cppcms-users — Disussions of users of CppCMS

[Cppcms-users] Test - Ignore

[Artyom B. <art...@ya...>]

Ignore Please

Re: [Cppcms-users] Creating a database web-frontend with interactive search in CppCMS

[augustin <aug...@ov...>]

Alec,

Good luck with your exams, and have fun with cppcms.

Blessings,

Augustin.



-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.

Re: [Cppcms-users] Creating a database web-frontend with interactive search in CppCMS

[Alec T. <ale...@gm...>]

Okay, well I'm amidst exams ATM, I'll start coding (heavily) in 20
days. In the meantime, I'll play around with examples and see if I can
get my template done =]

On Mon, Jun 6, 2011 at 4:35 AM, Artyom Beilis <art...@ya...> wrote:
>> From: Alec Taylor <ale...@gm...>
>
>>
>> I'll Checkout from the SVN and start coding in
>> ~week
>
> Actually it is better to use the latest stable version unless you
> have some bugs fixed in svn.
>
> Currently it is quite close to svn-trunk.
>
> Because svn trunk still can become broken once in a while
> during development.
>
> In any case svn version is quite stable.
>
>>
>> Here are some  names (off the top of my head), add your own below (or
>> to the wiki  :P):
>>
>> [list of names]
>>
>
> Actually there was a good name once as potential replacement but it was decided
> not to change cppcms as it is already well known name and
> replacement of the name should also be followed by namespace
> renaming (causing breaking existing software)
>
> So it was decided to keep not-so-good-but-well-known-name CppCMS
>
> Artyom
>
> ------------------------------------------------------------------------------
> Simplify data backup and recovery for your virtual environment with vRanger.
> Installation's a snap, and flexible recovery options mean your data is safe,
> secure and there when you need it. Discover what all the cheering's about.
> Get your free trial download today.
> http://p.sf.net/sfu/quest-dev2dev2
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>

Re: [Cppcms-users] Creating a database web-frontend with interactive search in CppCMS

[Artyom B. <art...@ya...>]

> From: Alec Taylor <ale...@gm...>

> 
> I'll Checkout from the SVN and start coding in
> ~week 

Actually it is better to use the latest stable version unless you
have some bugs fixed in svn.

Currently it is quite close to svn-trunk.

Because svn trunk still can become broken once in a while
during development.

In any case svn version is quite stable.

> 
> Here are some  names (off the top of my head), add your own below (or
> to the wiki  :P):
>
> [list of names]
> 

Actually there was a good name once as potential replacement but it was decided
not to change cppcms as it is already well known name and
replacement of the name should also be followed by namespace
renaming (causing breaking existing software)

So it was decided to keep not-so-good-but-well-known-name CppCMS

Artyom

Re: [Cppcms-users] Creating a database web-frontend with interactive search in CppCMS

[Alec T. <ale...@gm...>]

Thanks for the info, I'll Checkout from the SVN and start coding in
~week (I'm a student currently half-way through my IT Degree)

Here are some names (off the top of my head), add your own below (or
to the wiki :P):

webC++

sharpC++

netC++

speedC++

NetScalC++

ScalC++

UncppNet

accpt-Net (accpt means accompaniment, add an extra p in for fun)

ScalNet

NetScal



On Sun, Jun 5, 2011 at 9:45 PM, augustin <aug...@ov...> wrote:
>
> Speaking of:
>
>> Keep in mind that the questions you may have, others may also have them in
>>  the future. Your help in ensuring that each question need only be asked
>>  once (by improving the wiki as we go along) will be very appreciated.
>
>
>> > BTW: Are there any plans to disambiguate the name of this project,
>> > since it isn't a CMS?
>
> http://art-blog.no-ip.info/wikipp/en/page/faq/diff/19vs20
>
>
> Augustin.
>
>
>
> --
> Friends: http://www.reuniting.info/
> My projects:
> http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/
> http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
> http://openteacher.info/ http://minguo.info/
> http://www.wechange.org/ http://searching911.info/
>
>
>
>
>
>
>
>
>
>
>
>
> .
>
> ------------------------------------------------------------------------------
> Simplify data backup and recovery for your virtual environment with vRanger.
> Installation's a snap, and flexible recovery options mean your data is safe,
> secure and there when you need it. Discover what all the cheering's about.
> Get your free trial download today.
> http://p.sf.net/sfu/quest-dev2dev2
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>

Re: [Cppcms-users] Creating a database web-frontend with interactive search in CppCMS

[augustin <aug...@ov...>]

Speaking of:

> Keep in mind that the questions you may have, others may also have them in
>  the future. Your help in ensuring that each question need only be asked
>  once (by improving the wiki as we go along) will be very appreciated.

 
> > BTW: Are there any plans to disambiguate the name of this project,
> > since it isn't a CMS?

http://art-blog.no-ip.info/wikipp/en/page/faq/diff/19vs20


Augustin.



-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.

Re: [Cppcms-users] Creating a database web-frontend with interactive search in CppCMS

[augustin <aug...@ov...>]

Alec,

Welcome back on the list.

First of all, when we reply to you, we assume you have at least a fair amount 
of experience with C++ (you were asking about PHP and C++ frameworks at the 
same time). If you don't, that's ok to (I am myself a newbie) but then be 
ready to spend a fair amount of time to learn the basics.

On Sunday 05 June 2011 06:51:39 pm Alec Taylor wrote:
> Augustin: Thanks for the suggestions, however I've found the examples
> to be broken (not showing anything other than title), i.e.:
> http://art-blog.no-ip.info/cppcms_ref_v0_99/hello_8cpp-example.html

See:
http://art-blog.no-ip.info/wikipp/en/page/cppcms_1x_tut_hello

Download the source code of cppcms. See the sourceforge project for the exact 
download procedure. cppcms uses svn.

In the source code, see within the example folder. Compile the examples and 
run them. Study their code. Copy them elsewhere and modify the code to start 
playing with the various features.



> Please tell me where I can find CppCMS documentation.

http://art-blog.no-ip.info/wikipp/en/page/main

Registration to the wiki is easy. 
Keep in mind that the questions you may have, others may also have them in the 
future. Your help in ensuring that each question need only be asked once (by 
improving the wiki as we go along) will be very appreciated.

> BTW: Are there any plans to disambiguate the name of this project,
> since it isn't a CMS?

The developer asked for suggestions for a better name 3 years ago, but 
received 0 replies:
http://art-blog.no-ip.info/cppcms/blog/post/23
I am not sure Artyom would be open today to changing the name.... provided we 
have a nice alternative!

See also the FAQ:
"If CppCMS is not a CMS, where does this name come from?"
http://art-blog.no-ip.info/wikipp/en/page/faq

See the developer's blog for more info:
http://art-blog.no-ip.info/cppcms/blog


Have fun!


Augustin.


-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.

Re: [Cppcms-users] Creating a database web-frontend with interactive search in CppCMS

[Alec T. <ale...@gm...>]

Thanks for the replies (and sorry for this late one)

Augustin: Thanks for the suggestions, however I've found the examples
to be broken (not showing anything other than title), i.e.:
http://art-blog.no-ip.info/cppcms_ref_v0_99/hello_8cpp-example.html

Aris: Thanks, I'll take a look. My search results will be ranked by
accuracy, you see the results will be very statistical (66% correct)
&etc

Please tell me where I can find CppCMS documentation.

Thanks,

Alec Taylor

BTW: Are there any plans to disambiguate the name of this project,
since it isn't a CMS?

On Fri, Apr 29, 2011 at 11:06 AM, Aris Setyawan <ari...@gm...> wrote:
> Hi Alec,
>
> For fulltext search, you must use fulltext index. If you are mysql
> user, use spynx index (sphinxsearch.com), for postgresql user, use
> tsearch2 index and for sqlite user use FTS3/4 index.
>
> I'm now also developing web search app. For maximizing performance,
> I'm not using javascript at all. html + css is enough for me.
>
> Don't forget to use ranking in your search result.
>
> On Thu, Apr 28, 2011 at 8:34 PM, augustin <aug...@ov...> wrote:
>> Good Evening Alec,
>>
>> On Thursday 28 April 2011 09:17:37 pm Alec Taylor wrote:
>>> Please tell me if it's possible to do in CppCMS, and if so, how.
>>
>> Yes, it's possible. The cppcms library is well conceived, and I find the cppdb
>> (DB layer) library especially easy to use.
>>
>> How? Well, check the wiki for simple tutorials and documentation , and check
>> all the examples delivered with the source code. That will get you started in
>> understanding how to use the libraries.
>>
>> The templating layer will allow you to achieve the look you are after, like in
>> your mockup. Again, check the wiki for more information.
>>
>> Blessings,
>>
>> Augustin.
>>
>>
>>
>>
>> --
>> Friends: http://www.reuniting.info/
>> My projects:
>> http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/
>> http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
>> http://openteacher.info/ http://minguo.info/
>> http://www.wechange.org/ http://searching911.info/
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> .
>>
>> ------------------------------------------------------------------------------
>> WhatsUp Gold - Download Free Network Management Software
>> The most intuitive, comprehensive, and cost-effective network
>> management toolset available today.  Delivers lowest initial
>> acquisition cost and overall TCO of any competing solution.
>> http://p.sf.net/sfu/whatsupgold-sd
>> _______________________________________________
>> Cppcms-users mailing list
>> Cpp...@li...
>> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>>
>
> ------------------------------------------------------------------------------
> WhatsUp Gold - Download Free Network Management Software
> The most intuitive, comprehensive, and cost-effective network
> management toolset available today.  Delivers lowest initial
> acquisition cost and overall TCO of any competing solution.
> http://p.sf.net/sfu/whatsupgold-sd
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
>

Re: [Cppcms-users] json_rpc and json questions

[Artyom B. <art...@ya...>]

Small note, instead of:

  return_response("Not authorised");
  return false;

You can also return a error on HTTP level (not JSON-RPC):

  context().response().make_error_response(cppcms::http::response::forbidden);
  return false;

But I think this is generally bad idea to make a error response on transport
level rather then on JSON-RPC level.

Artyom

Re: [Cppcms-users] json_rpc and json questions

[Artyom B. <art...@ya...>]

> So if I do just that, the request parsing is made ​​twice.  One for obtain 
> method name for check_privileges and another into json_rpc_server::main() 
> function.  I guess there are no other solution in actual design. Is it true? 

If the privilege control is made on the level of each specific method like

   get(id) // for all
   update(id,new_value) // for privilege user

Then you should check it inside method and not outside, 

  void update(std::string const &id,my_object const &vale) 
  {
     if(!check_permissions()) {
           return;
     }
     objects_[id]=value;
  }

  void get(std::string const &id)
  {
     return_result(objects_[id]);
  }

  bool check_permissions()
  {
       if(session().get("role","nobody")!="admin") {
           return_response("Not authorised");
           return false;
       }
       return true;
  }

The point you should not decouple parts of code that
are truly connected.  This way you know from looking
to the method its permissions.

Consider you add in future new method  add() that is for priviliged users
thus you have to add permissions control in one place and method
itself in other.

It is dangerous practice.

Artyom

Re: [Cppcms-users] json_rpc and json questions

[Daniel V. <chi...@gm...>]

On Thu, 2011-05-19 at 00:07 -0700, Artyom Beilis wrote:

> 
> This is why it is virtual. You want pre/post-actions?
> 
>   void main(std::string param)
>   {
>       // some pre-action
>       cppcms::rpc::json_rpc_server::main(param);
>       // some post-actions
>   }
> 
> 


The privilege validations depends on session() object (user and his
rights) and method name. Then:

  void main(std::string param)
  {
     if( check_privileges(context()) )
     {
        cppcms::rpc::json_rpc_server::main(param);
     }
     else { 
         // not authorized
     }   
  }


So if I do just that, the request parsing is made ​​twice.  One for
obtain method name for check_privileges and another into
json_rpc_server::main() function.  I guess there are no other solution
in actual design. Is it true? 

One solution is give to the user of json_rpc_server class the
possibility to add pre action after parsing json request but before
dispatching. 

Thanks.

Re: [Cppcms-users] CPPDB and Oracle support

[Artyom B. <art...@ya...>]

----- Original Message ----
> From: Daniel Vallejos <chi...@gm...>
> To: cpp...@li...
> Sent: Thu, May 19, 2011 7:17:31 AM
> Subject: [Cppcms-users] CPPDB and Oracle support
> 
> Hello.
> 
> Oracle Support is on cppdb roadmap?
> 

cppdb supports ODBC. So given a good ODBC driver it should support
Oracle or any other RDBMS.

However:

1. I do not have an installation of Oracle database itself.
2. I have no experience with OCI API.

So I'm not going to implement one myself.

So if you need a native driver for Orcale you need to
implement one - it should not be that hard if you know OCI.

Or you need to find somebody who is willing to contribute one.

In any case you can look on ODBC and if you can successfully use
ODBC with Oracle please report me.

There is also a Unit test (that may need some tweaks for Orcale)
that can give you a way to know of ODBC/Oracle works well.

I'd would be glad if you can test cppdb with Oracle and maybe
tweak a Unit test a little if needed to make sure it passes.

In any case if you need a good library for Oracle that
supports other RDMBSs as well you can take a look on SOCI.
>From my experience it does not support other DB too well
and it does not have CppDB's features like caching of prepared
statements.


Best,
 Artyom

Re: [Cppcms-users] json_rpc and json questions

[Artyom B. <art...@ya...>]

>Question 1.
>
>I am implementing a json rpc server (cppcms::rpc::json_rpc_server) with 
>privileges.
>
>One possible solution is for each rpc method do the following:
>
>void my_rpc_server::get_sensible_data()
>{
>if ( check_privileges() ) {
>
>// method implementation 
>//...
>}
>else {
>return_error("not authorized");
>} 
>};

It depends on the level you check the privileges and how do you want to handle 
an answer
and protocol.

If you use HTTP session to check the privileges then you can do different 
things:


1. Return 403 Forbidden status on HTTP level before you even access to RPC 
service.

   You can overload the main function and check permissions, but this is 
generally
   limited for entire service not specific methods.

   This is something like "prevent executable from running as you don't have 
permissions
   to run it at OS leve"

2. Return a error code for specific call on the protocol level.

   This is something like "tell the chroot() or unlink() system calls return
   error code EACCES"

It is really a question of your design and the design of the protocol.

If some methods are forbidden, return error code is best, if all the service
is forbidden return 403 as above.

>
>The problem is that I think this solution is not good because I would like to 
>decouple the privilege system of the rpc methods definitions.
>
>Another theoretical solution would be extend virtual main method to add desired 

>pre actions:
>
>class my_rcp : public cppcms::rpc::json_rpc_server
>{
>virtual void main(std::string);
>};
>
>
>void main(std::string //unused)
>{
>json_rpc_request req = parse_json_rpc_request(context()); // Ideally this method 
>
>should be protected and implemented in json_rpc_server.
>

This is why it is virtual. You want pre/post-actions?

  void main(std::string param)
  {
      // some pre-action
      cppcms::rpc::json_rpc_server::main(param);
      // some post-actions
  }


>// Added pre actions
>
>if ( check_privileges(req.method_name) )
>{
>dispatch_request(req);   // Ideally this method should be protected and 
>implemented in json_rpc_server.
>}
>else {
>return_error("not authorised");
>}
>}
>
>
>AFAIK that is not posible with current json_rpc_server implementation. 
>

AFAIK it should work fine, just overload main.

>
>Question 2.
>
>I can't implement a json rpc method with a json::value argument, for example the 
>
>following example not compile:
>
>void my_rpc_server::my_method(cppcms::json::value v)
>{
>}
>
>One posible workaround is define a json::value to json::value trait:
>
>namespace cppcms { namespace json {
>    template<>
>    struct traits<value> {               
>        static value get(value const &v) {                   
>            return v;      
>        }                   
>        static void set(value &v,value const &in) {                   
>            v = in;          
>        }                   
>    };
>}} // end cppcms::json
>
>But I guest it is not optimal because nedless copy overhead.
>


If you want "native" json parameter you should create method
that receives an array as parameter

my_method(cppcms::json::array const &v)
{
}

And the use something like boost::bind to create a reference.

   bind("sum",boost::bind(&json_service::my_method,this,_1),method_role);

Or if you wan't want to use boost

   struct binder {
        binder(my_service &s) : s_(&s)
        void operator()(cppcms::json::array const &v) const
        {
          s_->my_method(v);
        }
        my_service *s_;
   };

And then

   bind("sum",binder(this),method_role);

Actually you bring a good point, I need to provide such
binder my default something like: cppcms::rpc::json_native_method
in addition to json_method

Fill a feature request for the last one so I will not forget.

Best.
  Artyom

[Cppcms-users] CPPDB and Oracle support

[Daniel V. <chi...@gm...>]

Hello.

Oracle Support is on cppdb roadmap?

Thank you.

[Cppcms-users] json_rpc and json questions

[Daniel V. <chi...@gm...>]

Hello,

Question 1.

I am implementing a json rpc server (cppcms::rpc::json_rpc_server) with
privileges.

One possible solution is for each rpc method do the following:

void my_rpc_server::get_sensible_data()
{
	if ( check_privileges() ) {

		// method implementation		
		//...
	}
	else {
		return_error("not authorized");
	}	
};

The problem is that I think this solution is not good because I would
like to decouple the privilege system of the rpc methods definitions.

Another theoretical solution would be extend virtual main method to add
desired pre actions:

class my_rcp : public cppcms::rpc::json_rpc_server
{
	virtual void main(std::string);
};


void main(std::string //unused)
{
	json_rpc_request req = parse_json_rpc_request(context()); // Ideally
this method should be protected and implemented in json_rpc_server.

	// Added pre actions

	if ( check_privileges(req.method_name) )
	{
		dispatch_request(req);   // Ideally this method should be protected
and implemented in json_rpc_server.
	}
	else {
		return_error("not authorised");
	}
}


AFAIK that is not posible with current json_rpc_server implementation. 

What is the best approach to add pre actions (ie: check privileges) to 
rpc method invocation?.

Question 2.

I can't implement a json rpc method with a json::value argument, for
example the following example not compile:

void my_rpc_server::my_method(cppcms::json::value v)
{
}

One posible workaround is define a json::value to json::value trait:

namespace cppcms { namespace json {
    template<>
    struct traits<value> {               
        static value get(value const &v) {                   
            return v;      
        }                   
        static void set(value &v,value const &in) {                   
            v = in;          
        }                   
    };
}} // end cppcms::json

But I guest it is not optimal because nedless copy overhead.

Thank you.

Re: [Cppcms-users] CppDb Errors

[Artyom B. <art...@ya...>]

> From: Daniel Vallejos <chi...@gm...>
> 
> Hello.
> 
> I need catch common sql errors in cppdb, such as duplicate key  insert
> error or other constraint errors.
> 
> How to accomplish this in a  portable way with  cppdb?
> 

Just catch cppdb_error exception.

There is no portable way to get specific error.

Also it is better to check the constraints before you execute
something and not relay on them in normal flow.

i.e.:

   cppdb::translation tr(sql);

   cppdb::result r;
   sql << "SELECT id FROM foo where id=?" << id >> cppdb::row >> r;
   if(r.empty()) {
      sql << "INSERT ... " ... ;
      tr.commit();
   }
   else {
      tr.rollback();
      // Handle error.
   }

Artyom

[Cppcms-users] CppDb Errors

[Daniel V. <chi...@gm...>]

Hello.

I need catch common sql errors in cppdb, such as duplicate key insert
error or other constraint errors.

How to accomplish this in a portable way with cppdb?

Thanks.

Re: [Cppcms-users] C++ locale and boost.locale

[augustin <aug...@ov...>]

On Thursday 05 May 2011 12:52:44 am Artyom wrote:
> >         getline(ss, r);
> >          return r;
>
> It is better to call
> 
>    return ss.str()

Oops! This one is very embarrassing given the time I wasted trying to figure 
out a simple way to extract the string from the stream!

Thank you very much for all your comments on my code. They're duly noted. I 
appreciate it.

>  http://cppcms.sourceforge.net/cppcms_ref_v0_99/namespacecppcms_1_1filters.
> html
> 
> So just write in the template
> 
> 
>    <% tt | datetime %>
> 
> Or even with more fine grained control
> 
>    <% tt | strftime("%Y-%m-%d %H:%M:%S %Z") %>
> 
> 
> where tt is just time_t in your context :-)

Ah, yes! You just answered my earlier question about locales in a cppcms 
context. This is actually what I meant. I knew my approach didn't feel right.

It works as it should, now.

Thanks again,

augustin.



-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.

Re: [Cppcms-users] C++ locale and boost.locale

[Artyom <art...@ya...>]

> 
> BTW, thanks to your help, I got the  missing piece, so that I finally managed 
>to 
>
> do what I wanted to  do:
> 
> #include <sstream>
> #include  <boost/locale.hpp>
> 
> std::string timestamp_to_string(time_t  const& tt) {
>         using namespace  boost::locale;
>         generator gen;
>          std::locale::global(gen("en_US.UTF-8"));
>          std::stringstream ss;
>          ss.imbue(std::locale());
>         ss << as::date  << tt << " ";
>         ss << as::time  << tt << std::endl;
>         std::string r =  "";
>         getline(ss, r);
>          return r;
> }
> 

Few points:

> timestamp_to_string(time_t  const& tt) 

It is better to write

   timestamp_to_string(time_t tt) 

As time_t is integral type and it is better
to pass it by value.

>         using namespace  boost::locale;
>         generator gen;
>          std::locale::global(gen("en_US.UTF-8"));
>          ss.imbue(std::locale());


Is very inefficient as locale generation is very
heavy procedure.

In CppCMS context you have this member function:

http://cppcms.sourceforge.net/cppcms_ref_v0_99/classcppcms_1_1http_1_1context.html#39fde738210daf84deee691f45260e5b


So basically inside cppcms::application based class just write

  std::locale loc = context().locale();

or

  std::locale loc = context().locale("en_US.UTF-8");

It returns the locale defined in configuration file for you
and it would cache the locale automatically for future use.

>          std::locale::global(gen("en_US.UTF-8"));
>          ss.imbue(std::locale());


Even if you use it, it is better to write

    ss.imbue(gen("en_US.UTF-8"));

And not set global locale and then create a locale
instance from global one.

>         ss << as::date  << tt << " ";
>         ss << as::time  << tt << std::endl;

It is better to write

   ss << as::datetime << tt;

>         std::string r =  "";

r is empty by default, no need r="";

>         getline(ss, r);
>          return r;

It is better to call

   return ss.str()

Of course without "<<std::endl"




> It took me 10-20 hours to research, read and  understand enough to come up with 
>
> the seemingly simple piece of code above.  :-/ 
> 
> I use it to format unix timestamps before pushing it to my 
> cppcms::base_content object.
> 
> I don't know if it's the right approach,  but it works...
> 

It is much simpler, you have filers namespace

  http://cppcms.sourceforge.net/cppcms_ref_v0_99/namespacecppcms_1_1filters.html

So just write in the template


   <% tt | datetime %> 

Or even with more fine grained control

   <% tt | strftime("%Y-%m-%d %H:%M:%S %Z") %> 


where tt is just time_t in your context :-)

> 
> 
> Augustin.


Best,
  Artyom

Re: [Cppcms-users] C++ locale and boost.locale

[augustin <aug...@ov...>]

On Wednesday 04 May 2011 09:29:39 pm Artyom wrote:
>  http://www.boost.org/doc/libs/1_46_1/more/getting_started/unix-
variants.html
> http://www.boost.org/doc/libs/1_46_1/more/getting_started/windows.html

Ha! Even by reading the official boost documentation page about linking, I 
wouldn't have guessed what linker option to chose. 

For Linux/g++, at least, it wouldn't hurt mentioning the -lboost_locale in the 
docs.

> Was too damn busy, I'll try to get to it this week.
> Drop me a mail if I forget once again.

There's no actual hurry as far as I am concerned, but something just came up 
today and I may have to travel to Europe for two weeks later this month, so I 
won't be available then, if I do go.



BTW, thanks to your help, I got the missing piece, so that I finally managed to 
do what I wanted to do:

#include <sstream>
#include <boost/locale.hpp>

std::string timestamp_to_string(time_t const& tt) {
        using namespace boost::locale;
        generator gen;
        std::locale::global(gen("en_US.UTF-8"));
        std::stringstream ss;
        ss.imbue(std::locale());
        ss << as::date << tt << " ";
        ss << as::time << tt << std::endl;
        std::string r = "";
        getline(ss, r);
        return r;
}

It took me 10-20 hours to research, read and understand enough to come up with 
the seemingly simple piece of code above. :-/ 

I use it to format unix timestamps before pushing it to my 
cppcms::base_content object.

I don't know if it's the right approach, but it works...



Augustin.

-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.

Re: [Cppcms-users] C++ locale and boost.locale

[Artyom <art...@ya...>]

> > > Is  this  platform/compiler dependent? Can't this kind of thing be 
> > >  documented? Anyway, in this case, I found it was -lboost_locale.
> > 
> > This is something Boost general not related
> > to Boost.Locale  specifically.
> 
> Yes, but it affects every single boost library, including  yours. Maybe you can 
>
> add such necessary information to the boost.locale  documentation.
> 

It is not something trivial especially when you use MSVC with its
endless incompatible build variants.

See:

  http://www.boost.org/doc/libs/1_46_1/more/getting_started/unix-variants.html
  http://www.boost.org/doc/libs/1_46_1/more/getting_started/windows.html


> ever considered using git?

Subversion has a huge advantage of working
over https behind firewalls and in various
even untrusted networks. So if you work
from University network or just some public
network there is no replacement for https
support.

GIT does not allow to perform commits over
https protocol like SVN.

> (I am still waiting for your review of my  form.h patch, 
> though...).

I'm sorry, you are right.

Was too damn busy, I'll try to get to it this week.
Drop me a mail if I forget once again.

Regards, 
  Artyom

Re: [Cppcms-users] C++ locale and boost.locale

[augustin <aug...@ov...>]

On Wednesday 04 May 2011 08:37:33 pm augustin wrote:
> In the boost.locale documentation, as part of the introduction, you ought
>  to make a list of assumed knowledge, with some links like those I provided
>  above, to redirect people like me to background information.

Oops. You actually already do:
http://cppcms.sourceforge.net/boost_locale/html/std_locales.html 
I read it too quickly the first time and didn't understand what I was reading.
It's slowly sinking in...

Sorry for the noise.

Augustin.




-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.

Re: [Cppcms-users] C++ locale and boost.locale

[augustin <aug...@ov...>]

On Wednesday 04 May 2011 03:35:35 pm Artyom wrote:
> See: http://thread.gmane.org/gmane.comp.lib.boost.devel/218008
> [...]
> See section about locale generation.

Thanks. The discussion thread you link to, and your mention prompted me to do 
some further research and I finally figured out that I was missing yet another 
whole chapter of this very complex, yet fascinating language (C++).
I hadn't grasped how the concept of locale was deeply embedded within the 
language itself:
http://www.cplusplus.com/reference/std/locale/ 
http://stdcxx.apache.org/doc/stdlibug/VII.html 
chapter 22 of c++0x,
chapter 27 of c++03.

With only 400+ hours of experience with C++, I still count as a very beginner. 
It's been 400 hours worth of shocks at discovering how many facets there are 
to the language, and how many ancillary skills (compiling, linking, gnu tools, 
Makefile, cmake, etc.)  and libraries (std, STL, std::locale, boost, cppcms, 
cppdb and others I am using...) I have yet to master to become, hopefully one 
day, a decent C++ programmer. 

In the boost.locale documentation, as part of the introduction, you ought to 
make a list of assumed knowledge, with some links like those I provided above, 
to redirect people like me to background information.

I'm going to be reading about all this now...

> http://stackoverflow.com/questions/3467153/is-there-an-intelligent-way-to-k
> now-
> 
> > the-name-of-the-library-to-link-to-at-compile
> > Is  this platform/compiler dependent? Can't this kind of thing be 
> > documented? Anyway, in this case, I found it was -lboost_locale.
> 
> This is something Boost general not related
> to Boost.Locale specifically.

Yes, but it affects every single boost library, including yours. Maybe you can 
add such necessary information to the boost.locale documentation.


> Take a look on libs/locale/examples
> with full code to see what you miss.

Oh, yes, I did miss that! Very cool, thanks. 

Thanks again for your constant support. Your answers were very helpful in 
pointing me in the right direction. I have much study to do, now... ;)

Btw, since I have a svn checkout (ever considered using git? Yet another thing 
I need to learn: how to link a local git working directory to a svn upstream 
source. I know there are tutorials about it, that I can find easily... there is 
so much to do and to learn!)... I was saying: since I checked out boost.locale 
from svn, let me know if you'd like me to prepare a patch for the 
documentation.... (I am still waiting for your review of my form.h patch, 
though...).


Blessings,

Augustin.







-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.

Re: [Cppcms-users] boost.locale and date formatting

[Artyom <art...@ya...>]

> From: augustin <aug...@ov...>
> 
> (second try to send this to the list)
> 
> Hello,
> 
> Speaking of  boost(er).locale, I am having problems formatting dates.
> 
> I tried the  simple example from  here:
> [...]


See: http://thread.gmane.org/gmane.comp.lib.boost.devel/218008

> I'm sure I am missing something simple. 
> Probably, I need to set the locale first, but it's not clear to me how to do 
> it.

See section about locale generation.

> 
> Also, a grip I have with most/all boost libraries, is that one  never knows 
> what option to pass to the  linker:
> 
http://stackoverflow.com/questions/3467153/is-there-an-intelligent-way-to-know-
> the-name-of-the-library-to-link-to-at-compile
> Is  this platform/compiler dependent? Can't this kind of thing be  documented?
> Anyway, in this case, I found it was -lboost_locale.
> 

This is something Boost general not related
to Boost.Locale specifically.

> For  boost.locale, more, fully working, compilable examples would  help.
> 

Take a look on libs/locale/examples
with full code to see what you miss.

> -
> 
> Most boost libraries documentation are written for very  experienced C++ 
> developers. From a newbie perspective, I find most  documentation very obtuse. 
>I 
>
> checked boost.date_time for a long while, and  still couldn't understand 
>enough  
>
> to actually make use of it. I still  have much to learn before I can understand 
>
> a lot of the boost  documentation.
> 
> -


This specific "problem" would be added to FAQ.

> 
> Generally speaking, within the context of a  cppcms application, what is the 
> best/recommended way to format dates, times  and such?
> Basically, I have a unix timestamp, and I want to output it in  human readable 

> form, both in the UTC and locale  timezone.
> 

Read the tutorial there are an examples how to make UTC, local timezone 
(default)
or custom time zone output (there is a manipulator for this purpose and 
formatting
flags).

Also the way you want to display dates is application specific and dependent
on use case.

Best,
  Artyom

[Cppcms-users] boost.locale and date formatting

[augustin <aug...@ov...>]

(second try to send this to the list)

Hello,

Speaking of boost(er).locale, I am having problems formatting dates.

I tried the simple example from here:
http://cppcms.sourceforge.net/boost_locale/html/formatting_and_parsing.html

#include <iostream>
#include <sstream>
#include <boost/locale.hpp>

using std::cout;
using std::endl;
using namespace boost::locale;

int main() {
        double now=time(0);
        cout << "Today is "<< as::date << now << " and tommorrow is " << 
now+24*3600 << endl;
        cout << "Current time is "<< as::time << now << endl;
        cout << "The current weekday is "<< as::ftime("%A") << now << endl;
}

g++ main.cpp -o run.main -lboost_locale

and the output is:

Today is 1.30439e+09 and tommorrow is 1.30448e+09
Current time is 1.30439e+09
The current weekday is 1.30439e+09

I tried with booster.locale that comes with cppcms, and I downloaded from svn 
and compiled the very latest boost.locale. I get the same result with both.

I'm sure I am missing something simple. 
Probably, I need to set the locale first, but it's not clear to me how to do 
it.

Also, a grip I have with most/all boost libraries, is that one never knows 
what option to pass to the linker:
http://stackoverflow.com/questions/3467153/is-there-an-intelligent-way-to-know-
the-name-of-the-library-to-link-to-at-compile
Is this platform/compiler dependent? Can't this kind of thing be documented?
Anyway, in this case, I found it was -lboost_locale.

For boost.locale, more, fully working, compilable examples would help.

-

Most boost libraries documentation are written for very experienced C++ 
developers. From a newbie perspective, I find most documentation very obtuse. I 
checked boost.date_time for a long while, and still couldn't understand enough  
to actually make use of it. I still have much to learn before I can understand 
a lot of the boost documentation.

-

Generally speaking, within the context of a cppcms application, what is the 
best/recommended way to format dates, times and such?
Basically, I have a unix timestamp, and I want to output it in human readable 
form, both in the UTC and locale timezone.


Thanks,

Augustin.




-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.

-------------------------------------------------------
-- 
Friends: http://www.reuniting.info/
My projects:
http://astralcity.org/ http://3enjeux.overshoot.tv/ http://linux.overshoot.tv/ 
http://overshoot.tv/ http://charityware.info/ http://masquilier.org/
http://openteacher.info/ http://minguo.info/ 
http://www.wechange.org/ http://searching911.info/












.