what's the deal with this my_pair business? Either use std::pair or leave a comment explaining why you did this. I suspect it has to do with POD-ness and static initializers? if so, call it pod_pair.<div><br></div><div>
Even better, you can probably get away with having one struct type with int or unsigned (whichever signedness enum values have by default) as their first member, and avoid having a bunch of bit-identical instantiations of std::map for all these different types.</div>
<div><br></div><div>You should see whether the maps are even needed at all. A linear search may be faster for small fixed sets of data like this.<br><div><br></div><div>
Also, in LLVM, most (all?) code uses one space to separate the initial // from the comment text itself. You are consistently using 2.</div><div><br></div><div>in writeHexStream, please make</div><div>const char *hex = "0123456789ABCDEF";</div>
<div>into</div><div>static const char hex[] = "0123456789ABCDEF";</div><div><br></div><div><div>+llvm::raw_ostream &writeHexNumber(llvm::raw_ostream &Out, unsigned long long N) {</div><div>+ if ( N > 10 )</div>
<div>+ Out << "0x";</div><div>+ Out.write_hex(N);</div><div>+ return Out;</div><div>+}</div></div><div><br></div><div>I think you mean >= 10, otherwise, it'll just output an inexplicable "A" with no "0x".</div>
<div><br></div><div>+const char endl = '\n';</div><div><br></div><div>why? either just write '\n' in the code, or as in many cases it seems in the code, just include "\n" in the string literal instead of causing a pointless extra call to raw_ostream's operator<<. e.g. instead of</div>
<div>+ Out << " Type: " << endl;</div><div>just do</div><div>+ Out << " Type: \n"</div><div><div><br></div><div>--Sean Silva<br><br><div class="gmail_quote">
On Tue, Jun 12, 2012 at 8:35 AM, Marshall Clow <span dir="ltr"><<a href="mailto:mclow.lists@gmail.com" target="_blank">mclow.lists@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div>On Jun 9, 2012, at 2:44 AM, Michael Spencer wrote:<br>
<br>
> On Fri, Jun 8, 2012 at 11:09 AM, Marshall Clow <<a href="mailto:mclow.lists@gmail.com" target="_blank">mclow.lists@gmail.com</a>> wrote:<br>
>> [ resending to llvm-commits instead of -dev ]<br>
>><br>
>> This patch adds a tool called 'obj2yaml', which takes an object file, and produces a YAML representation of the file.<br>
>> This is a companion tool to Michael Spencer's proposed "yaml2obj" tool.<br>
>><br>
>> This tool processes only COFF files; the plan is to enhance it to handle other object formats, as well as object archives.<br>
>> The primary uses for this tool is expected to be:<br>
>> 1) Debugging object files that are produced by llvm and lld.<br>
>> 2) Automated testing suites.<br>
>><br>
>> The goal is to have obj2yaml and yaml2obj be inverse operations, so that one can test either by using the other.<br>
>> So, both obj -> yaml -> obj and yaml -> obj -> yaml are expected to be idempotent.<br>
>><br>
>> -- Marshall<br>
>><br>
>> Marshall Clow Idio Software <mailto:<a href="mailto:mclow.lists@gmail.com" target="_blank">mclow.lists@gmail.com</a>><br>
>><br>
>> A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait).<br>
>> -- Yu Suzuki<br>
><br>
> The design looks good, however there are quite a few LLVM style issues.<br>
<br>
</div>Extensively reworked. Now uses the facilities in llvm/Object/COFF.h for byte-sex awareness.<br>
Still a few 80 column issues, and the relocation dumping is not done - but since yaml2obj doesn't build relocation entries yet, this may be moot. ;-)<br>
<div><div><br>
-- Marshall<br>
<br>
Marshall Clow Idio Software <mailto:<a href="mailto:mclow.lists@gmail.com" target="_blank">mclow.lists@gmail.com</a>><br>
<br>
A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait).<br>
-- Yu Suzuki<br>
</div></div><br>_______________________________________________<br>
llvm-commits mailing list<br>
<a href="mailto:llvm-commits@cs.uiuc.edu" target="_blank">llvm-commits@cs.uiuc.edu</a><br>
<a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits</a><br>
<br></blockquote></div><br></div></div>
</div>