| View previous topic :: View next topic |
| Author |
Message |
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
 Posted: Sun Sep 22, 2024 5:13 pm Post subject: eix doubt, what does ^t after the version mean? |
 |
|
Hello guys,
I run
| Code: | eix vagrant
[I] app-emulation/vagrant
Available versions: (~)2.4.0-r1^t (~)2.4.1^t {test RUBY_TARGETS="ruby31 ruby32 ruby33"}
Installed versions: 2.4.1^t(12:42:23 PM 08/30/2024)(-test RUBY_TARGETS="ruby31 ruby32 -ruby33")
Homepage: https://vagrantup.com/
Description: A tool for building and distributing development environments
[I] dev-ruby/vagrant_cloud
Available versions: (~)3.1.1^t (~)3.1.1-r1^t {doc test RUBY_TARGETS="ruby31 ruby32 ruby33"}
Installed versions: 3.1.1-r1^t(02:43:52 PM 09/02/2024)(-doc -test RUBY_TARGETS="ruby31 ruby32 -ruby33")
Homepage: https://github.com/hashicorp/vagrant_cloud
Description: Vagrant Cloud API Library
|
what does ^t mean?
Is there anywhere I can refer to?
Thanks for your attention.
Last edited by Rockman8 on Thu Oct 03, 2024 8:46 pm; edited 2 times in total |
|
| Back to top |
|
 |
Hu
Administrator
Joined: 06 Mar 2007
Posts: 22650
|
| Posted: Sun Sep 22, 2024 5:38 pm Post subject: |
|
|
| Per the manual page man eix, section OUTPUT, this indicates the package has RESTRICT=test. |
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Sun Sep 22, 2024 10:33 pm Post subject: |
|
|
found it, thanks
It was in plain sight, sorry for the trouble, tried using ^ instead of ˆ and all those letters cluttered together made me miss the whole thing.
| Code: |
1.0*ilvsˆfmpbstuidP{tbz2,gpkg:3,pak:2}
This is version 1.0 which is subject to PROPERTIES="interactive live virtual set" as well as RESTRICT="fetch mirror primaryuri binchecks strip test userpriv installsources bindist parallel" Moreover, a *.tbz2, three *.gpkg.tar, and two *.xpak files (binary packages without or with FEATURES=binpkg-multi-instance or BINPKG_FORMAT=gpkg, respectively) exist for that version in PKGDIR.
5.0-r3(5.0R3)ˆf or 5.0-r3:5.0R3ˆf
This is version 5.0-r3 which will be installed into SLOT 5.0R3 and has fetch restrictions.
|
|
|
| Back to top |
|
|
Hu
Administrator
Joined: 06 Mar 2007
Posts: 22650
|
| Posted: Sun Sep 22, 2024 11:31 pm Post subject: |
|
|
| No need to apologize. This is a hard one to find via text search, so I pointed you to the section instead of just citing a search term. When I was preparing my first reply, I knew what I was looking for (that ^t means RESTRICT=test), but to find the citation, I still had to resort to reading the manual page, because search wasn't working well. Searching for test would technically find it, but there were so many other hits for test in other parts of the text that finding it via search was tedious. |
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Mon Sep 23, 2024 12:56 am Post subject: |
|
|
| Hu wrote: | | No need to apologize. This is a hard one to find via text search, so I pointed you to the section instead of just citing a search term. When I was preparing my first reply, I knew what I was looking for (that ^t means RESTRICT=test), but to find the citation, I still had to resort to reading the manual page, because search wasn't working well. Searching for test would technically find it, but there were so many other hits for test in other parts of the text that finding it via search was tedious. |
Even though you pointed me where to look at, would you believe that I was struggling to find that sign ( ˆ )?
Now out of curiosity
This symbol ( ˆ )is present on the eix manual Output section
| Code: | | 1.0*ilvsˆfmpbstuidP{tbz2,gpkg:3,pak:2} |
is different from eix actual output ( ^ )
| Code: | [I] app-emulation/vagrant
Available versions: (~)2.4.0-r1^t (~)2.4.1^t {test RUBY_TARGETS="ruby31 ruby32 ruby33"}
Installed versions: 2.4.1^t(12:42:23 PM 08/30/2024)(-test RUBY_TARGETS="ruby31 ruby32 -ruby33")
Homepage: https://vagrantup.com/
Description: A tool for building and distributing development environments
[I] dev-ruby/vagrant_cloud
Available versions: (~)3.1.1^t (~)3.1.1-r1^t {doc test RUBY_TARGETS="ruby31 ruby32 ruby33"}
Installed versions: 3.1.1-r1^t(02:43:52 PM 09/02/2024)(-doc -test RUBY_TARGETS="ruby31 ruby32 -ruby33")
Homepage: https://github.com/hashicorp/vagrant_cloud
Description: Vagrant Cloud API Library |
perhaps changing the symbol used in the manual would ease a little for those searching the manual since the actual symbol usage is different ( ˆ != ^ )? |
|
| Back to top |
|
|
Hu
Administrator
Joined: 06 Mar 2007
Posts: 22650
|
| Posted: Mon Sep 23, 2024 2:38 am Post subject: |
|
|
| That seems like a useful improvement, in my opinion. |
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Mon Sep 23, 2024 2:45 am Post subject: |
|
|
| Hu wrote: | | That seems like a useful improvement, in my opinion. |
So where should I send my suggestion, so that it would be evaluated? |
|
| Back to top |
|
|
Banana
Moderator
Joined: 21 May 2004
Posts: 1730
Location: Germany
|
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Tue Sep 24, 2024 12:06 am Post subject: |
|
|
Well just got my answer from https://github.com/vaeth/eix
And it seems to be a locale issue.
I'm gonna quote both answers that I got from there
vaeth said
| Quote: |
In the manpage as well as in eix the symbol ^ (the ASCII symbol, not the utf8 symbol) is specified.
If your manpage or terminal program displays this symbol in one case differently, it is IMHO the fault of the manpage program (or manpage formatter) or terminal program.
That being said, if you want to change the symbol in the eix output, you can do so locally by setting FORMAT_RESTRICTSEPARATOR in e.g. some file in /etc/eixrc correspondingfly, see the output of eix --dump|grep FORMAT_RESTRICTSEPARATOR. However, by default, I will change neither the manpage nor eix to a non-ASCII-symbol. |
chiitoo said
| Quote: |
I suppose this is a localisation issue of some kind, somewhere.
While the sources [1] seem to indeed use a caret, it becomes a circumflex for me as well when viewing the manual normally.
If I do something like LANG="" man eix, then the caret is there instead.
My usual LANG is set to en_GB.UTF-8.
https://github.com/vaeth/eix/blob/main/manpage/en-eix.1.in#L1084 |
So how to set manpages to lang="" without having to add LANG="" to man eix
my make.conf has LC_MESSAGES=C.utf8
but eselect locale is set to en_US.utf8
setting my LC_MESSAGES="" would solve the issue? Does eselect locale override make.conf settings? |
|
| Back to top |
|
|
dmpogo
Advocate
Joined: 02 Sep 2004
Posts: 3425
Location: Canada
|
| Posted: Tue Sep 24, 2024 12:56 am Post subject: |
|
|
| How interesting it became ! I wonder, what would be the reason for any locale to change ASCII symbols into something else. That sounds pretty if not dangerous then counterintuitive, given that ASCII part is at the core of coding. |
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Tue Sep 24, 2024 4:26 pm Post subject: |
|
|
| dmpogo wrote: | | How interesting it became ! I wonder, what would be the reason for any locale to change ASCII symbols into something else. That sounds pretty if not dangerous then counterintuitive, given that ASCII part is at the core of coding. |
If you go into terminal and try to find the eix's Output section perhaps you gonna see a little square there, well that's what I saw which got me confused and made me wonder what encoding I'm running in terminal?
edit:
Running
Found out that I'm running UTF-8 enconding.
| Code: | env
LANG=en_US.utf8
LANGUAGE="" |
And according to this post
https://stackoverflow.com/questions/5306153/how-to-get-terminals-character-encoding
Since this about terminal emulator I'm assuming that man pages is using env. variables to set the locale
As in
| Code: | $man man
Finding manual pages
-L locale, --locale=locale
man will normally determine your current locale by a call to the C function setlocale(3) which interrogates various environment variables, possibly including $LC_MESSAGES and $LANG. To temporarily override the determined value, use this option to supply a locale string directly to man. Note that it will not take effect until the search for pages actually begins. Output such as the help message will always be displayed in the initially determined locale.
|
Searching and decopressing the file got me this
| Code: | $file eix.1
eix.1: troff or preprocessor input, ASCII text, with very long lines (424) |
given all that for now there are only a few counter-intuitive solutions, as they all involve adding env. variables or options and the user must be aware that man pages are being encoded to utf8 or UTF-8.
| Code: | LANG="" man eix
man -E ascii eix or man --encode=ascii eix
man -E latin1 eix or man --encode=latin1 eix |
|
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Thu Sep 26, 2024 5:12 pm Post subject: |
|
|
Create a file on /etc/env.d/
| Code: | /etc/end.d/00manopt
=========================================
MANOPT="-Elatin1
#--encoding=latin1" |
or
| Code: | /etc/end.d/00manopt
=========================================
MANOPT="-Eascii"
#--encoding=ascii |
run
| Code: | | env-update && source /etc/profile |
or if you do not wish to temper with envVar then run
| Code: | man -Elatin1 eix or man -E latin1 eix
or
man --encoding=latin1 eix
=========================================
man -Eascii eix or man -E ascii eix
or
man --encoding=ascii eix |
Last edited by Rockman8 on Fri Sep 27, 2024 4:59 pm; edited 1 time in total |
|
| Back to top |
|
|
flexibeast
Guru
Joined: 04 Apr 2022
Posts: 449
Location: Naarm/Melbourne, Australia
|
| Posted: Fri Sep 27, 2024 8:51 am Post subject: |
|
|
The fundamental issue here isn't locales. What's actually going on here is that man pages aren't being written correctly (whether directly, or indirectly via some generating tool).
By default, on Gentoo, the program used to render man pages is groff(1). groff(1) is the frontend to the GNU roff document formatting system; roff is a markup language. The source of man pages is roff, using either the man(7) macro set (typical on Linux) or the mdoc(7) macro set (typical on BSDs). groff, as well as another program to render man pages, mandoc(1), can handle both.
In roff, a bare '^' gets represented in the output as a circumflex, as described in the following table from the groff_char(7) man page:
| Code: | | ^ ˆ modifier circumflex \(ha circumflex/caret/“hat” |
mandoc(1) does the same, as described in the mandoc_char(7) man page:
| Code: | | ^ U+02C6 modifier letter circumflex |
Man pages shouldn't be using bare '^'; they should be using '\(ha' in their source to represent that character. So they either need to be fixed manually, or the tools that generate such output need to be fixed. |
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Fri Sep 27, 2024 4:59 pm Post subject: |
|
|
| Quote: | | Man pages shouldn't be using bare '^'; they should be using '\(ha' in their source to represent that character. So they either need to be fixed manually, or the tools that generate such output need to be fixed. |
Indeed, changed the character ^ by \(ha like so
| Code: | eix.1
...
5.0-r3(5.0R3)\(haf " or " 5.0-r3:5.0R3^f
... |
A simple test can be done by just running
| Code: | | echo "5.0-r3(5.0R3)\(haf " or " 5.0-r3:5.0R3^f"|groff -Tutf8|sed '/^$/d' |
Output
| Code: | | 5.0‐r3(5.0R3)^f or 5.0‐r3:5.0R3ˆf |
The difference was there, I could find the ^ (ascii) by just typing it ( regex search \^ ) and the other character ˆ (modifier circumflex) was there unaffected, my locale was set to utf8 and the eix manpage is ascii encoded.
On | Code: | | man 7 groff_man_style |
on Notes section
| Code: | Notes
Some tips on troubleshooting your man pages follow.
o Some ASCII characters look funny or copy and paste wrong.
On devices with large glyph repertoires, like UTF-8-capable terminals and PDF, several keyboard glyphs are mapped to code points outside the Unicode basic Latin range because that usually results in better typography in the general case. When documenting GNU/Linux command or C language syntax, however, this translation is sometimes not desirable.
To get a "literal"... ...should be input.
--------------------------------------------
' \(aq
- \-
\ \(rs
^ \(ha
` \(ga
~ \(ti
--------------------------------------------
|
|
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Fri Sep 27, 2024 5:29 pm Post subject: |
|
|
Just reopened the issue on github.
Sent a pull request. |
|
| Back to top |
|
|
Rockman8
n00b
Joined: 19 Sep 2023
Posts: 38
|
| Posted: Sat Oct 19, 2024 1:40 pm Post subject: |
|
|
The pull request has been merged!
Thank you everyone for all the directions given, especially flexibeast for pointing out the real issue. |
|
| Back to top |
|
|
|
Portage & Programming
