Make various string_util functions take StringPieces instead of char[].

Make various string_util functions take StringPieces instead of char[].

This allows most callers to pass string literals or results of ASCIIToUTF16(),
rather than construct error-prone character arrays that must be manually
nul-terminated.

Also cleans up various bits of the callers.

BUG=104260
TEST=none

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=273765

[Peter Kasting, 2014-05-20 18:22:51]

In a few cases I removed named string constants passed to e.g. RemoveChars()
when they were only used once and naming them didn't seem to me to make things
clearer, but it any of these strike you as less clear, let me know.

[Peter Kasting, 2014-05-22 23:33:54]

Ping

[darin (slow to review), 2014-05-22 23:49:44]

https://codereview.chromium.org/296593003/diff/40001/chrome/browser/guest_vie...
File chrome/browser/guest_view/ad_view/ad_view_guest.cc (right):

https://codereview.chromium.org/296593003/diff/40001/chrome/browser/guest_vie...
chrome/browser/guest_view/ad_view/ad_view_guest.cc:54:
DCHECK(StartsWithASCII(error_type, "net::", true));
Ditto.

https://codereview.chromium.org/296593003/diff/40001/chrome/browser/guest_vie...
File chrome/browser/guest_view/web_view/web_view_guest.cc (right):

https://codereview.chromium.org/296593003/diff/40001/chrome/browser/guest_vie...
chrome/browser/guest_view/web_view/web_view_guest.cc:811:
DCHECK(StartsWithASCII(error_type, "net::", true));
I don't know that it is guaranteed that the error description of blink-generated
errors will all start with "net::". I think this DCHECK is suspicious or at
least quite fragile.

https://codereview.chromium.org/296593003/diff/40001/chrome/common/localized_...
File chrome/common/localized_error.cc (right):

https://codereview.chromium.org/296593003/diff/40001/chrome/common/localized_...
chrome/common/localized_error.cc:583: DCHECK(StartsWithASCII(ascii_error_string,
"net::", true));
This case seems fine since the code is calling ErrorToString.

[Peter Kasting, 2014-05-22 23:54:49]

https://codereview.chromium.org/296593003/diff/40001/chrome/browser/guest_vie...
File chrome/browser/guest_view/web_view/web_view_guest.cc (right):

https://codereview.chromium.org/296593003/diff/40001/chrome/browser/guest_vie...
chrome/browser/guest_view/web_view/web_view_guest.cc:811:
DCHECK(StartsWithASCII(error_type, "net::", true));
On 2014/05/22 23:49:46, darin wrote:
> I don't know that it is guaranteed that the error description of
blink-generated
> errors will all start with "net::". I think this DCHECK is suspicious or at
> least quite fragile.

net::ErrorToString() prefixes "net::" onto all return values.  If you're worried
that won't be true in the future, I could make some of these callsites
conditional.

https://codereview.chromium.org/296593003/diff/40001/chrome/common/localized_...
File chrome/common/localized_error.cc (right):

https://codereview.chromium.org/296593003/diff/40001/chrome/common/localized_...
chrome/common/localized_error.cc:583: DCHECK(StartsWithASCII(ascii_error_string,
"net::", true));
On 2014/05/22 23:49:46, darin wrote:
> This case seems fine since the code is calling ErrorToString.

I'm confused.  Both other cases are also calling ErrorToString().  How is this
different?  Or did you mean to say something else?

[Peter Kasting, 2014-05-29 23:17:41]

Ping for questions above

[darin (slow to review), 2014-05-30 00:03:17]

LGTM

https://codereview.chromium.org/296593003/diff/40001/chrome/browser/guest_vie...
File chrome/browser/guest_view/web_view/web_view_guest.cc (right):

https://codereview.chromium.org/296593003/diff/40001/chrome/browser/guest_vie...
chrome/browser/guest_view/web_view/web_view_guest.cc:811:
DCHECK(StartsWithASCII(error_type, "net::", true));
On 2014/05/22 23:54:50, Peter Kasting wrote:
> On 2014/05/22 23:49:46, darin wrote:
> > I don't know that it is guaranteed that the error description of
> blink-generated
> > errors will all start with "net::". I think this DCHECK is suspicious or at
> > least quite fragile.
> 
> net::ErrorToString() prefixes "net::" onto all return values.  If you're
worried
> that won't be true in the future, I could make some of these callsites
> conditional.

Oh, nevermind. I was mistakenly thinking the string came from a WebURLError. You
can't be sure that a WebURLError's description string came from the network
stack. Sorry for the confusion on my part!

[Peter Kasting, 2014-05-30 00:04:17]

The CQ bit was checked by pkasting@chromium.org

[commit-bot: I haz the power, 2014-05-30 00:04:58]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/pkasting@chromium.org/296593003/40001

[commit-bot: I haz the power, 2014-05-30 01:07:43]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  chromium_presubmit on tryserver.chromium
(http://build.chromium.org/p/tryserver.chromium/builders/chromium_presubmit/bu...)

[commit-bot: I haz the power, 2014-05-30 01:13:46]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-05-30 01:13:47]

Try jobs failed on following builders:
  chromium_presubmit on tryserver.chromium
(http://build.chromium.org/p/tryserver.chromium/builders/chromium_presubmit/bu...)

[Peter Kasting, 2014-05-30 01:36:00]

The CQ bit was checked by pkasting@chromium.org

[commit-bot: I haz the power, 2014-05-30 01:39:15]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/pkasting@chromium.org/296593003/60001

[commit-bot: I haz the power, 2014-05-30 08:16:13]

Change committed as 273765