Inspection tool run
I've been trying to get Boost.Math off the Sinners List on the Inspection Report at http://boost.cowic.de/rc/inspect-trunk.html#math. It last ran 25 July 12, so I can't see if I have succeeded yet. When can I expect it to run again? Paul --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow@hetp.u-net.com
On Aug 9, 2012, at 1:22 AM, Paul A. Bristow <pbristow@hetp.u-net.com> wrote:
I've been trying to get Boost.Math off the Sinners List on the Inspection Report at http://boost.cowic.de/rc/inspect-trunk.html#math.
It last ran 25 July 12, so I can't see if I have succeeded yet.
When can I expect it to run again?
Paul - It's easy to build and run the inspect tool yourself. $ cd $BOOST/tools/inspect/build ; $BOOST/b2 $ cd $BOOST ; dist/bin/inspect > /tmp/inspect.html -- Marshall Marshall Clow Idio Software <mailto:mclow.lists@gmail.com> A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait). -- Yu Suzuki
-----Original Message----- From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Marshall Clow Sent: Thursday, August 09, 2012 2:11 PM To: boost@lists.boost.org Subject: Re: [boost] Inspection tool run
On Aug 9, 2012, at 1:22 AM, Paul A. Bristow <pbristow@hetp.u-net.com> wrote:
I've been trying to get Boost.Math off the Sinners List on the Inspection Report at http://boost.cowic.de/rc/inspect-trunk.html#math.
It last ran 25 July 12, so I can't see if I have succeeded yet.
When can I expect it to run again?
Paul - It's easy to build and run the inspect tool yourself.
$ cd $BOOST/tools/inspect/build ; $BOOST/b2 $ cd $BOOST ; dist/bin/inspect > /tmp/inspect.html
Thanks - so I now I can know my un-expiated sins :-( Paul PS I copied the inspect.exe to the folder containing my b2 (so in my path) and ran it in the local boost/math/ folder thereby by just listing the Boost.Math sins. I have enough sins of my own to view without wanting to view all the other authors sins ;-) (Might this be suggested as good practice - or have I failed to RTFM, as ever ...)
On 9 August 2012 14:46, Paul A. Bristow <pbristow@hetp.u-net.com> wrote:
PS I copied the inspect.exe to the folder containing my b2 (so in my path)
and ran it in the local boost/math/ folder thereby by just listing the Boost.Math sins.
I have enough sins of my own to view without wanting to view all the other authors sins ;-)
(Might this be suggested as good practice - or have I failed to RTFM, as ever ...)
I think it's generally best to run it from the root so that the link checker can check links properly. Although running from a subdirectory will certainly be faster if you're just fixing the headers.
-----Original Message----- From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Daniel James Sent: Thursday, August 09, 2012 3:05 PM To: boost@lists.boost.org Subject: Re: [boost] Inspection tool run
On 9 August 2012 14:46, Paul A. Bristow <pbristow@hetp.u-net.com> wrote:
PS I copied the inspect.exe to the folder containing my b2 (so in my path)
and ran it in the local boost/math/ folder thereby by just listing the Boost.Math sins.
I have enough sins of my own to view without wanting to view all the other authors sins ;-)
(Might this be suggested as good practice - or have I failed to RTFM, as ever ...)
I think it's generally best to run it from the root so that the link checker can check links properly. Although running from a subdirectory will certainly be faster if you're just fixing the headers.
Yes - that does check more - but takes much longer. It also brings in lots of local files that are not in SVN :-( Is there any easy way of the Inspection tool not 'seeing' those? Paul PS I found it difficult to find some pesky tabs - could we print the line number? --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow@hetp.u-net.com
On Thu, Aug 9, 2012 at 4:22 AM, Paul A. Bristow <pbristow@hetp.u-net.com> wrote:
I've been trying to get Boost.Math off the Sinners List on the Inspection Report at http://boost.cowic.de/rc/inspect-trunk.html#math.
It last ran 25 July 12, so I can't see if I have succeeded yet.
When can I expect it to run again?
Hum... It should run every day at 9:00 AM Eastern US time. Ouch! Getting an I/O error. I tried to do a fresh checkout of trunk and got the same error: svn checkout https://svn.boost.org/svn/boost/trunk boost-inspect-trunk2 ... A boost-inspect-trunk2/boost/asio/detail/dev_poll_reactor.hpp A boost-inspect-trunk2/boost/asio/detail/win_thread.hpp A boost-inspect-trunk2/boost/asio/detail/win_iocp_handle_write_op.hpp svn: E200030: disk I/O error, executing statement 'COMMIT TRANSACTION;' svn: E200030: sqlite: disk I/O error svn: E200030: sqlite: disk I/O error svn: E200030: no such savepoint: s28750, executing statement 'RELEASE s28750' svn: E200030: no such savepoint: s28750, executing statement 'ROLLBACK TO s28750' I'll try this on another machine to see if it reoccurs. --Beman
-----Original Message----- From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Beman Dawes Sent: Thursday, August 09, 2012 8:48 PM To: boost@lists.boost.org Subject: Re: [boost] Inspection tool run
On Thu, Aug 9, 2012 at 4:22 AM, Paul A. Bristow <pbristow@hetp.u-net.com> wrote:
I've been trying to get Boost.Math off the Sinners List on the Inspection Report at http://boost.cowic.de/rc/inspect-trunk.html#math.
It last ran 25 July 12, so I can't see if I have succeeded yet.
When can I expect it to run again?
Hum... It should run every day at 9:00 AM Eastern US time.
Ouch! Getting an I/O error.
I tried to do a fresh checkout of trunk and got the same error:
svn checkout https://svn.boost.org/svn/boost/trunk boost-inspect-trunk2 ... A boost-inspect-trunk2/boost/asio/detail/dev_poll_reactor.hpp A boost-inspect-trunk2/boost/asio/detail/win_thread.hpp A boost-inspect-trunk2/boost/asio/detail/win_iocp_handle_write_op.hpp svn: E200030: disk I/O error, executing statement 'COMMIT TRANSACTION;' svn: E200030: sqlite: disk I/O error svn: E200030: sqlite: disk I/O error svn: E200030: no such savepoint: s28750, executing statement 'RELEASE s28750' svn: E200030: no such savepoint: s28750, executing statement 'ROLLBACK TO s28750'
I'll try this on another machine to see if it reoccurs.
Thanks for looking into this. Paul
Le 09/08/12 10:22, Paul A. Bristow a écrit :
I've been trying to get Boost.Math off the Sinners List on the Inspection Report at http://boost.cowic.de/rc/inspect-trunk.html#math.
Hi, in the report there are a lot of inspection errors that are extracted from the compiler warnings, such as libs\thread\test\no_implicit_assign_from_lvalue_thread.cpp: (line 19) Non-ASCII: //../../../boost/system/error_code.hpp:214:36: warning: Ôboost::system::posix_categoryÕ defined but not used [-Wunused-variable libs\thread\test\no_implicit_move_from_lvalue_thread.cpp: (line 18) Non-ASCII: //../../../boost/system/error_code.hpp:214:36: warning: Ôboost::system::posix_categoryÕ defined but not used [-Wunused-variable libs\thread\test\sync\conditions\condition_variable\assign_fail.cpp: (line 32) Non-ASCII: //../../../boost/system/error_code.hpp:214:36: warning: Ôboost::system::posix_categoryÕ defined but not used [-Wunused-variable libs\thread\test\sync\conditions\condition_variable\copy_fail.cpp: (line 31) Non-ASCII: //../../../boost/system/error_code.hpp:214:36: warning: Ôboost::system::posix_categoryÕ defined but not used [-Wunused-variable libs\thread\test\sync\conditions\condition_variable_any\assign_fail.cpp: (line 31) Non-ASCII: //../../../boost/system/error_code.hpp:214:36: warning: Ôboost::system::posix_categoryÕ defined but not used [-Wunused-variable libs\thread\test\sync\conditions\condition_variable_any\copy_fail.cpp: (line 31) Non-ASCII: //../../../boost/system/error_code.hpp:214:36: warning: Ôboost::system::posix_categoryÕ defined but not used [-Wunused-variable libs\thread\test\sync\futures\future\copy_assign_fail.cpp: What is going wrong? Best, Vicente
On 12 August 2012 10:48, Vicente J. Botet Escriba <vicente.botet@wanadoo.fr> wrote:
in the report there are a lot of inspection errors that are extracted from the compiler warnings, such as
libs\thread\test\no_implicit_assign_from_lvalue_thread.cpp: (line 19) Non-ASCII: //../../../boost/system/error_code.hpp:214:36: warning: Ôboost::system::posix_categoryÕ defined but not used [-Wunused-variable
The warnings have been copied into comments in the file 'no_implicit_assign_from_lvalue_thread.cpp', and contain non-ascii characters (the single quotes).
Le 12/08/12 11:55, Daniel James a écrit :
On 12 August 2012 10:48, Vicente J. Botet Escriba <vicente.botet@wanadoo.fr> wrote:
in the report there are a lot of inspection errors that are extracted from the compiler warnings, such as
libs\thread\test\no_implicit_assign_from_lvalue_thread.cpp: (line 19) Non-ASCII: //../../../boost/system/error_code.hpp:214:36: warning: Ôboost::system::posix_categoryÕ defined but not used [-Wunused-variable The warnings have been copied into comments in the file 'no_implicit_assign_from_lvalue_thread.cpp', and contain non-ascii characters (the single quotes).
Oh, thanks. I should inspect the files before posting ;-) Best, Vicente
Hi, Currently I'm working on a spatial index and was considering using methods templates until I've tested it on GCC. Consider the following simple example: template <typename T1, typename T2> struct boom {}; struct aaa { template <typename T> void boom() {} }; template <typename T> struct bbb { void test() { T t; t.template boom<int>(); // <- here GCC resolves struct boom } }; This kind of code is already used in Boost. Try to compile this on GCC: template <typename T1, typename T2> struct apply {}; template <typename T1, typename T2> struct replace {}; template <typename T1, typename T2> struct remove_if {}; template <typename T1, typename T2> struct transfer {}; template <typename T1, typename T2> struct convert_to_container {}; template <typename T1, typename T2> struct push_match {}; #include <boost/optional.hpp> #include <boost/ptr_container/ptr_array.hpp> #include <boost/ptr_container/ptr_list.hpp> #include <boost/assign.hpp> #include <boost/xpressive/xpressive.hpp> This is not the case on VC++ or clang. Also, not the case if structs can't be found by the compiler (e.g. are in the different namespace). But what if some library defines some struct template and includes other library which defines method template with the same name? Is GCC standard-compliant in this case? Regards, Adam
On Sunday 30 September 2012 19:18:32 Adam Wulkiewicz wrote:
Hi,
Currently I'm working on a spatial index and was considering using methods templates until I've tested it on GCC. Consider the following simple example:
template <typename T1, typename T2> struct boom {};
struct aaa { template <typename T> void boom() {} };
template <typename T> struct bbb { void test() { T t; t.template boom<int>(); // <- here GCC resolves struct boom } };
Assuming bbb<aaa>::test() is called, confirmed with GCC 4.7. Looks like a compiler bug to me. The "t" expression is type dependent on the template parameter (14.6.2.2), which makes the "boom" method call dependent as well. Name lookup should be performed at the point of the template instantiation, where aaa::boom should be found. This should probably be reported to the GCC dev team.
AMDG On 09/30/2012 10:18 AM, Adam Wulkiewicz wrote:
Hi,
Currently I'm working on a spatial index and was considering using methods templates until I've tested it on GCC. Consider the following simple example:
template <typename T1, typename T2> struct boom {};
struct aaa { template <typename T> void boom() {} };
template <typename T> struct bbb { void test() { T t; t.template boom<int>(); // <- here GCC resolves struct boom } };
<snip>
This is not the case on VC++ or clang. Also, not the case if structs can't be found by the compiler (e.g. are in the different namespace). But what if some library defines some struct template and includes other library which defines method template with the same name? Is GCC standard-compliant in this case?
Definitely not. The names should be resolved as members of the class, not using unqualified lookup. In Christ, Steven Watanabe P.S. Please don't reply to an existing thread with new subject. Start a new thread instead.
participants (8)
-
Adam Wulkiewicz -
Andrey Semashev -
Beman Dawes -
Daniel James -
Marshall Clow -
Paul A. Bristow -
Steven Watanabe -
Vicente J. Botet Escriba