iwyu: Include stringprintf.h where appropriate, part 1.

iwyu: Include stringprintf.h where appropriate, part 1.

BUG=82098
TEST=none

R=thakis@chromium.org

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=84754

[Nico, 2011-05-10 02:57:50]

LGTM

But: It would be great if you could file a bug for the iwyu work, add that to
all your CLs, and in that bug describe:

* your approach (I assume you ran iwyu once and now you're somehow slicing that
patch into smaller pieces)
* build time effects / savings
* total loc +/- by the combined patch

http://codereview.chromium.org/6997006/diff/1/base/json/string_escape.cc
File base/json/string_escape.cc (right):

http://codereview.chromium.org/6997006/diff/1/base/json/string_escape.cc#newc...
base/json/string_escape.cc:67: base::StringAppendF(dst, "\\u%04X", as_uint);
hm, these are not necessary because of ADL. does iwyu add these? do we prefer
explicitness over adl? it's a bit less flexible.

[James Hawkins, 2011-05-10 04:01:43]

Added a bug, but have to add more data to it at a later point.

http://codereview.chromium.org/6997006/diff/1/base/json/string_escape.cc
File base/json/string_escape.cc (right):

http://codereview.chromium.org/6997006/diff/1/base/json/string_escape.cc#newc...
base/json/string_escape.cc:67: base::StringAppendF(dst, "\\u%04X", as_uint);
On 2011/05/10 02:57:51, Nico wrote:
> hm, these are not necessary because of ADL. does iwyu add these? do we prefer
> explicitness over adl? it's a bit less flexible.

No, I added them.  I prefer them, but I don't think the style guide says
anything about this. My rationale is that the style guide bans the use of using
entire namespaces (specifically because of ambiguity), and in my mind relyingon
ADL to get this right is even worse than using namespace base at the top of the
file.