Case-insensitive search for non-ASCII characters in auto-complete of Drive files.

chrome/browser/chromeos/drive/search_metadata.h

Index: chrome/browser/chromeos/drive/search_metadata.h
diff --git a/chrome/browser/chromeos/drive/search_metadata.h b/chrome/browser/chromeos/drive/search_metadata.h
index dbe0ae2466252675e5c83b40ac79cee2b1d636f9..13e13363d6047c76b37c67b7528117575255c37e 100644
--- a/chrome/browser/chromeos/drive/search_metadata.h
+++ b/chrome/browser/chromeos/drive/search_metadata.h
@@ -35,7 +35,7 @@ void SearchMetadata(
 // Finds |query| in |text| while ignoring case. Returns true if |query| is

[satorux1, 2013/05/27 05:00:24]

please mention that non-ASCII characters are supported.

BTW, does this perform any normalization against hiragana/katakana? If so, it'd
be nice to add some tests.

[kinaba, 2013/05/27 05:50:26]

It does. Added a test.

 // found. Returns false if |query| is not found, or empty. |highlighted_text|
 // will have the original text with matched portions highlighted with <b> tag
-// (multiple portions can be highlighted). Meta characters are escaped like
+// (only the first match is highlighted). Meta characters are escaped like
 // &lt;. The original contents of |highlighted| will be lost.
 bool FindAndHighlight(const std::string& text,
                       const std::string& query,