[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[lmi] Code inspection requested
From: |
Greg Chicares |
Subject: |
[lmi] Code inspection requested |
Date: |
Fri, 14 May 2010 22:02:22 +0000 |
User-agent: |
Thunderbird 2.0.0.24 (Windows/20100228) |
I changed quite a few files here:
http://svn.savannah.gnu.org/viewvc?view=rev&root=lmi&revision=4951
Though I tried to be very careful, and ran every test I have, there's
a chance that I nevertheless made a mistake. I think the best way to
find any mistake in this changeset is inspection rather than testing.
The likeliest mistake is a parity reversal, such as
- if(foo.end() == std::find(foo.begin(), foo.end(), bar))
+ if(contains(foo, bar) // wrong: should be !contains
It doesn't matter if I missed a refactoring opportunity somewhere;
all that matters is whether I introduced an error.
I generally avoid making such sweeping changes, but I always have to
stop and think what the verbose idiom above means in context, so I
see a lot of value in writing a clearer substitute. For instance,
without "map_lookup.hpp", I would write this the way it originally
appeared, but then I couldn't get it at a glance:
- button_bitmaps const& bitmaps = all_button_bitmaps();
- button_bitmaps::const_iterator it =
- bitmaps.find(std::make_pair(button, state));
- if(it == bitmaps.end())
- {
- fatal_error() << "Unknown or missing button bitmap" << LMI_FLUSH;
- }
- return it->second;
+ return map_lookup(all_button_bitmaps(), std::make_pair(button, state));
- [lmi] Code inspection requested,
Greg Chicares <=