Make CPDF_Document cache CPDF_Pages

We cache pages not by page number, which can bounce around as pages
are inserted or removed, but by page dictionary's object number.

Since the page may be created under one function and used under
another, we can't take the shortcut of not instantiating a render
cache nor not parsing the page.

Change-Id: I9a325cda8b3141153544ac53e78a51a44e6b411a
Reviewed-on: https://pdfium-review.googlesource.com/32830
Commit-Queue: Tom Sepez <tsepez@chromium.org>
Reviewed-by: dsinclair <dsinclair@chromium.org>

5 files changed, 32 insertions, 8 deletions

diff --git a/core/fpdfapi/edit/cpdf_pagecontentgenerator_unittest.cpp b/core/fpdfapi/edit/cpdf_pagecontentgenerator_unittest.cpp
index 1244b12331..859f5b3991 100644
--- a/core/fpdfapi/edit/cpdf_pagecontentgenerator_unittest.cpp
+++ b/core/fpdfapi/edit/cpdf_pagecontentgenerator_unittest.cpp
@@ -129,7 +129,7 @@ TEST_F(CPDF_PageContentGeneratorTest, ProcessGraphics) {
   auto pDoc = pdfium::MakeUnique<CPDF_Document>(nullptr);
   pDoc->CreateNewDoc();
   CPDF_Dictionary* pPageDict = pDoc->CreateNewPage(0);
-  auto pTestPage = pdfium::MakeRetain<CPDF_Page>(pDoc.get(), pPageDict, false);
+  RetainPtr<CPDF_Page> pTestPage = pDoc->GetOrCreatePDFPage(pPageDict);
   CPDF_PageContentGenerator generator(pTestPage.Get());
   std::ostringstream buf;
   TestProcessPath(&generator, &buf, pPathObj.get());
@@ -168,7 +168,7 @@ TEST_F(CPDF_PageContentGeneratorTest, ProcessStandardText) {
   auto pDoc = pdfium::MakeUnique<CPDF_Document>(nullptr);
   pDoc->CreateNewDoc();
   CPDF_Dictionary* pPageDict = pDoc->CreateNewPage(0);
-  auto pTestPage = pdfium::MakeRetain<CPDF_Page>(pDoc.get(), pPageDict, false);
+  RetainPtr<CPDF_Page> pTestPage = pDoc->GetOrCreatePDFPage(pPageDict);
   CPDF_PageContentGenerator generator(pTestPage.Get());
   auto pTextObj = pdfium::MakeUnique<CPDF_TextObject>();
   CPDF_Font* pFont = CPDF_Font::GetStockFont(pDoc.get(), "Times-Roman");
@@ -231,7 +231,7 @@ TEST_F(CPDF_PageContentGeneratorTest, ProcessText) {
   auto pDoc = pdfium::MakeUnique<CPDF_Document>(nullptr);
   pDoc->CreateNewDoc();
   CPDF_Dictionary* pPageDict = pDoc->CreateNewPage(0);
-  auto pTestPage = pdfium::MakeRetain<CPDF_Page>(pDoc.get(), pPageDict, false);
+  RetainPtr<CPDF_Page> pTestPage = pDoc->GetOrCreatePDFPage(pPageDict);
   CPDF_PageContentGenerator generator(pTestPage.Get());
 
   std::ostringstream buf;
diff --git a/core/fpdfapi/page/cpdf_page.cpp b/core/fpdfapi/page/cpdf_page.cpp
index 7c0a3234c9..c5c85635b2 100644
--- a/core/fpdfapi/page/cpdf_page.cpp
+++ b/core/fpdfapi/page/cpdf_page.cpp
@@ -21,11 +21,11 @@
 
 CPDF_Page::CPDF_Page(CPDF_Document* pDocument,
                      CPDF_Dictionary* pPageDict,
-                     bool bPageCache)
+                     bool bUseRenderCache)
     : CPDF_PageObjectHolder(pDocument, pPageDict),
       m_PageSize(100, 100),
       m_pPDFDocument(pDocument) {
-  if (bPageCache)
+  if (bUseRenderCache)
     m_pPageRender = pdfium::MakeUnique<CPDF_PageRenderCache>(this);
   if (!pPageDict)
     return;
diff --git a/core/fpdfapi/page/cpdf_page.h b/core/fpdfapi/page/cpdf_page.h
index 0bb99b73f1..b94326e47b 100644
--- a/core/fpdfapi/page/cpdf_page.h
+++ b/core/fpdfapi/page/cpdf_page.h
@@ -12,6 +12,7 @@
 #include "core/fpdfapi/page/cpdf_pageobjectholder.h"
 #include "core/fxcrt/fx_coordinates.h"
 #include "core/fxcrt/fx_system.h"
+#include "core/fxcrt/observable.h"
 #include "core/fxcrt/retain_ptr.h"
 #include "core/fxcrt/unowned_ptr.h"
 #include "third_party/base/optional.h"
@@ -22,7 +23,9 @@ class CPDF_Object;
 class CPDF_PageRenderCache;
 class CPDF_PageRenderContext;
 
-class CPDF_Page : public Retainable, public CPDF_PageObjectHolder {
+class CPDF_Page : public Retainable,
+                  public Observable<CPDF_Page>,
+                  public CPDF_PageObjectHolder {
  public:
   class View {};  // Caller implements as desired, empty here due to layering.
   class Extension : public Retainable {};  // XFA page parent class, layering.
@@ -66,7 +69,7 @@ class CPDF_Page : public Retainable, public CPDF_PageObjectHolder {
  private:
   CPDF_Page(CPDF_Document* pDocument,
             CPDF_Dictionary* pPageDict,
-            bool bPageCache);
+            bool bUseRenderCache);
   ~CPDF_Page() override;
 
   void StartParse();
diff --git a/core/fpdfapi/parser/cpdf_document.cpp b/core/fpdfapi/parser/cpdf_document.cpp
index 7968d1b019..2e4baabe91 100644
--- a/core/fpdfapi/parser/cpdf_document.cpp
+++ b/core/fpdfapi/parser/cpdf_document.cpp
@@ -602,6 +602,19 @@ bool CPDF_Document::InsertNewPage(int iPage, CPDF_Dictionary* pPageDict) {
   return true;
 }
 
+RetainPtr<CPDF_Page> CPDF_Document::GetOrCreatePDFPage(
+    CPDF_Dictionary* pPageDict) {
+  std::pair<uint32_t, uint32_t> key = {pPageDict->GetObjNum(),
+                                       pPageDict->GetGenNum()};
+  if (m_PageMap[key])
+    return RetainPtr<CPDF_Page>(m_PageMap[key].Get());
+
+  auto pPage = pdfium::MakeRetain<CPDF_Page>(this, pPageDict, true);
+  pPage->ParseContent();
+  m_PageMap[key].Reset(pPage.Get());
+  return pPage;
+}
+
 void CPDF_Document::DeletePage(int iPage) {
   CPDF_Dictionary* pPages = GetPagesDict();
   if (!pPages)
diff --git a/core/fpdfapi/parser/cpdf_document.h b/core/fpdfapi/parser/cpdf_document.h
index c2774b163a..c240e77567 100644
--- a/core/fpdfapi/parser/cpdf_document.h
+++ b/core/fpdfapi/parser/cpdf_document.h
@@ -8,6 +8,7 @@
 #define CORE_FPDFAPI_PARSER_CPDF_DOCUMENT_H_
 
 #include <functional>
+#include <map>
 #include <memory>
 #include <set>
 #include <utility>
@@ -64,6 +65,7 @@ class CPDF_Document : public CPDF_IndirectObjectHolder {
   const CPDF_Dictionary* GetInfo() const { return m_pInfoDict.Get(); }
   CPDF_Dictionary* GetInfo() { return m_pInfoDict.Get(); }
 
+  RetainPtr<CPDF_Page> GetOrCreatePDFPage(CPDF_Dictionary* pPageDict);
   void DeletePage(int iPage);
   int GetPageCount() const;
   bool IsPageLoaded(int iPage) const;
@@ -167,7 +169,13 @@ class CPDF_Document : public CPDF_IndirectObjectHolder {
   std::unique_ptr<CPDF_DocRenderData> m_pDocRender;
   std::unique_ptr<JBig2_DocumentContext> m_pCodecContext;
   std::unique_ptr<CPDF_LinkList> m_pLinksContext;
-  std::vector<uint32_t> m_PageList;  // Page number to page's dict objnum.
+
+  // Page number (index) to page's dict objnum.
+  std::vector<uint32_t> m_PageList;
+
+  // Dict {objnum, gennum} to page mapping.
+  std::map<std::pair<uint32_t, uint32_t>, CPDF_Page::ObservedPtr> m_PageMap;
+
   UnownedPtr<Extension> m_pExtension;
 };