May 2023 - Boost - lists.preview.boost.org

Re: [boost] charconv for C++11
by Ruben Perez 12 May '23

12 May '23

> > > What does charconv do? I need this in MySQL.

1 0

Re: [boost] charconv for C++11
by Matt Borland 12 May '23

12 May '23

> I also still think that whoever came up with that specification needs a > whap upside the head, but then I've never encountered a reason to print > a floating-point value in hex, so I assume there's some justification > for it. Gavin, Hexfloats are useful for comparison of numerical accuracy. We use them in the Boost.Math unit tests when outputting ULP distances. https://github.com/boostorg/math/blob/develop/test/math_unit_test.hpp#L136 Matt

1 0

Re: [boost] charconv for C++11
by Matt Borland 12 May '23

12 May '23

> On May 12, 2023, at 8:59 AM, Alexander Grund via Boost <boost(a)lists.boost.org> wrote: > > Am 12.05.23 um 02:52 schrieb Gavin Lambert via Boost: >> Mere moments ago, quoth I: >>>> Hex format will be of the form +1.0e2p+5. The integer part will >>>> always be 0 or 1. The exponent will be with a p instead of an e like >>>> in base 10 formats. >> (Also, FWIW, gcc fails to parse that with the leading +. And I >> couldn't even find any online calculators that understand this format.) > According to cppreference the floating point parser should not support it: > > "the plus sign is not recognized outside of the exponent (only the > minus sign is permitted at the beginning)" > > So I'd say that is a bug in Boost charconv given that it wants to be > as-if the standard one. > You two are correct that a value with a leading plus sign should be EINVAL rather than ignoring it during parsing. I opened an issue, and will address it. Matt

1 0

Re: [boost] charconv for C++11
by Alexander Grund 12 May '23

12 May '23

Am 12.05.23 um 02:52 schrieb Gavin Lambert via Boost: > Mere moments ago, quoth I: >>> Hex format will be of the form +1.0e2p+5. The integer part will >>> always be 0 or 1. The exponent will be with a p instead of an e like >>> in base 10 formats. > (Also, FWIW, gcc fails to parse that with the leading +. And I > couldn't even find any online calculators that understand this format.) According to cppreference the floating point parser should not support it: "the plus sign is not recognized outside of the exponent (only the minus sign is permitted at the beginning)" So I'd say that is a bug in Boost charconv given that it wants to be as-if the standard one.

1 0

Re: [boost] Extra file in boost serialization for 1.82.0
by Robert Ramey 12 May '23

12 May '23

On 5/11/23 9:15 PM, Robert Ramey via Boost wrote: > There is a little more to it than this though. Some time ago I modified > my usage of Git so that I have my development updates in the development > branch - so far so good. But then I noticed that all the same > information incremental files were repeated in the master branch. I > found this redundant, repetitive and confusing. So I took to using > "squash and merge" when incorporating the changes on the development > branch. This option is available on GitHub. This seemed to give me what > I wanted: a detailed history of development and a concise changes in the > history of releases. It seems that something is awry here as the "squash > & merge" option doesn't merge file deletions. When I catch up, I'll > look into this. > Note this this procedure has some effects which were unanticipated. a) when someone checks in a change into the master branch without telling me, it is wiped out by the next merger to master. At first I thought this was a bug, now I'm thinking this is more of feature as it takes a long time to track down cases where someone has made changes which leave the master permanently out of sync with the development branch. b) The development and master branches are always the same at the time of release. Again - a feature not a bug. c) Of course it probably does annoy he who checked in a change directly to master, but he shouldn't be doing that anyway. Robert Ramey

1 0

Re: [boost] Extra file in boost serialization for 1.82.0
by Glen Fernandes 12 May '23

12 May '23

On Thu, May 11, 2023 at 9:20 PM Marshall Clow via Boost < boost(a)lists.boost.org> wrote: > > On May 11, 2023, at 12:01 PM, Bryan St. Amour via Boost < > boost(a)lists.boost.org> wrote: > > > > Hi, > > > > I noticed that under boost/serialization, there is a file > "collection_size_type copy.hpp" (notice the space in the file name.) > Comparing it with the contents of collection_size_type.hpp, it looks like > "collection_size_type copy.hpp" is newer. Can someone verify which one is > the correct collection_size_type header file? > > This is https://github.com/boostorg/serialization/issues/237 < > https://github.com/boostorg/serialization/issues/237> > And https://github.com/boostorg/serialization/issues/221 < > https://github.com/boostorg/serialization/issues/221> > > Sadly, this has been a problem for several releases. > > Robert removed it from develop a while ago but it's still in master. Robert, your merges to master aren't removing the file, you need to fix your branches. Glen

2 1

Re: [boost] Extra file in boost serialization for 1.82.0
by Marshall Clow 12 May '23

12 May '23

> On May 11, 2023, at 12:01 PM, Bryan St. Amour via Boost <boost(a)lists.boost.org> wrote: > > Hi, > > I noticed that under boost/serialization, there is a file "collection_size_type copy.hpp" (notice the space in the file name.) Comparing it with the contents of collection_size_type.hpp, it looks like "collection_size_type copy.hpp" is newer. Can someone verify which one is the correct collection_size_type header file? This is https://github.com/boostorg/serialization/issues/237 <https://github.com/boostorg/serialization/issues/237> And https://github.com/boostorg/serialization/issues/221 <https://github.com/boostorg/serialization/issues/221> Sadly, this has been a problem for several releases. — Marshall

1 0

Re: [boost] charconv for C++11
by Gavin Lambert 12 May '23

12 May '23

Mere moments ago, quoth I: >> Hex format will be of the form +1.0e2p+5. The integer part will >> always be 0 or 1. The exponent will be with a p instead of an e like >> in base 10 formats. > This is some strange definition of "hex" with which I was previously > unfamiliar. > > It also contradicts itself, by saying "a p instead of an e" while the > example contains both. (And isn't a very good example, lacking actual > hex digits.) It later occurred to me that the "e" could be intended as a hex digit, thus "1.0e2 with exponent +5". This still makes it a poor example, though. It might perhaps be intended as an example of a vexing parse, but in that case it shouldn't be the *only* example (and perhaps should be elsewhere). (Also, FWIW, gcc fails to parse that with the leading +. And I couldn't even find any online calculators that understand this format.)

1 0

Re: [boost] charconv for C++11
by Gavin Lambert 12 May '23

12 May '23

On 12/05/2023 10:59, Peter Dimov wrote: > The work-in-progress documentation is here: https://develop.charconv.cpp.al/ > Hex format will be of the form +1.0e2p+5. The integer part will > always be 0 or 1. The exponent will be with a p instead of an e like > in base 10 formats. This is some strange definition of "hex" with which I was previously unfamiliar. It also contradicts itself, by saying "a p instead of an e" while the example contains both. (And isn't a very good example, lacking actual hex digits.) I've now found the reference to it -- https://en.cppreference.com/w/cpp/utility/to_chars says that "hex" uses as-if printf %a, and https://en.cppreference.com/w/cpp/io/c/fprintf says printf %a appears *somewhat* like the above (but I still think there is an error in the above). I also still think that whoever came up with that specification needs a whap upside the head, but then I've never encountered a reason to print a floating-point value in hex, so I assume there's some justification for it.

1 0

Re: [boost] charconv for C++11
by Peter Dimov 11 May '23

11 May '23

Robert Ramey wrote: > > In addition to all that, from/to_chars are standard C++17, so Charconv > > can be used by libraries that want to support C++11 (or 14) today, but > > at the same time want to be able to switch to the standard functions > > later when increasing the minimal supported standard to C++17. > > So this is an implementation in C++11 of a library which is already accepted > (and presumably shipping) in C++17 ? And the only user who would use this is > someone who has C++11 but can't upgrade to C++17? > Doesn't seem like very many people to me. Users... and library authors whose libraries still support C++11/14, such as Boost.JSON. > I didn't find an documentation for this. The work-in-progress documentation is here: https://develop.charconv.cpp.al/

1 0