Replacing post-increment with pre-increment for consistency

content/renderer/media/webmediaplayer_util.cc

Index: content/renderer/media/webmediaplayer_util.cc
diff --git a/content/renderer/media/webmediaplayer_util.cc b/content/renderer/media/webmediaplayer_util.cc
index a245f038b0ae25efd3081aa4537d1e777e2cf209..ed7e0cdfd6bd3b04709fc3f25fe65573678bc590 100644
--- a/content/renderer/media/webmediaplayer_util.cc
+++ b/content/renderer/media/webmediaplayer_util.cc
@@ -31,7 +31,7 @@ base::TimeDelta ConvertSecondsToTimestamp(double seconds) {
 blink::WebTimeRanges ConvertToWebTimeRanges(
     const media::Ranges<base::TimeDelta>& ranges) {
   blink::WebTimeRanges result(ranges.size());
-  for (size_t i = 0; i < ranges.size(); i++) {
+  for (size_t i = 0; i < ranges.size(); ++i) {

[Ami GONE FROM CHROMIUM, 2013/12/19 21:21:01]

CL description says "to improve efficiency" but AFAICT Release builds of this
file before and after this change generate bit-exact results, because clang is
smart enough to make this change itself.  Do you have a toolchain where this is
not the case?

Using pre-increment instead of post-increment for PODs is rarely/never a
performance win, but it can be a readability win due to consistency with non-POD
pre-increment (where efficiency _can_ be an issue).

So, this change LGTM as long as you update the summary & description to
something more apt, like: "Replace post-increment with pre-increment for
consistency".

     result[i].start = ranges.start(i).InSecondsF();
     result[i].end = ranges.end(i).InSecondsF();
   }