Fix loading page using hint tables.chromium/2897

When linearized document have hint table,
The FPDFAvail_IsPageAvail return true, but
FPDF_LoadPage return nullptr, for non first pages.

This happens, bacause document not use hint tables, to load page.

To fix this, I force save the page's ObjNum in document.

R=npm, dsinclair

Review-Url: https://chromiumcodereview.appspot.com/2437773003

4 files changed, 56 insertions, 2 deletions

diff --git a/BUILD.gn b/BUILD.gn
index 1af2c28c55..c9f85fa0ed 100644
--- a/BUILD.gn
+++ b/BUILD.gn
@@ -1634,6 +1634,7 @@ test("pdfium_unittests") {
     "core/fpdfapi/page/fpdf_page_parser_old_unittest.cpp",
     "core/fpdfapi/page/fpdf_page_parser_unittest.cpp",
     "core/fpdfapi/parser/cpdf_array_unittest.cpp",
+    "core/fpdfapi/parser/cpdf_document_unittest.cpp",
     "core/fpdfapi/parser/cpdf_object_unittest.cpp",
     "core/fpdfapi/parser/cpdf_parser_unittest.cpp",
     "core/fpdfapi/parser/cpdf_simple_parser_unittest.cpp",
diff --git a/core/fpdfapi/parser/cpdf_data_avail.cpp b/core/fpdfapi/parser/cpdf_data_avail.cpp
index e23aa8ecfb..e52d0611f9 100644
--- a/core/fpdfapi/parser/cpdf_data_avail.cpp
+++ b/core/fpdfapi/parser/cpdf_data_avail.cpp
@@ -1625,6 +1625,8 @@ CPDF_DataAvail::DocAvailStatus CPDF_DataAvail::IsPageAvail(
       nResult = m_pHintTables->CheckPage(dwPage, pHints);
       if (nResult != DataAvailable)
         return nResult;
+      // We should say to the document, which object is the page.
+      m_pDocument->SetPageObjNum(dwPage, GetPage(dwPage)->GetObjNum());
       m_pagesLoadState.insert(dwPage);
       return DataAvailable;
     }
@@ -1763,11 +1765,18 @@ CPDF_Dictionary* CPDF_DataAvail::GetPage(int index) {
       uint32_t dwObjNum = 0;
       bool bPagePosGot = m_pHintTables->GetPagePos(index, &szPageStartPos,
                                                    &szPageLength, &dwObjNum);
+      if (!dwObjNum)
+        return nullptr;
+      // Page object already can be parsed in document.
+      CPDF_Object* pPageDict = m_pDocument->GetIndirectObject(dwObjNum);
+      if (pPageDict)
+        return pPageDict->GetDict();
+
       if (!bPagePosGot)
         return nullptr;
 
       m_syntaxParser.InitParser(m_pFileRead, (uint32_t)szPageStartPos);
-      CPDF_Object* pPageDict = ParseIndirectObjectAt(0, dwObjNum, m_pDocument);
+      pPageDict = ParseIndirectObjectAt(0, dwObjNum, m_pDocument);
       if (!pPageDict)
         return nullptr;
 
diff --git a/core/fpdfapi/parser/cpdf_document.cpp b/core/fpdfapi/parser/cpdf_document.cpp
index c5f64a790c..5f64c14d31 100644
--- a/core/fpdfapi/parser/cpdf_document.cpp
+++ b/core/fpdfapi/parser/cpdf_document.cpp
@@ -425,7 +425,9 @@ CPDF_Document::CPDF_Document(std::unique_ptr<CPDF_Parser> pParser)
 
 CPDF_Document::~CPDF_Document() {
   delete m_pDocPage;
-  CPDF_ModuleMgr::Get()->GetPageModule()->ClearStockFont(this);
+  if (CPDF_ModuleMgr::Get()->GetPageModule()) {
+    CPDF_ModuleMgr::Get()->GetPageModule()->ClearStockFont(this);
+  }
   m_pByteStringPool.DeleteObject();  // Make weak.
 }
 
diff --git a/core/fpdfapi/parser/cpdf_document_unittest.cpp b/core/fpdfapi/parser/cpdf_document_unittest.cpp
new file mode 100644
index 0000000000..128b8d91c4
--- /dev/null
+++ b/core/fpdfapi/parser/cpdf_document_unittest.cpp
@@ -0,0 +1,42 @@
+// 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.
+
+#include "core/fpdfapi/parser/cpdf_document.h"
+
+#include <memory>
+
+#include "core/fpdfapi/parser/cpdf_parser.h"
+#include "core/fpdfapi/parser/cpdf_dictionary.h"
+#include "core/fxcrt/fx_memory.h"
+#include "testing/gtest/include/gtest/gtest.h"
+
+namespace {
+
+using ScopedDictionary =
+    std::unique_ptr<CPDF_Dictionary, ReleaseDeleter<CPDF_Dictionary>>;
+
+}  // namespace
+
+TEST(cpdf_document, UseCachedPageObjNumIfHaveNotPagesDict) {
+  // ObjNum can be added in CPDF_DataAvail::IsPageAvail, and PagesDict
+  // can be not exists in this case.
+  // (case, when hint table is used to page check in CPDF_DataAvail).
+  std::unique_ptr<CPDF_Parser> parser(new CPDF_Parser());
+  CPDF_Document document(std::move(parser));
+  ScopedDictionary dict(new CPDF_Dictionary());
+  const int page_count = 100;
+  dict->SetIntegerFor("N", page_count);
+  document.LoadLinearizedDoc(dict.get());
+  ASSERT_EQ(page_count, document.GetPageCount());
+  CPDF_Object* page_stub = new CPDF_Dictionary();
+  const uint32_t obj_num = document.AddIndirectObject(page_stub);
+  const int test_page_num = 33;
+
+  EXPECT_FALSE(document.IsPageLoaded(test_page_num));
+  EXPECT_EQ(nullptr, document.GetPage(test_page_num));
+
+  document.SetPageObjNum(test_page_num, obj_num);
+  EXPECT_TRUE(document.IsPageLoaded(test_page_num));
+  EXPECT_EQ(page_stub, document.GetPage(test_page_num));
+}