Add PDF engine method to get all text in the document

This will be used by the Lens Overlay to pass a fraction of the document
to be used for suggest signals while the full PDF is being processed.
Retrieving the text in the Lens overlay controller will come in a
followup CL.

Bug: 379344946
Change-Id: Ie52df82022916a3bb367150207d1b70e03fbce8a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/6032353
Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: Duncan Mercer <mercerd@google.com>
Cr-Commit-Position: refs/heads/main@{#1385831}

pdf/pdfium/pdfium_engine.cc

@@ -2248,6 +2248,24 @@ void PDFiumEngine::DisplayAnnotations(bool display) {
   InvalidateAllPages();
 }
 
+std::u16string PDFiumEngine::GetAllText(uint32_t size_limit) {
+  std::u16string all_pages_text;
+
+  for (auto& page : pages_) {
+    // Add the current page's text to the output.
+    auto range = PDFiumRange::AllTextOnPage(page.get());
+    all_pages_text.append(range.GetText());
+
+    // Truncate and exit early if over the size limit.
+    if (all_pages_text.size() > size_limit) {
+      all_pages_text.resize(size_limit);
+      break;
+    }
+  }
+
+  return all_pages_text;
+}
+
 void PDFiumEngine::InvalidateAllPages() {
   CancelPaints();
   StopFind();

pdf/pdfium/pdfium_engine.h

@@ -222,6 +222,11 @@ class PDFiumEngine : public DocumentLoader::Client, public IFSDK_PAUSE {
   void SetDocumentLayout(DocumentLayout::PageSpread page_spread);
   void DisplayAnnotations(bool display);
 
+  // Returns the text contained in the PDF. If the size of the text is larger
+  // than `size_limit`, the remaining text will be truncated, and a string with
+  // the first `size_limit` characters will be returned.
+  std::u16string GetAllText(uint32_t size_limit);
+
   // Applies the document layout options proposed by a call to
   // PDFiumEngineClient::ProposeDocumentLayout(), returning the overall size of
   // the new effective layout.

pdf/pdfium/pdfium_engine_unittest.cc

@@ -1097,6 +1097,32 @@ TEST_P(PDFiumEngineTest, DrawTextSelectionsBigtableMicro) {
       *engine, /*page_index=*/0, "bigtable_micro_selection.png");
 }
 
+TEST_P(PDFiumEngineTest, GetAllText) {
+  NiceMock<MockTestClient> client;
+  std::unique_ptr<PDFiumEngine> engine =
+      InitializeEngine(&client, FILE_PATH_LITERAL("hello_world2.pdf"));
+  ASSERT_TRUE(engine);
+
+  auto text = engine->GetAllText(/*size_limit=*/100);
+  constexpr char16_t kExpectedText[] =
+      u"Hello, world!\r\nGoodbye, world!Hello, world!\r\nGoodbye, world!";
+
+  EXPECT_EQ(kExpectedText, text);
+}
+
+TEST_P(PDFiumEngineTest, GetAllTextSizeLimit) {
+  NiceMock<MockTestClient> client;
+  std::unique_ptr<PDFiumEngine> engine =
+      InitializeEngine(&client, FILE_PATH_LITERAL("hello_world2.pdf"));
+  ASSERT_TRUE(engine);
+
+  auto text = engine->GetAllText(/*size_limit=*/30);
+  ASSERT_EQ(30u, text.size());
+
+  constexpr char16_t kExpectedText[] = u"Hello, world!\r\nGoodbye, world!";
+  EXPECT_EQ(kExpectedText, text);
+}
+
 TEST_P(PDFiumEngineTest, LinkNavigates) {
   NiceMock<MockTestClient> client;
   std::unique_ptr<PDFiumEngine> engine =