Re: Should "df --portability" allow thousands separators?

[Peter D.]

On Monday 26 February 2007 15:11, Paul Eggert wrote:
> "Peter D." <address@hidden> writes:
> > I have an environment variable set that triggers thousands separators. 
> > It can be overridden with an extra command line option, but I suspect
> > that a "-P" should be sufficient.
>
> Can you please be more specific?  Which environment variable is that?

Sure.  This line in /etc/bashrc sets thousands separators in ls, du and df.  
    export BLOCK_SIZE=\'1
The escaped apostrophe does the magic.  

> What does the output look like, when you set that variable?  

Here are a couple of examples with the environment variable BLOCK_SIZE set 
as above...

address@hidden ~
->df .
Filesystem           1B-blocks      Used Available Use% Mounted on
/dev/hdc7            4,120,768,512 1,770,029,056 2,350,739,456  43% /home
address@hidden ~
->df -P .
Filesystem            1-blocks      Used Available Capacity Mounted on
/dev/hdc7            4,120,768,512 1,770,029,056 2,350,739,456      
43% /home
address@hidden ~
->df -P --block-size=1 .
Filesystem            1-blocks      Used Available Capacity Mounted on
/dev/hdc7            4120768512 1770029056 2350739456      43% /home
address@hidden ~
->df -P --block-size=\'1 .
Filesystem            1-blocks      Used Available Capacity Mounted on
/dev/hdc7            4,120,768,512 1,770,029,056 2,350,739,456      
43% /home

Sorry about the line wrap, also my bash prompt is a multicolored two line 
thing, but I am sure you can make sense of if.  

> Also, 
> what locale are you using (what is the output of the 'locale'
> command)?

address@hidden ~
->locale
LANG=en_US.UTF-8
LC_CTYPE=en_US.UTF-8
LC_NUMERIC=en_AU.UTF-8
LC_TIME=en_US.UTF-8
LC_COLLATE=en_US.UTF-8
LC_MONETARY=en_AU.UTF-8
LC_MESSAGES=en_US.UTF-8
LC_PAPER=en_AU.UTF-8
LC_NAME=en_AU.UTF-8
LC_ADDRESS=en_AU.UTF-8
LC_TELEPHONE=en_AU.UTF-8
LC_MEASUREMENT=en_AU.UTF-8
LC_IDENTIFICATION=en_AU.UTF-8
LC_ALL=

> It could well be that 'df' is conforming to POSIX, but nonetheless
> could be improved.

I am sufficiently ignorant of POSIX that df could be in completely compliant 
and I would not know.  

Unfortunately there are cases where code (wrongly?) calls "df -P" and fails 
because of the commas.  Here is a solved example, 
<http://qa.mandriva.com/show_bug.cgi?id=28489>.

Maybe code that calls "df --portability" and fails because of commas is just 
wrong.  I don't know.  


-- 
sig goes here...
Peter D.