From: Larry Evans (jcampbell3_at_[hidden])
Date: 2003-04-23 09:52:21
John Torjo wrote:
> Since there seems to be some confusion about this, I will post the message I
> sent to Reece in the first place:
>
> Hi Reece,
>
> I've looked at col_io library -
> http://groups.yahoo.com/group/boost/files/col_io/, since I've written a
> (note: at that time, I thought Reece developed it)
> library that allows thread-safe logging
> (http://groups.yahoo.com/group/boost/files/thread_safe_log.zip).
>
> I wanted to see if they "get along" and unfortunately the answer is
> something like "sometimes yes, sometimes no".
>
> The point is that I really like the col_io library.
Thanks very much.
>
> I would have a few remarks if it's ok with you:
> 1. at this time, you donot care about formatting.
> So, something like this will most likely not work as expected:
>
> std::ofstream out("out.txt");
> marg_stream m(out);
> locale l( "german");
> out.imbue( l);
> m << 5.23 << std::endl;
I'll have to try it, but I think the ctor either flushes the
out.rdbuf or copies it to m.rdbuf; hence, it MIGHT work.
>
> 2. binding marg_stream to a std::ostream& couples them too much IMHO
> (that is, the marg_stream variable is coupled to the other stream)
> This actually came to me when I wanted to use col_io together with my
> thread_safe_log library.
[snip]
OK.
> Both indent_ostream and indent_ostream_wrapper are meant to be used as RAII
> (resource acquisition is initialization) objects - they are destroyed as
> soon as they are not needed anymore.
[snip]
>
> Please let me know what you think.
>
> Best,
> John
John, I haven't looked closely at your indenter_streambuf below, but
it looks similar to marg_ostreambuf, which I mentioned in my previous post.
Have you tried using marg_ostreambuf with ostringstream as shown in
the main.cpp code which was also mentioned in my previous post?