Add UUID, strlcpy, and their tests to util

Add UUID, c16lcpy (strlcpy for char16*), and their tests to util.

These are dependencies of the upcoming MinidumpStringWriter.

R=rsesek@chromium.org

Committed: https://chromium.googlesource.com/crashpad/crashpad/+/4d6b867a1fbd

[Robert Sesek, 2014-08-01 17:28:18]

https://codereview.chromium.org/430003003/diff/1/util/misc/uuid.h
File util/misc/uuid.h (right):

https://codereview.chromium.org/430003003/diff/1/util/misc/uuid.h#newcode28
util/misc/uuid.h:28: struct UUID {
No default constructor to zero-initialize everything?

https://codereview.chromium.org/430003003/diff/1/util/misc/uuid.h#newcode44
util/misc/uuid.h:44: uint8_t data_4[8];
You may want to split data_4 into two fields: data_4[2] and data_5[6]. Because
the string representation of a UUID actually contains five fields:
29C24D52-A794-4887-B392-C855C155A28C, per the RFC:

"""
 UUID                   = time-low "-" time-mid "-"
                               time-high-and-version "-"
                               clock-seq-and-reserved
                               clock-seq-low "-" node
      time-low               = 4hexOctet
      time-mid               = 2hexOctet
      time-high-and-version  = 2hexOctet
      clock-seq-and-reserved = hexOctet
      clock-seq-low          = hexOctet
      node                   = 6hexOctet
      hexOctet               = hexDigit hexDigit
"""

And just for contrast, uuid_t on Mac is just a unsigned char[16].

https://codereview.chromium.org/430003003/diff/1/util/misc/uuid.h#newcode47
util/misc/uuid.h:47: 
A little surprised to not see an ostream operator or a ToString() method.

[Mark Mentovai, 2014-08-01 18:05:49]

https://codereview.chromium.org/430003003/diff/1/util/misc/uuid.h
File util/misc/uuid.h (right):

https://codereview.chromium.org/430003003/diff/1/util/misc/uuid.h#newcode28
util/misc/uuid.h:28: struct UUID {
rsesek wrote:
> No default constructor to zero-initialize everything?

I kinda wanted to keep it POD, but maybe it’s enough that it’s standard layout
now that we’re using C++11?

https://codereview.chromium.org/430003003/diff/1/util/misc/uuid.h#newcode44
util/misc/uuid.h:44: uint8_t data_4[8];
rsesek wrote:
> You may want to split data_4 into two fields: data_4[2] and data_5[6]. Because
> the string representation of a UUID actually contains five fields:
> 29C24D52-A794-4887-B392-C855C155A28C, per the RFC:

This definition came from Microsoft’s GUID definition, but we don’t actually use
that anywhere (which is why I renamed this to UUID). I was concerned about
byte-swapping, but if we always treat it as a uint8_t[16], we don’t really need
to worry about that. I’ll change it.

[Robert Sesek, 2014-08-01 18:37:03]

LGTM

https://codereview.chromium.org/430003003/diff/20001/util/misc/uuid.cc
File util/misc/uuid.cc (right):

https://codereview.chromium.org/430003003/diff/20001/util/misc/uuid.cc#newcode30
util/misc/uuid.cc:30: COMPILE_ASSERT(std::is_standard_layout<UUID>::value,
Faaaancy.

https://codereview.chromium.org/430003003/diff/20001/util/misc/uuid.h
File util/misc/uuid.h (right):

https://codereview.chromium.org/430003003/diff/20001/util/misc/uuid.h#newcode43
util/misc/uuid.h:43: //! \brief Formats the %UUID per RFC 4122 3.
A § could look nice here.

[Mark Mentovai, 2014-08-01 18:39:59]

Committed patchset #3 manually as 4d6b867a1fbd (presubmit successful).