Re: About word splitting in read

[Sven Wischnowsky <wischnow@xxxxxxxxxxxxxxxxxxxxxxx>]

Andrej Borsenkow wrote:

> The last patch reminded me ...
> 
> Currently we can
> 
> - either read in the whole line (with length unknown in advance) theat
> is IFS-splitted
> - or read in fixed number of raw characters
> 
> Several times I wished, we could read in the whole line but in *raw*
> mode, that is, without any word splitting. It is somewhat hard to
> implement this, as we end up with too long/incomplete line so we need
> some sort of buffered IO on top of read etc
> 
> How hard would be to add this to read? With semantic "read in the whole
> line and assign it to scalar unchanged".

Errr... in what way does `read -r foo' not do what you want?

Bye
 Sven


--
Sven Wischnowsky                         wischnow@xxxxxxxxxxxxxxxxxxxxxxx