Move PDF text search code into its own standalone function.

The code in PdfViewWebPlugin::SearchString() has little to do with the
rest of PdfViewWebPlugin. Make it a standalone function, and use it to
replace the text search code used for testing in FindTextTestClient. The
function can also be easily unit tested in the future.

Change-Id: Ic359757c0c299be1b7f792158902c60e391bd916
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4137334
Code-Coverage: findit-for-me@appspot.gserviceaccount.com <findit-for-me@appspot.gserviceaccount.com>
Reviewed-by: Alan Screen <awscreen@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1369464}

pdf/BUILD.gn

@@ -139,6 +139,8 @@ if (enable_pdf) {
       "pdfium/pdfium_unsupported_features.h",
       "preview_mode_client.cc",
       "preview_mode_client.h",
+      "text_search.cc",
+      "text_search.h",
       "ui/document_properties.cc",
       "ui/document_properties.h",
       "ui/file_name.cc",

pdf/pdf_view_web_plugin.cc

@@ -29,7 +29,6 @@
 #include "base/functional/callback.h"
 #include "base/i18n/char_iterator.h"
 #include "base/i18n/rtl.h"
-#include "base/i18n/string_search.h"
 #include "base/i18n/time_formatting.h"
 #include "base/logging.h"
 #include "base/memory/raw_ptr.h"
@@ -70,6 +69,7 @@
 #include "pdf/pdfium/pdfium_engine.h"
 #include "pdf/pdfium/pdfium_engine_client.h"
 #include "pdf/post_message_receiver.h"
+#include "pdf/text_search.h"
 #include "pdf/ui/document_properties.h"
 #include "pdf/ui/file_name.h"
 #include "pdf/ui/thumbnail.h"
@@ -1157,14 +1157,7 @@ std::vector<PDFiumEngineClient::SearchStringResult>
 PdfViewWebPlugin::SearchString(const std::u16string& needle,
                                const std::u16string& haystack,
                                bool case_sensitive) {
-  base::i18n::RepeatingStringSearch searcher(
+  return TextSearch(/*needle=*/needle, /*haystack=*/haystack, case_sensitive);
-      /*find_this=*/needle, /*in_this=*/haystack, case_sensitive);
-  std::vector<SearchStringResult> results;
-  int match_index;
-  int match_length;
-  while (searcher.NextMatchResult(match_index, match_length))
-    results.push_back({.start_index = match_index, .length = match_length});
-  return results;
 }
 
 void PdfViewWebPlugin::DocumentLoadComplete() {

pdf/pdfium/findtext_unittest.cc

@@ -10,6 +10,7 @@
 #include "pdf/pdfium/pdfium_engine.h"
 #include "pdf/pdfium/pdfium_test_base.h"
 #include "pdf/test/test_client.h"
+#include "pdf/text_search.h"
 #include "testing/gmock/include/gmock/gmock.h"
 
 using testing::_;
@@ -37,22 +38,7 @@ class FindTextTestClient : public TestClient {
     EXPECT_FALSE(needle.empty());
     EXPECT_FALSE(haystack.empty());
     EXPECT_EQ(case_sensitive, expected_case_sensitive_);
-
+    return TextSearch(/*needle=*/needle, /*haystack=*/haystack, case_sensitive);
-    std::vector<SearchStringResult> results;
-
-    size_t pos = 0;
-    while (true) {
-      pos = haystack.find(needle, pos);
-      if (pos == std::u16string::npos)
-        break;
-
-      SearchStringResult result;
-      result.length = needle.size();
-      result.start_index = pos;
-      results.push_back(result);
-      pos += needle.size();
-    }
-    return results;
   }
 
  private:

pdf/text_search.cc

@@ -0,0 +1,30 @@
+// Copyright 2024 The Chromium Authors
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#include "pdf/text_search.h"
+
+#include <string>
+#include <vector>
+
+#include "base/i18n/string_search.h"
+#include "pdf/pdfium/pdfium_engine_client.h"
+
+namespace chrome_pdf {
+
+std::vector<PDFiumEngineClient::SearchStringResult> TextSearch(
+    const std::u16string& needle,
+    const std::u16string& haystack,
+    bool case_sensitive) {
+  base::i18n::RepeatingStringSearch searcher(
+      /*find_this=*/needle, /*in_this=*/haystack, case_sensitive);
+  std::vector<PDFiumEngineClient::SearchStringResult> results;
+  int match_index;
+  int match_length;
+  while (searcher.NextMatchResult(match_index, match_length)) {
+    results.push_back({.start_index = match_index, .length = match_length});
+  }
+  return results;
+}
+
+}  // namespace chrome_pdf

pdf/text_search.h

@@ -0,0 +1,22 @@
+// Copyright 2024 The Chromium Authors
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef PDF_TEXT_SEARCH_H_
+#define PDF_TEXT_SEARCH_H_
+
+#include <string>
+#include <vector>
+
+#include "pdf/pdfium/pdfium_engine_client.h"
+
+namespace chrome_pdf {
+
+std::vector<PDFiumEngineClient::SearchStringResult> TextSearch(
+    const std::u16string& needle,
+    const std::u16string& haystack,
+    bool case_sensitive);
+
+}  // namespace chrome_pdf
+
+#endif  // PDF_TEXT_SEARCH_H_