Debug Windows compiler complaint fixes

This delivery fixes all the remaining Debug Windows compiler complaints. There are still ~5 Release-only complaints.

[robertphillips, 2012-05-14 13:40:53]

http://codereview.appspot.com/6208055/diff/1/tests/PDFPrimitivesTest.cpp
File tests/PDFPrimitivesTest.cpp (left):

http://codereview.appspot.com/6208055/diff/1/tests/PDFPrimitivesTest.cpp#oldc...
tests/PDFPrimitivesTest.cpp:10: #include <string>
This causes complaints r.e. exception handling and isn't really necessary

[TomH, 2012-05-14 13:45:44]

Modulo PDF-nit, LGTM.

http://codereview.appspot.com/6208055/diff/1/bench/InterpBench.cpp
File bench/InterpBench.cpp (right):

http://codereview.appspot.com/6208055/diff/1/bench/InterpBench.cpp#newcode88
bench/InterpBench.cpp:88: curr += step;
This is now coding-standards compliant, but harder to follow. I suppose that
with 80col you had to do it. :(

http://codereview.appspot.com/6208055/diff/1/src/pdf/SkPDFDocument.cpp
File src/pdf/SkPDFDocument.cpp (right):

http://codereview.appspot.com/6208055/diff/1/src/pdf/SkPDFDocument.cpp#newcod...
src/pdf/SkPDFDocument.cpp:266: fTrailerDict->insertInt("Size", int(objCount));
I hate to do this, but we should loop in vandebo@ here to verify PDF
correctness?

[robertphillips, 2012-05-14 14:05:09]

Added vandebo@ to assess the PDF file changes.

http://codereview.appspot.com/6208055/diff/1/bench/InterpBench.cpp
File bench/InterpBench.cpp (right):

http://codereview.appspot.com/6208055/diff/1/bench/InterpBench.cpp#newcode88
bench/InterpBench.cpp:88: curr += step;
80 cols is sometimes very cruel :(

[reed1, 2012-05-14 15:22:04]

The 80-col rule was made for man, not man for the 80-col rule...

[Steve VanDeBogart, 2012-05-14 16:19:19]

LGTM

[robertphillips, 2012-05-14 16:34:04]

committed as r3924