Don't display file byte counts with thousands separator

[bknowles]

Folks,

When using the -B option with exa -l, the output will include actual byte counts for each file in question, but those byte counts will include a thousands separator. This is inconsistent with the behaviour of /bin/ls when providing byte counts, and would require that a developer using this command as part of a pipeline will need to be able to detect the thousands separator and remove it, if they are going to try to do any calculations on this number.

I understand the desire to be able to display a large number with a thousands separator, but I do not believe that this should be the default in this case.

Of course, I'm assuming that if you do provide a thousands separator, that this is language and locale appropriate, so while we might see a comma here in the US, other places might see a simple space character or a period or some other character that the language/locale specifies as the thousands separator.

[ogham]

Of course, I'm assuming that if you do provide a thousands separator, that this is language and locale appropriate, so while we might see a comma here in the US, other places might see a simple space character or a period or some other character that the language/locale specifies as the thousands separator.

It does indeed go through the locale routines.

I'm going to defer this issue to the very old feature request #13, which is to provide some sort of structured output, which will probably be JSON. Tools like exa and ls certainly look almost-machine-parseable, but having them be fully machine-parseable is not a road I want to go down without an explicit flag to enable this mode.

[themadsens]

The thousand sep behaviour is extremely obnoxious when cutting/pasting into calculations.
Please consider a flag to control this