Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-25 Thread Eliot Moss 

It seems that git send-email is in the git-email package,
not installed by default when installing git ...

Regards - EM

--
Problem reports:   http://cygwin.com/problems.html
FAQ:   http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-25 Thread Brian Inglis 

On 2016-08-25 08:52, Corinna Vinschen wrote:

On Aug 25 08:38, Brian Inglis wrote:

On 2016-08-25 02:07, Corinna Vinschen wrote:

You don't need mailx, btw.  After you created your patch with
`git format-patch', you can send it with `git send-email' :)


Looked for that, and everything there, but:

$ uname -srvmo
CYGWIN_NT-10.0 2.5.2(0.297/5/3) 2016-06-23 14:29 x86_64 Cygwin
$ git --version
git version 2.8.3
$ man git | fgrep git-send-email
   git-send-email(1)
$ git help send-email
No manual entry for gitsend-email
$ git help git-send-email
No manual entry for git-send-email
$ man git-send-email
No manual entry for git-send-email
$ ls -1 /usr/share/man/man1/git* | wc -l
143
$ ls /usr/share/man/man1/git-send-email*
ls: cannot access '/usr/share/man/man1/git-send-email*': No such file or 
directory
$ git send-email
git: 'send-email' is not a git command. See 'git --help'.


Huh?  WJFFM:

$ uname -a
CYGWIN_NT-10.0 vmbert10 2.6.0(0.304/5/3) 2016-08-25 14:03 x86_64 Cygwin
$ cygcheck -c git
Cygwin Package Information
Package  VersionStatus
git  2.8.3-1OK
$ git --version
git version 2.8.3
$ ls -l /usr/share/man/man1/git-send-email*
-rw-r--r-- 1 corinna vinschen 6835 May 24 18:55 
/usr/share/man/man1/git-send-email.1.gz
$ man git send-email
GIT-SEND-EMAIL(1) Git Manual GIT-SEND-EMAIL(1)
NAME
   git-send-email - Send a collection of patches as emails
SYNOPSIS
   git send-email [options] ...
   git send-email --dump-aliases
$ git send-email
No patch files specified!
git send-email [options] 
git send-email --dump-aliases


Eric pointed out that git-email is a separate package,
as I now find is git-completion, and others...
Hatess them, preciouss...

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

--
Problem reports:   http://cygwin.com/problems.html
FAQ:   http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-25 Thread Eric Blake 
On 08/25/2016 09:38 AM, Brian Inglis wrote:
>> You don't need mailx, btw.  After you created your patch with
>> `git format-patch', you can send it with `git send-email' :)
> 
> Looked for that, and everything there, but:
> 
> $ uname -srvmo
> CYGWIN_NT-10.0 2.5.2(0.297/5/3) 2016-06-23 14:29 x86_64 Cygwin
> $ git --version
> git version 2.8.3
> $ man git | fgrep git-send-email
>git-send-email(1)
> $ git help send-email
> No manual entry for gitsend-email

Rerun setup.exe, and download git-email.

-- 
Eric Blake   eblake redhat com+1-919-301-3266
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-25 Thread Corinna Vinschen 
On Aug 25 08:38, Brian Inglis wrote:
> On 2016-08-25 02:07, Corinna Vinschen wrote:
> > You don't need mailx, btw.  After you created your patch with
> > `git format-patch', you can send it with `git send-email' :)
> 
> Looked for that, and everything there, but:
> 
> $ uname -srvmo
> CYGWIN_NT-10.0 2.5.2(0.297/5/3) 2016-06-23 14:29 x86_64 Cygwin
> $ git --version
> git version 2.8.3
> $ man git | fgrep git-send-email
>git-send-email(1)
> $ git help send-email
> No manual entry for gitsend-email
> $ git help git-send-email
> No manual entry for git-send-email
> $ man git-send-email
> No manual entry for git-send-email
> $ ls -1 /usr/share/man/man1/git* | wc -l
> 143
> $ ls /usr/share/man/man1/git-send-email*
> ls: cannot access '/usr/share/man/man1/git-send-email*': No such file or 
> directory
> $ git send-email
> git: 'send-email' is not a git command. See 'git --help'.

Huh?  WJFFM:

$ uname -a
CYGWIN_NT-10.0 vmbert10 2.6.0(0.304/5/3) 2016-08-25 14:03 x86_64 Cygwin
$ cygcheck -c git
Cygwin Package Information
Package  VersionStatus
git  2.8.3-1OK
$ git --version
git version 2.8.3
$ ls -l /usr/share/man/man1/git-send-email*
-rw-r--r-- 1 corinna vinschen 6835 May 24 18:55 
/usr/share/man/man1/git-send-email.1.gz
$ man git send-email 
GIT-SEND-EMAIL(1) Git Manual GIT-SEND-EMAIL(1)

NAME
   git-send-email - Send a collection of patches as emails

SYNOPSIS
   git send-email [options] ...
   git send-email --dump-aliases
[...]
$ git send-email

No patch files specified!

git send-email [options] 
git send-email --dump-aliases
[...]


Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat


signature.asc
Description: PGP signature

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-25 Thread Eliot Moss 

I agree -- my git has no "send-email" command.
The "format-patch" commands looks useful in this
context, however.  (My git is cygwin got, by the way!)

Cheers -- Eliot Moss

--
Problem reports:   http://cygwin.com/problems.html
FAQ:   http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-25 Thread Brian Inglis 

On 2016-08-25 02:07, Corinna Vinschen wrote:

On Aug 24 14:47, Brian Inglis wrote:


On 2016-08-24 12:29, Corinna Vinschen wrote:

On Aug 24 11:30, Brian Inglis wrote:

On 2016-08-24 02:22, Corinna Vinschen wrote:

On Aug 23 23:36, Brian Inglis wrote:

On 2016-08-23 22:15, Brian Inglis wrote:

On 2016-08-23 10:11, Corinna Vinschen wrote:

On Aug 23 07:27, Brian Inglis wrote:

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:
* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?
* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);
* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.

I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.

GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.

i.e. str[n]casecmp_l should be defined under #if __POSIX_VISIBLE >= 200809
but not defined under #ifdef __GNU_VISIBLE in string*s*.h,
and defined under #ifdef __GNU_VISIBLE but not defined under
#if __POSIX_VISIBLE >= 200809 in *string*.h;
strerror_l should be under #if __POSIX_VISIBLE  >= 200809 in *string*.h,
or its #includes.

They were already declared in string,h.

Sorry for the poor explanation, but what I was failing to say clearly
was that:
* there does not appear to be any strerror_l declaration in string.h
[...]
and that str[n]casecmp_l conditionals __GNU_VISIBLE and __POSIX_VISIBLE >= 
200809
appear to be flipped around between string.h and strings.h declarations in:
[...]

Care to send patches to the newlib list?  Patches (git format-patch)
rule over descriptions alone :}


Knew there were good reasons I avoided git for a decade!
See https://en.wikipedia.org/wiki/Git_(slang)
Developers never heard of KISS, unlike you folks at Cygwin ;^> and the folks at 
hg.

git show attached in case my patch email does not get thru or is scrambled some 
way,
so it does not apply with git am, after my screwing around with git and mailx.


It applied, but I had to make a few changes, see my reply to the
newlib list.  What you shouldn't do is to put the entire log
message into one line.  The first line of your log message becomes
the subject of your patch email.  Keep it short and add the
rest of the log message starting in line 3.  Line 2 stays clear,
it separates the subject from the body.


Seems like mailx -t does not handle multi-line subjects and .sig
properly: may have some digging to do there. Other mail clients
do not support plain text attachments: I've used most in Cygwin,
and built mailx from s-nail based on Heirloom, as Heirloom mailx
just failed to email  from Cygwin.
 

You don't need mailx, btw.  After you created your patch with
`git format-patch', you can send it with `git send-email' :)


Looked for that, and everything there, but:

$ uname -srvmo
CYGWIN_NT-10.0 2.5.2(0.297/5/3) 2016-06-23 14:29 x86_64 Cygwin
$ git --version
git version 2.8.3
$ man git | fgrep git-send-email
   git-send-email(1)
$ git help send-email
No manual entry for gitsend-email
$ git help git-send-email
No manual entry for git-send-email
$ man git-send-email
No manual entry for git-send-email
$ ls -1 /usr/share/man/man1/git* | wc -l
143
$ ls /usr/share/man/man1/git-send-email*
ls: cannot access '/usr/share/man/man1/git-send-email*': No such file or 
directory
$ git send-email
git: 'send-email' is not a git command. See 'git --help'.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

--
Problem reports:   http://cygwin.com/problems.html
FAQ:   http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-25 Thread Corinna Vinschen 
On Aug 24 14:47, Brian Inglis wrote:
> 
> On 2016-08-24 12:29, Corinna Vinschen wrote:
> > On Aug 24 11:30, Brian Inglis wrote:
> > > On 2016-08-24 02:22, Corinna Vinschen wrote:
> > > > On Aug 23 23:36, Brian Inglis wrote:
> > > > > On 2016-08-23 22:15, Brian Inglis wrote:
> > > > > > On 2016-08-23 10:11, Corinna Vinschen wrote:
> > > > > > > On Aug 23 07:27, Brian Inglis wrote:
> > > > > > > > Compared lists of locale_t headers and functions for POSIX, 
> > > > > > > > Cygwin,
> > > > > > > > and glibc, attached below for comparison, and found:
> > > > > > > > * missed string.h(strerror_l) on my first check;
> > > > > > > >not sure if you can implement that easily on Windows?
> > > > > > > > * GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);
> > > > > > > > * GNU also defines string.h(str[n]casecmp_l) functions as an 
> > > > > > > > extension,
> > > > > > > >as well as in POSIX specified strings.h.
> > > > > > > I just applied a couple of patches to add the missing strerror_l,
> > > > > > > strptime_l and wcsftime_l.  I also added the missing 
> > > > > > > str[n]casecmp_l
> > > > > > > prototypes to strings.h.  I'll create a new test release in a bit.
> > > > > > GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h 
> > > > > > also.
> > > > > i.e. str[n]casecmp_l should be defined under #if __POSIX_VISIBLE >= 
> > > > > 200809
> > > > > but not defined under #ifdef __GNU_VISIBLE in string*s*.h,
> > > > > and defined under #ifdef __GNU_VISIBLE but not defined under
> > > > > #if __POSIX_VISIBLE >= 200809 in *string*.h;
> > > > > strerror_l should be under #if __POSIX_VISIBLE  >= 200809 in 
> > > > > *string*.h,
> > > > > or its #includes.
> > > > They were already declared in string,h.
> > > Sorry for the poor explanation, but what I was failing to say clearly
> > > was that:
> > > * there does not appear to be any strerror_l declaration in string.h
> > > [...]
> > > and that str[n]casecmp_l conditionals __GNU_VISIBLE and __POSIX_VISIBLE 
> > > >= 200809
> > > appear to be flipped around between string.h and strings.h declarations 
> > > in:
> > > [...]
> > Care to send patches to the newlib list?  Patches (git format-patch)
> > rule over descriptions alone :}
> 
> Knew there were good reasons I avoided git for a decade!
> See https://en.wikipedia.org/wiki/Git_(slang)
> Developers never heard of KISS, unlike you folks at Cygwin ;^> and the folks 
> at hg.
> 
> git show attached in case my patch email does not get thru or is scrambled 
> some way,
> so it does not apply with git am, after my screwing around with git and mailx.

It applied, but I had to make a few changes, see my reply to the
newlib list.  What you shouldn't do is to put the entire log
message into one line.  The first line of your log message becomes
the subject of your patch email.  Keep it short and add the
rest of the log message starting in line 3.  Line 2 stays clear,
it separates the subject from the body.

You don't need mailx, btw.  After you created your patch with
`git format-patch', you can send it with `git send-email' :)


Thanks,
Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat


signature.asc
Description: PGP signature

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-24 Thread Brian Inglis 

On 2016-08-24 12:29, Corinna Vinschen wrote:

On Aug 24 11:30, Brian Inglis wrote:

On 2016-08-24 02:22, Corinna Vinschen wrote:

On Aug 23 23:36, Brian Inglis wrote:

On 2016-08-23 22:15, Brian Inglis wrote:

On 2016-08-23 10:11, Corinna Vinschen wrote:

On Aug 23 07:27, Brian Inglis wrote:

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:
* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?
* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);
* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.

I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.

GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.

i.e. str[n]casecmp_l should be defined under #if __POSIX_VISIBLE >= 200809
but not defined under #ifdef __GNU_VISIBLE in string*s*.h,
and defined under #ifdef __GNU_VISIBLE but not defined under
#if __POSIX_VISIBLE >= 200809 in *string*.h;
strerror_l should be under #if __POSIX_VISIBLE  >= 200809 in *string*.h,
or its #includes.

They were already declared in string,h.

Sorry for the poor explanation, but what I was failing to say clearly
was that:
* there does not appear to be any strerror_l declaration in string.h
[...]
and that str[n]casecmp_l conditionals __GNU_VISIBLE and __POSIX_VISIBLE >= 
200809
appear to be flipped around between string.h and strings.h declarations in:
[...]

Care to send patches to the newlib list?  Patches (git format-patch)
rule over descriptions alone :}


Knew there were good reasons I avoided git for a decade!
See https://en.wikipedia.org/wiki/Git_(slang)
Developers never heard of KISS, unlike you folks at Cygwin ;^> and the folks at 
hg.

git show attached in case my patch email does not get thru or is scrambled some 
way,
so it does not apply with git am, after my screwing around with git and mailx.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada
commit 3be4400034a2bb9c28300b088ecb6271eaa44a2b
Author: Brian Inglis 
Date:   Wed Aug 24 13:46:25 2016 -0600

add string.h(strerror_l) as POSIX, change string.h(str[n]casecmp_l) from 
POSIX
to GNU, change strings.h(str[n]casecmp_l) from GNU to POSIX, use _EXFUN

diff --git a/newlib/libc/include/string.h b/newlib/libc/include/string.h
index 8f4fd2e..96703b5 100644
--- a/newlib/libc/include/string.h
+++ b/newlib/libc/include/string.h
@@ -48,13 +48,14 @@ char*_EXFUN(strtok,(char *__restrict, const char 
*__restrict));
 size_t  _EXFUN(strxfrm,(char *__restrict, const char *__restrict, size_t));
 
 #if __POSIX_VISIBLE >= 200809
-extern int strcasecmp_l (const char *, const char *, locale_t);
-extern int strncasecmp_l (const char *, const char *, size_t, locale_t);
-extern int strcoll_l (const char *, const char *, locale_t);
-extern size_t strxfrm_l (char *__restrict, const char *__restrict, size_t,
-locale_t);
+int _EXFUN(strcoll_l,(const char *, const char *, locale_t);
+char   *_EXFUN(strerror_l,(int, locale_t);
+size_t  _EXFUN(strxfrm_l,(char *__restrict, const char *__restrict, size_t, 
locale_t));
+#endif
+#if __GNU_VISIBLE
+int _EXFUN(strcasecmp_l,(const char *, const char *, locale_t));
+int _EXFUN(strncasecmp_l,(const char *, const char *, size_t, locale_t));
 #endif
-
 #if __MISC_VISIBLE || __POSIX_VISIBLE
 char   *_EXFUN(strtok_r,(char *__restrict, const char *__restrict, char 
**__restrict));
 #endif
diff --git a/newlib/libc/include/strings.h b/newlib/libc/include/strings.h
index b462d07..60e995b 100644
--- a/newlib/libc/include/strings.h
+++ b/newlib/libc/include/strings.h
@@ -33,10 +33,10 @@ int  _EXFUN(ffs,(int));
 int _EXFUN(strcasecmp,(const char *, const char *));
 int _EXFUN(strncasecmp,(const char *, const char *, size_t));
 
-#if __GNU_VISIBLE
-extern int strcasecmp_l (const char *, const char *, locale_t);
-extern int strncasecmp_l (const char *, const char *, size_t, locale_t);
-#endif /* _GNU_VISIBLE */
+#if __POSIX_VISIBLE >= 200809
+int _EXFUN(strcasecmp_l,(const char *, const char *, locale_t);
+int _EXFUN(strncasecmp_l,(const char *, const char *, size_t, locale_t);
+#endif /* __POSIX_VISIBLE >= 200809 */
 
 _END_STD_C
 
--
Problem reports:   http://cygwin.com/problems.html
FAQ:   http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-24 Thread Corinna Vinschen 
On Aug 24 11:30, Brian Inglis wrote:
> On 2016-08-24 02:22, Corinna Vinschen wrote:
> > On Aug 23 23:36, Brian Inglis wrote:
> > > On 2016-08-23 22:15, Brian Inglis wrote:
> > > > On 2016-08-23 10:11, Corinna Vinschen wrote:
> > > > > On Aug 23 07:27, Brian Inglis wrote:
> > > > > > Compared lists of locale_t headers and functions for POSIX, Cygwin,
> > > > > > and glibc, attached below for comparison, and found:
> > > > > > * missed string.h(strerror_l) on my first check;
> > > > > >not sure if you can implement that easily on Windows?
> > > > > > * GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);
> > > > > > * GNU also defines string.h(str[n]casecmp_l) functions as an 
> > > > > > extension,
> > > > > >as well as in POSIX specified strings.h.
> > > > > I just applied a couple of patches to add the missing strerror_l,
> > > > > strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
> > > > > prototypes to strings.h.  I'll create a new test release in a bit.
> > > > GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.
> > > i.e. str[n]casecmp_l should be defined under #if __POSIX_VISIBLE >= 200809
> > > but not defined under #ifdef __GNU_VISIBLE in string*s*.h,
> > > and defined under #ifdef __GNU_VISIBLE but not defined under
> > > #if __POSIX_VISIBLE >= 200809 in *string*.h;
> > > strerror_l should be under #if __POSIX_VISIBLE  >= 200809 in *string*.h,
> > > or its #includes.
> > They were already declared in string,h.
> 
> Sorry for the poor explanation, but what I was failing to say clearly
> was that:
> 
> * there does not appear to be any strerror_l declaration in string.h
> [...]
> and that str[n]casecmp_l conditionals __GNU_VISIBLE and __POSIX_VISIBLE >= 
> 200809
> appear to be flipped around between string.h and strings.h declarations in:
> [...]

Care to send patches to the newlib list?  Patches (git format-patch)
rule over descriptions alone :}


Thanks,
Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat


signature.asc
Description: PGP signature

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-24 Thread Brian Inglis 

On 2016-08-24 02:22, Corinna Vinschen wrote:

On Aug 23 23:36, Brian Inglis wrote:

On 2016-08-23 22:15, Brian Inglis wrote:

On 2016-08-23 10:11, Corinna Vinschen wrote:

On Aug 23 07:27, Brian Inglis wrote:

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:
* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?
* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);
* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.

I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.

GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.

i.e. str[n]casecmp_l should be defined under #if __POSIX_VISIBLE >= 200809
but not defined under #ifdef __GNU_VISIBLE in string*s*.h,
and defined under #ifdef __GNU_VISIBLE but not defined under
#if __POSIX_VISIBLE >= 200809 in *string*.h;
strerror_l should be under #if __POSIX_VISIBLE  >= 200809 in *string*.h,
or its #includes.

They were already declared in string,h.


Sorry for the poor explanation, but what I was failing to say clearly
was that:

* there does not appear to be any strerror_l declaration in string.h
or any header searching using grep or pickaxe on:

https://sourceware.org/git/?p=newlib-cygwin.git=search=HEAD=grep=strerror_l

and that str[n]casecmp_l conditionals __GNU_VISIBLE and __POSIX_VISIBLE >= 
200809
appear to be flipped around between string.h and strings.h declarations in:

https://sourceware.org/git/?p=newlib-cygwin.git;a=blob;f=newlib/libc/include/string.h

#if __POSIX_VISIBLE >= 200809
extern int strcasecmp_l (const char *, const char *, locale_t);
extern int strncasecmp_l (const char *, const char *, size_t, locale_t);
extern int strcoll_l (const char *, const char *, locale_t);
extern size_t strxfrm_l (char *__restrict, const char *__restrict, size_t,
 locale_t);
#endif

should probably be:

#if __POSIX_VISIBLE >= 200809
extern int strcoll_l (const char *, const char *, locale_t);
extern size_t strxfrm_l (char *__restrict, const char *__restrict, size_t,
 locale_t);
#endif
#if __GNU_VISIBLE
extern int strcasecmp_l (const char *, const char *, locale_t);
extern int strncasecmp_l (const char *, const char *, size_t, locale_t);
#endif

and in:

https://sourceware.org/git/?p=newlib-cygwin.git;a=blob;f=newlib/libc/include/strings.h

#if __GNU_VISIBLE
extern int strcasecmp_l (const char *, const char *, locale_t);
extern int strncasecmp_l (const char *, const char *, size_t, locale_t);
#endif /* _GNU_VISIBLE */

should probably be:

#if __POSIX_VISIBLE >= 200809
extern int strcasecmp_l (const char *, const char *, locale_t);
extern int strncasecmp_l (const char *, const char *, size_t, locale_t);
#endif /* __POSIX_VISIBLE >= 200809 */

if I am reading the code and docs correctly.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

--
Problem reports:   http://cygwin.com/problems.html
FAQ:   http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-24 Thread Corinna Vinschen 
On Aug 23 23:36, Brian Inglis wrote:
> On 2016-08-23 22:15, Brian Inglis wrote:
> > On 2016-08-23 10:11, Corinna Vinschen wrote:
> > > On Aug 23 07:27, Brian Inglis wrote:
> > > > Compared lists of locale_t headers and functions for POSIX, Cygwin,
> > > > and glibc, attached below for comparison, and found:
> > > > 
> > > > * missed string.h(strerror_l) on my first check;
> > > >not sure if you can implement that easily on Windows?
> > > > 
> > > > * GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);
> > > > 
> > > > * GNU also defines string.h(str[n]casecmp_l) functions as an extension,
> > > >as well as in POSIX specified strings.h.
> > > 
> > > I just applied a couple of patches to add the missing strerror_l,
> > > strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
> > > prototypes to strings.h.  I'll create a new test release in a bit.
> > 
> > GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.
> 
> i.e. str[n]casecmp_l should be defined under #if __POSIX_VISIBLE >= 200809
> but not defined under #ifdef __GNU_VISIBLE in string*s*.h,
> and defined under #ifdef __GNU_VISIBLE but not defined under
> #if __POSIX_VISIBLE >= 200809 in *string*.h;
> strerror_l should be under #if __POSIX_VISIBLE  >= 200809 in *string*.h,
> or its #includes.

They were already declared in string,h.


Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat


signature.asc
Description: PGP signature

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-23 Thread Brian Inglis 

On 2016-08-23 22:15, Brian Inglis wrote:

On 2016-08-23 10:11, Corinna Vinschen wrote:

On Aug 23 07:27, Brian Inglis wrote:

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:

* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?

* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);

* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.


I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.


GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.


i.e. str[n]casecmp_l should be defined under #if __POSIX_VISIBLE >= 200809
but not defined under #ifdef __GNU_VISIBLE in string*s*.h,
and defined under #ifdef __GNU_VISIBLE but not defined under
#if __POSIX_VISIBLE >= 200809 in *string*.h;
strerror_l should be under #if __POSIX_VISIBLE  >= 200809 in *string*.h,
or its #includes.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

--
Problem reports:   http://cygwin.com/problems.html
FAQ:   http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-23 Thread Brian Inglis 

On 2016-08-23 10:11, Corinna Vinschen wrote:

On Aug 23 07:27, Brian Inglis wrote:

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:

* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?

* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);

* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.


I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.


GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

--
Problem reports:   http://cygwin.com/problems.html
FAQ:   http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-23 Thread Corinna Vinschen 
On Aug 23 07:27, Brian Inglis wrote:
> Compared lists of locale_t headers and functions for POSIX, Cygwin,
> and glibc, attached below for comparison, and found:
> 
> * missed string.h(strerror_l) on my first check;
>not sure if you can implement that easily on Windows?
> 
> * GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);
> 
> * GNU also defines string.h(str[n]casecmp_l) functions as an extension,
>as well as in POSIX specified strings.h.

I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.


Thanks a lot,
Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat


signature.asc
Description: PGP signature

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

2016-08-23 Thread Brian Inglis 

On 2016-08-20 15:37, Corinna Vinschen wrote:

I uploaded a new Cygwin test release 2.6.0-0.11.
(Still) two interesting changes in this test release:
- Fix the bug reported in
  https://cygwin.com/ml/cygwin/2016-08/msg00357.html
- Add missing nl_langinfo_l functions as reported in
  https://cygwin.com/ml/cygwin/2016-08/msg00349.html
The 2.6.0 release is going to introducing the locale_t datatype, as well
as all functions related to locale_t locales and per-thread locales per
POSIX-1.2008.
So, rather than just providing a single, per-process locale, you can now
create new locales ("newlocale") and set it as locale for the current
thread ("uselocale") or use it directly with one of the new functions
taking a locale_t as parameter (i.e. isalpha_l).
The full list of new interfaces is:
- POSIX:
newlocale, freelocale, duplocale, uselocale, nl_langinfo_l
isalnum_l, isalpha_l, isblank_l, iscntrl_l, isdigit_l, isgraph_l,
islower_l, isprint_l, ispunct_l, isspace_l, isupper_l,
iswalnum_l, iswalpha_l, iswblank_l, iswcntrl_l, iswctype_l, iswdigit_l,
iswgraph_l, iswlower_l, iswprint_l, iswpunct_l, iswspace_l, iswupper_l,
iswxdigit_l, isxdigit_l
tolower_l, toupper_l, towctrans_l, towlower_l, towupper_l,
wctrans_l, wctype_l
strcasecmp_l, strcoll_l, strncasecmp_l, strxfrm_l
wcscasecmp_l, wcscoll_l, wcstrncasecmp_l, wcstrxfrm_l
strfmon_l, strftime_l
- GNU extensions:
isascii_l, toascii_l
strtol_l, stroll_l, strtoul_l, strtoull_l
strtod_l, strtofl, strtold_l
wcstol_l, wcsoll_l, wcstoul_l, wcstoull_l
wcstod_l, wcstofl, wcstold_l


Hi folks,

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:

* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?

* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);

* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

POSIX   CYGWIN  GLIBC
headers ctype.h
functions   isalnum_l   isalnum_l   isalnum_l
functions   isalpha_l   isalpha_l   isalpha_l
functions   isblank_l   isblank_l   isblank_l
functions   iscntrl_l   iscntrl_l   iscntrl_l
functions   isdigit_l   isdigit_l   isdigit_l
functions   isgraph_l   isgraph_l   isgraph_l
functions   islower_l   islower_l   islower_l
functions   isprint_l   isprint_l   isprint_l
functions   ispunct_l   ispunct_l   ispunct_l
functions   isspace_l   isspace_l   isspace_l
functions   isupper_l   isupper_l   isupper_l
functions   isxdigit_l  isxdigit_l  isxdigit_l
functions   tolower_l   tolower_l   tolower_l
functions   toupper_l   toupper_l   toupper_l
functions   toascii_l   toascii_l
functions   isascii_l   isascii_l
headers wctype.h
functions   iswalnum_l  iswalnum_l  iswalnum_l
functions   iswalpha_l  iswalpha_l  iswalpha_l
functions   iswblank_l  iswblank_l  iswblank_l
functions   iswcntrl_l  iswcntrl_l  iswcntrl_l
functions   iswctype_l  iswctype_l  iswctype_l
functions   iswdigit_l  iswdigit_l  iswdigit_l
functions   iswgraph_l  iswgraph_l  iswgraph_l
functions   iswlower_l  iswlower_l  iswlower_l
functions   iswprint_l  iswprint_l  iswprint_l
functions   iswpunct_l  iswpunct_l  iswpunct_l
functions   iswspace_l  iswspace_l  iswspace_l
functions   iswupper_l  iswupper_l  iswupper_l
functions   iswxdigit_l iswxdigit_l iswxdigit_l
functions   towctrans_l towctrans_l towctrans_l
functions   towlower_l  towlower_l  towlower_l
functions   towupper_l  towupper_l  towupper_l
functions   wctrans_l   wctrans_l   wctrans_l
functions   wctype_lwctype_lwctype_l
headers locale.h
functions   duplocale   duplocale   duplocale
functions   freelocale  freelocale  freelocale
functions   newlocale   newlocale   newlocale
functions   uselocale   uselocale   uselocale
headers wchar.h
functions   wcscasecmp_lwcscasecmp_lwcscasecmp_l
functions   wcscoll_l   wcscoll_l   wcscoll_l
functions   wcsncasecmp_l   wcsncasecmp_l   wcsncasecmp_l
functions