- Boost-users - lists.boost.org

Fwd: SQL library with full embedded SQL-syntax
by Lyfar Dmitriy 17 Mar '07

17 Mar '07

Hello, sorry for my long responce. Rene Rivera wrote: > Sure... But we could use that same argument to propose that we write C++ > programs by writing an abstraction on top of C++ compilers in Python so > that we can remove the incompatibilities in the various C++ > implementations. Having had the displeasure of dealing with PL/SQL there > is one serious drawback to the language embedded approach, it ties ones > to the capabilities of the abstraction. May be you right, but I think that that comparison is incorrect, because of as far as I know SQL-standart doesn't include many useful features from various databases, opposite, the C++ standart, which is describes the whole language :) The main goal of the abstract layer for that library is, as Thorsten said, is to provide "portable" SQL for at least a simple queries. There is no goal to emulate features of each db for others at the first time. > Jorge Lodos wrote: > Sorry, I don't see how any of this applies -- just because the SQL is a string > doesn't mean it comes from an untrusted source. And, programmers that don't > validate input from untrusted sources deserve what they get.... I'd like to draw an intermediate conclusions: 1) The SQL library should provide a backdoor so you can still use string queries if you need to be non-portable or cutting egde etc. 2) Embedded SQL commands is type-safe, but strings are flexible and more useful in the 'real world' 3) It's hard to write one big abstract layer for all databases with single functionality and many people doesn't need in it usually. 4) ...? Phil Endecott wrote: > Well, an interface is not much use without some sort of implementation > behind it. Possible implementations for a "generic relational > interface" include an SQL database, on-disk data structures implemented > using something like stxxl, and in-memory data structures made > persistent using serialisation. A measure of a good generic relational > interface could be that it works well with all of these implementations. > > Thinking just about SQL backends, the basic question is whether we want to > > (a) hide the SQL, making the database accesses look more like C++ > function calls and data structures, or > > (b) make the SQL explicit, making the C++ code look more like SQL. > > I definitely vote for (a), and the possibility of moving an application > to something like stxxl is one of my motivations for that. If you interested in such libraries -- take a look at database template library(dtl) also. What do you think about containers overhead? Does the (a) variant means that the parsing of the result of query is something like parsing an Variant type? Is it always opportune? The people who wants to have type-safety of queries and platform independence will select the (b) variant, I hope :) ----- Best Regards, Dmitriy Lyfar

1 0

Bug in boost.filesystem?
by Tom Bachmann 17 Mar '07

17 Mar '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, Is the following code expected to be portable? namespace fs = boost::filesystem; fs::rename (current_p, skin_p / current_p.leaf ()); Where current_p and skin_p are fs::path, where current_p is valid and skin_p is the result of dereferencing a directory iterator (that was not at behind-end position at time of dereferencing). For current_p = foo/10NfRaHD1AAICPMKZIC50FvmaHw==.medium.jpg skin_p = foo/bar (directory) the code causes an exception: boost::filesystem::path: invalid name "10NfRaHD1AAICPMKZIC50FvmaHw==.medium.jpg" in path: "10NfRaHD1AAICPMKZIC50FvmaHw==.medium.jpg" (that is, I believe, at the point where a temporary path object is constructed from current_p.leaf ()). This appears quiet strange to me, as I never mention the file in the code, and I'd expect the library to treat such oddities. The following code works: fs::rename (current_p, skin_p / fs::path (current_p.leaf (), fs::native)); Is this the expected behavior? - -- - -ness- -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFF+/vPvD/ijq9JWhsRAhMdAKCDyLsi0EKzhUnVcM3GjKopqnpcCQCbBK4y vkNablkmw69y3xj6ulSNdvs= =izVN -----END PGP SIGNATURE-----

2 2

Boost path to use for VC++ 8?
by Jeff Dunlap 17 Mar '07

17 Mar '07

Hello, I am new to VC++ (ver 8) and would really like to use Boost! This is what I have done: 1) I unzipped the boost download to c:\boost_1_33_1 2) At the command prompt, I executed: c:\boost_1_33_1\>bjam "-sTOOLS=VC-8_0" install bjam ran fine and a new directory was created: c:\boost\include\boost-1_33_1\boost c:\boost\lib I had assumed that using bjam's install parameter would modify whatever the path settings for the VC IDE are supposed to be, but that does not appear to happen. There are alot of paths and I want to make sure that I use the right one. What are the paths that I am supposed to include in VC++ when using the bjam param 'install' above? Secondly, I was thinking that maybe the 'stage' param might be a better bjam choice since the boost libraries are built in a common directory, thus making it easier to use. Are there drawbacks to using boost this way? Thank you very much

2 2

RWmutex + sockets
by Bo Lorentsen 17 Mar '07

17 Mar '07

Hi ... I have been looking into your thread lib, and all seem fine and nice, but is it true that there where no RW mutex or is it just called something I did not know ? Also, ... are there some kind of URL / socket lib in boost ? /BL

3 3

cache basics
by Gottlob Frege 17 Mar '07

17 Mar '07

I had an idea for a general cache, based on the idea that caches are essentially trying to avoid re-calling expensive functions with the same parameters as last time. So typically you build a cache key from the params, put it in a map, etc. But what if we did things slightly differently... template <typename T> struct cache { typedef boost::function< T () > retriever_function; typedef std::map<retriever_function, T> map; map m; T retrieve(retriever_function retriever) { map::iterator it = m.find(retriever); if (it != m.end()) { return it->second; } else { T t = retriever(); m[retriever] = t; return t; } } }; Basically, instead of building a key from the params, build a key from the params + the function that takes the params. Then, as a bonus, you don't need to worry about packing/unpacking the params - just let bind() do that for you! The other bonus is that you can cache things that may be created via different methonds. So what's the problem? The problem is that boost::function and boost::bind don't have operator<(), so you can't put a boost::function in a map. :-( (actually it appears boost::bind does have an operator<, but it doesn't do what I want - it actually bind's the operator< into a function!) Any advice? Any chance boost::function might gain a < operator? (One that compared the underlying function and the params, possibly handled different numbe of params, etc - although I could live with requiring all the param types to be the same. ie basically boost::function would pass most of the work off to the function object (ie bind)) Do I instead need to use tuples and pack/unpack the params myself (and write overloads for 0 to n params, handle return_type messes, etc etc etc )? - Tony

3 2

[MPL] 2 Newbie Questions
by Scott Meyers 17 Mar '07

17 Mar '07

I'm trying to use the MPL for the first time, and I have two questions. First, I want to use BOOST_MPL_ASSERT to say that at least one of two conditions is true. For the time being, one of the conditions is always false, so this is what I've tried: BOOST_MPL_ASSERT(( false || boost::is_same<CallerConstraints, CodeConstraints::Ignore_Constraints> )); It won't compile, so I tried this: BOOST_MPL_ASSERT(( mpl::or_(false, boost::is_same<CallerConstraints, CodeConstraints::Ignore_Constraints>) )); It won't compile either. I'm guessing that I have to use mpl::false_ or something in place of false, but that doesn't work either. Can somebody please tell me the proper way to say what I want to say? My second question is about the existence of an MPL algorithm akin to the STL algorithm includes. Given two MPL vectors V1 and V2, I want to know if all the types in V1 are also in V2. I know about mpl::contains, but I can't find mpl::includes. Is there one, or do I need to write it myself? I have the MPL book, so references to that are fine, in addition to anything online. Thanks, Scott

4 7

[filesystem] Puzzling behaviour on MSVC 7.1 with ofstream and wpath in binary mode
by Eoin 17 Mar '07

17 Mar '07

Hello, My development platform is Win XP with sp2. I am using a very recent (no older than a week) copy of the boost cvs version 1.34rc and the compilers are MSVC 7.1 and 8.0 both with the service packs. I've encountered some puzzling behaviour with filesystem::wpath's and filesystem::ofstream's when opened in binary mode and compiled with VC++ 7.1. When I attempt to open a binary ofstream with a wpath as its filename it will fail if the file did not already exist. This appears to be contrary to when the filename argument is a narrow filesystem::path. When the binary mode is not specified both files are created. Attached is a very small app. demonstrating the inconsistency. This only occurs with 7.1, when compiled with 8.0 both files will be created if they did not already exist. Kind regards, Eoin. #include <iostream> #include <boost/filesystem/path.hpp> #include <boost/filesystem/fstream.hpp> namespace fs = boost::filesystem; int main() { fs::wpath wide_filename = L"./Wide.txt"; fs::path narrow_filename = "./Narrow.txt"; fs::ofstream wide_fs(wide_filename, fs::fstream::binary); std::cout << ((wide_fs.is_open()) ? "Wide file open." : "Wide file not open.") << std::endl; fs::ofstream narrow_fs(narrow_filename, fs::fstream::binary); std::cout << ((narrow_fs.is_open()) ? "Narrow file open." : "Narrow file not open.") << std::endl; return 0; }

1 0

Re: [Boost-users] Boost path to use for VC++ 8?
by Nathan McDaniel 16 Mar '07

16 Mar '07

Hi Jeff, I'm not real clear either on the difference between 'stage' and 'install' (seem to be different only in the default folders they place things), but this is what I'm doing: First, I build Boost with this command (I only want the release libraries and only statically linked--builds a bit faster): bjam "-sBUILD=release <runtime-link>static" "-sTOOLS=vc-8_0" --prefix=./boost-built install When it's done, you'll have a "boost-built" subfolder with a "boost" folder and a "lib" folder. I don't plan on routinely rebuilding 1.33.1, so I copy the "boost-built" folder into one of my project folders and delete the boost build directory/files. Then add the "boost-built\boost" subfolder to your include search path, and the "boost-built\lib" subfolder to your library search path. Nathan ---------- Forwarded message ---------- > From: Jeff Dunlap <jeff_j_dunlap(a)yahoo.com> > To: boost-users(a)lists.boost.org > Date: Thu, 15 Mar 2007 02:49:10 +0000 (UTC) > Subject: [Boost-users] Boost path to use for VC++ 8? > Hello, I am new to VC++ (ver 8) and would really like to use Boost! > > This is what I have done: > > 1) I unzipped the boost download to c:\boost_1_33_1 > > 2) At the command prompt, I executed: > c:\boost_1_33_1\>bjam "-sTOOLS=VC-8_0" install > > bjam ran fine and a new directory was created: > > c:\boost\include\boost-1_33_1\boost > c:\boost\lib > > > I had assumed that using bjam's install parameter would modify whatever > the > path settings for the VC IDE are supposed to be, but that does not appear > to > happen. There are alot of paths and I want to make sure that I use the > right > one. What are the paths that I am supposed to include in VC++ when using > the > bjam param 'install' above? > > Secondly, I was thinking that maybe the 'stage' param might be a better > bjam > choice since the boost libraries are built in a common directory, thus > making > it easier to use. Are there drawbacks to using boost this way? > > Thank you very much >

2 1

iostreams filter help
by Aljaz 16 Mar '07

16 Mar '07

Hello I've been trying to get boost iostreams filter to work. Heres the code I managed to write, but theres something I cannot understand.. It works fine with one small text file and it will process the lines okay, but when I try to use another text file (which is much bigger) it wont recognize any line and will return 0 lines counted (though file has around 2 million lines).. What am I doing wrong? What could be the cause of this? Best regards Thanks for help Heres the code: #include <boost/iostreams/device/file.hpp> #include <boost/iostreams/filtering_stream.hpp> #include <boost/iostreams/filter/newline.hpp> #include <boost/iostreams/filter/line.hpp> #include <boost/ref.hpp> #include <fstream> #include <ostream> #include <iostream> #include <string> using namespace std; using namespace boost::iostreams; namespace io = boost::iostreams; class long_line_counter : public io::line_filter { public: explicit long_line_counter(int max_length = 80) : max_(max_length), count_(0) { } int count() const { return count_; } private: std::string do_filter(const std::string& line) { ++count_; return line; } int max_; int count_; }; int main() { filtering_istream fin; long_line_counter llc(80); fin.push(boost::ref(llc)); fin.push(file_source("test2.txt", ios::in | ios::binary)); char buf[4012]; while ( fin.read(buf, 4012) && (llc.count() < 7000) ) { std::cout << "read" << "\n"; } std::cout << llc.count() << "\n"; return 0; }

1 0

[Serialization] issue with portable_binary_oarchive
by Dan Leibovich 16 Mar '07

16 Mar '07

Hi, I am trying to use the portable binary archive example. I need to serialize only integers and strings. I tried it on linux and solaris5.8. The result binary stream is not the same and deserialization fails. I used hex viewer and notice that the files are very similar except for some few differences which results from endienness issues. Can it be that the metadata is written differently on each platform. This is happens when I serialize derived class through a pointer to base class. (I used BOOS_CLASS_EXPORT) Below you will find a test case. It produce differ output on linux and solaris. Am I missing something? Thanks in advance Dan Leibovich #include <fstream> #include "portable_binary_oarchive.hpp" #include <boost/serialization/base_object.hpp> #include <boost/serialization/export.hpp> class Base { private: friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & x; } protected: int x; public: Base() {} ; Base(int x_) : x(x_) {} ; virtual void foo() {} ; }; class Derived1 : public Base { private: friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & boost::serialization::base_object<Base>(*this); ar & y; } protected: int y; public: Derived1() {} ; Derived1(int x_, int y_) : Base(x_), y(y_) {} ; virtual void foo() {x++;} }; class Derived2 : public Base { private: friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & boost::serialization::base_object<Base>(*this); ar & z; } protected: bool z; public: Derived2() {} ; Derived2(int x_) : Base(x_), z(true) {} ; virtual void foo() {x--;} }; class Container { Base* b1 ; Base* b2 ; friend class boost::serialization::access; template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & b1; ar & b2 ; } public: Container() {} ; Container(int x, int y) { b1 = new Derived1(x,y) ; b2 = new Derived2(x) ; } }; BOOST_CLASS_EXPORT(Derived1) ; BOOST_CLASS_EXPORT(Derived2) ; int main() { std::ofstream ofs("filename"); const Container c(1,2) ; portable_binary_oarchive oa(ofs); oa << c; }

2 19