configuring log4cplus failed about log4cplus HOT 7 CLOSED

First off, you really need to be more precise as to what does not work. I my crystal ball is broken. :)

            LOG4CPLUS_TEXT
            (
                "log4cplus.rootLogger=TRACE, stdout\n"
                "log4cplus.appender.stdout=log4cplus::ConsoleAppender\n"
                "log4cplus.appender.stdout.Threshold=TRACE\n"
                "log4cplus.appender.stdout.Target=System.out\n"
                "log4cplus.appender.stdout.layout=log4cplus::PatternLayout\n"
                "log4cplus.appender.stdout.layout.ConversionPattern=%D{%Y-%m-%d %H:%M:%S,%q} %p %t %c %L: %m%n\n"
            )

This will not work, IIRC. You have to wrap each individual string literal with LOG4CPLUS_TEXT().

If you are using master branch then you should use the log4cplus::Initializer approach instead of calling log4cplus::initialize() yourself. If you are using earlier branches, well, you are stuck with this call.

static intmax_t dummy = initialize();

Do not do this! Initialize at the top of your main() function instead (cf. static initialization order fiasco). Also, see my previous comment.

log4cplus::Logger LOGGER = log4cplus::Logger::getInstance(LOG4CPLUS_TEXT("int main(void)"));

There is no point naming your logger with full signature of your function. Since you are using PatternLayout you can use the %M specifier to get function name in the output. Better name the logger just "main".

from log4cplus.

Oh, and I have noticed two more issues:

::log4cplus::tistringstream const log4cplusProperties

Remove the const. It is obviously not a constant object.

* const_cast<::log4cplus::tistringstream *>(& log4cplusProperties)

Do not do this. Just remove the const on log4cplusProperties at its declaration point.

from log4cplus.

Hi, thanks for your comments.

1. It would be great if example code posted in http://sourceforge.net/p/log4cplus/wiki/CodeExamples/ is updated. ::log4cplus::initialize() was coped from there.
2. static intmax_t dummy = initialize(); has no problems. I have already kept the initialization order in mind.
3. Thanks for telling me that %M could get the function names. I will have a try.
4. It does not matter whether I specified const here.

By the way, http://sourceforge.net/p/log4cplus/wiki/Home/ hints that both 1.2.0-RC5 and 1.2.0-RC6 are released. But nothing was changed in https://github.com/log4cplus/log4cplus/releases.

Yours sincerely,
Xiaoshuang

from log4cplus.

It would be great if example code posted in http://sourceforge.net/p/log4cplus/wiki/CodeExamples/ is updated. ::log4cplus::initialize() was coped from there.

Good point. I will do that. TBH, I often forget these examples are there. :)

By the way, http://sourceforge.net/p/log4cplus/wiki/Home/ hints that both 1.2.0-RC5 and 1.2.0-RC6 are released. But nothing was changed in https://github.com/log4cplus/log4cplus/releases.

I have no idea how this part of the GitHub web works. Maybe I forgot to push the tags. I use GitHub only for the Git hosting and issues. Everything else is on SF.

from log4cplus.

Hi Václav, 1.2.0-RC6 in http://sourceforge.net/p/log4cplus/wiki/Home/ and http://sourceforge.net/p/log4cplus/news/2015/11/log4cplus-120-rc6-released/ shall be 1.1.3-RC6. Thanks.

from log4cplus.

By the way, http://sourceforge.net/p/log4cplus/wiki/Home/ hints that both 1.2.0-RC5 and 1.2.0-RC6 are released. But nothing was changed in https://github.com/log4cplus/log4cplus/releases.

This should be fixed. I always forget to push the tags.

Hi Václav, 1.2.0-RC6 in http://sourceforge.net/p/log4cplus/wiki/Home/ and http://sourceforge.net/p/log4cplus/news/2015/11/log4cplus-120-rc6-released/ shall be 1.1.3-RC6. Thanks.

This is fixed now.

from log4cplus.

I am opening #160 to keep track of the code examples update.

from log4cplus.