Revert of Update SplitString calls to new form (patchset #5 id:80001 of https://codereview.chromium.org/1272823003/ )

Reason for revert:
Caused Blink layout test failures in media/encrypted-media/encrypted-media-requestmediakeysystemaccess.html :

http://test-results.appspot.com/dashboards/flakiness_dashboard.html#showExpectations=true&tests=media%2Fencrypted-media%2Fencrypted-media-requestmediakeysystemaccess.html

Original issue's description:
> Update SplitString calls to new form
>
> Uses the new form for most (but not quite all) of the remaining users of the old form.
>
> Changes media mime util codec list parsing to expect no result from the string "," rather than two empty strings. The old SplitString call had a special case where if the input was empty, it would return empty, but if it had one split character, it would return two empty strings as results.
>
> The new one lets you choose but the options are either (1) empty string -> one empty string and "," -> two empty strings, or (2) map both to no results for when you don't want empty results. I'm pretty sure media codec parsing actually wants the latter behavior, so I updated the call to discard empty results and MimeUtilTest.ParseCodecString is updated.
>
> Committed: https://crrev.com/0aa7c64253cca8b636d52d1d01d94f96ab9c13fa
> Cr-Commit-Position: refs/heads/master@{#342238}

TBR=sky@chromium.org,dalecurtis@chromium.org,brettw@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1278973003

Cr-Commit-Position: refs/heads/master@{#342257}

1 files changed, 3 insertions, 4 deletions

diff --git a/ui/gfx/font_fallback_win.cc b/ui/gfx/font_fallback_win.cc
index e5771cf..6b9c620 100644
--- a/ui/gfx/font_fallback_win.cc
+++ b/ui/gfx/font_fallback_win.cc
@@ -180,8 +180,8 @@ namespace internal {
 void ParseFontLinkEntry(const std::string& entry,
                         std::string* filename,
                         std::string* font_name) {
-  std::vector<std::string> parts = base::SplitString(
-      entry, ",", base::TRIM_WHITESPACE, base::SPLIT_WANT_ALL);
+  std::vector<std::string> parts;
+  base::SplitString(entry, ',', &parts);
   filename->clear();
   font_name->clear();
   if (parts.size() > 0)
@@ -199,8 +199,7 @@ void ParseFontFamilyString(const std::string& family,
   // followed optionally by the font family name and a pair of integer scaling
   // factors.
   // TODO(asvitkine): Should we support these scaling factors?
-  *font_names = base::SplitString(
-      family, "&", base::TRIM_WHITESPACE, base::SPLIT_WANT_ALL);
+  base::SplitString(family, '&', font_names);
   if (!font_names->empty()) {
     const size_t index = font_names->back().find('(');
     if (index != std::string::npos) {