnum(0),

[Lei Zhang, 2016/09/20 23:48:23]

Use a delegate ctor?

: CPDF_SyntaxParser(nullptr) {}

[Tom Sepez, 2016/09/27 22:36:59]

done

       m_pFileAccess(nullptr),
       m_pFileBuf(nullptr),
-      m_BufSize(CPDF_ModuleMgr::kFileBufSize) {}
+      m_BufSize(CPDF_ModuleMgr::kFileBufSize),
+      m_pPool(nullptr) {}
+
+CPDF_SyntaxParser::CPDF_SyntaxParser(CFX_ByteStringPool* pPool)
+    : m_MetadataObjnum(0),
+      m_pFileAccess(nullptr),
+      m_pFileBuf(nullptr),
+      m_BufSize(CPDF_ModuleMgr::kFileBufSize),
+      m_pPool(pPool) {}
 
 CPDF_SyntaxParser::~CPDF_SyntaxParser() {
   FX_Free(m_pFileBuf);
 }
 
 FX_BOOL CPDF_SyntaxParser::GetCharAt(FX_FILESIZE pos, uint8_t& ch) {
   CFX_AutoRestorer<FX_FILESIZE> save_pos(&m_Pos);
   m_Pos = pos;
   return GetNextChar(ch);
 }
@@ skipping 348 matching lines @@
   if (word == "true" || word == "false")
     return new CPDF_Boolean(word == "true");
 
   if (word == "null")
     return new CPDF_Null;
 
   if (word == "(") {
     CFX_ByteString str = ReadString();
     if (m_pCryptoHandler && bDecrypt)
       m_pCryptoHandler->Decrypt(objnum, gennum, str);
-    return new CPDF_String(str, FALSE);
+    return new CPDF_String(MaybeIntern(str), FALSE);
   }
 
   if (word == "<") {
     CFX_ByteString str = ReadHexString();
     if (m_pCryptoHandler && bDecrypt)
       m_pCryptoHandler->Decrypt(objnum, gennum, str);
-
-    return new CPDF_String(str, TRUE);
+    return new CPDF_String(MaybeIntern(str), TRUE);
   }
 
   if (word == "[") {
     CPDF_Array* pArray = new CPDF_Array;
     while (CPDF_Object* pObj = GetObject(pObjList, objnum, gennum, true))
       pArray->Add(pObj);
 
     return pArray;
   }
 
   if (word[0] == '/') {
-    return new CPDF_Name(
-        PDF_NameDecode(CFX_ByteStringC(m_WordBuffer + 1, m_WordSize - 1)));
+    return new CPDF_Name(MaybeIntern(
+        PDF_NameDecode(CFX_ByteStringC(m_WordBuffer + 1, m_WordSize - 1))));
   }
 
   if (word == "<<") {
     int32_t nKeys = 0;
     FX_FILESIZE dwSignValuePos = 0;
 
     std::unique_ptr<CPDF_Dictionary, ReleaseDeleter<CPDF_Dictionary>> pDict(
-        new CPDF_Dictionary);
+        new CPDF_Dictionary(m_pPool));
     while (1) {
       CFX_ByteString key = GetNextWord(nullptr);
       if (key.IsEmpty())
         return nullptr;
 
       FX_FILESIZE SavedPos = m_Pos - key.GetLength();
       if (key == ">>")
         break;
 
       if (key == "endobj") {
@@ skipping 72 matching lines @@
   if (word == "true" || word == "false")
     return new CPDF_Boolean(word == "true");
 
   if (word == "null")
     return new CPDF_Null;
 
   if (word == "(") {
     CFX_ByteString str = ReadString();
     if (m_pCryptoHandler)
       m_pCryptoHandler->Decrypt(objnum, gennum, str);
-    return new CPDF_String(str, FALSE);
+    return new CPDF_String(MaybeIntern(str), FALSE);
   }
 
   if (word == "<") {
     CFX_ByteString str = ReadHexString();
     if (m_pCryptoHandler)
       m_pCryptoHandler->Decrypt(objnum, gennum, str);
-    return new CPDF_String(str, TRUE);
+    return new CPDF_String(MaybeIntern(str), TRUE);
   }
 
   if (word == "[") {
     std::unique_ptr<CPDF_Array, ReleaseDeleter<CPDF_Array>> pArray(
         new CPDF_Array);
     while (CPDF_Object* pObj = GetObject(pObjList, objnum, gennum, true))
       pArray->Add(pObj);
 
     return m_WordBuffer[0] == ']' ? pArray.release() : nullptr;
   }
 
   if (word[0] == '/') {
-    return new CPDF_Name(
-        PDF_NameDecode(CFX_ByteStringC(m_WordBuffer + 1, m_WordSize - 1)));
+    return new CPDF_Name(MaybeIntern(
+        PDF_NameDecode(CFX_ByteStringC(m_WordBuffer + 1, m_WordSize - 1))));
   }
 
   if (word == "<<") {
     std::unique_ptr<CPDF_Dictionary, ReleaseDeleter<CPDF_Dictionary>> pDict(
-        new CPDF_Dictionary);
+        new CPDF_Dictionary(m_pPool));
     while (1) {
       FX_FILESIZE SavedPos = m_Pos;
       CFX_ByteString key = GetNextWord(nullptr);
       if (key.IsEmpty())
         return nullptr;
 
       if (key == ">>")
         break;
 
       if (key == "endobj") {
@@ skipping 409 matching lines @@
     if (limit && m_Pos == limit)
       return -1;
   }
   return -1;
 }
 
 void CPDF_SyntaxParser::SetEncrypt(
     std::unique_ptr<CPDF_CryptoHandler> pCryptoHandler) {
   m_pCryptoHandler = std::move(pCryptoHandler);
 }
+
+CFX_ByteString CPDF_SyntaxParser::MaybeIntern(const CFX_ByteString& str) {
+  return m_pPool ? m_pPool->Intern(str) : str;
+}