Re: How does the input-line test work?

[Chet Ramey]

Hector Chu wrote:
> input-line.sh is:
>
> bash ./input-line.sub
> this line for input-line.sub
>
> with input-line.sub being:
>
> read line
> echo line read by $0 was \`$line\'
>
> The test is run by executing: bash < input-line.sh.
>
> How does input-line.sub manage to read from the correct point in the
> input file? Doesn't the parent bash read the whole lot into its
> buffer, so that the child bash starts reading from the end of the
> file?

No.  That's the point of the test.  Posix requires that shells rewind
or otherwise manipulate the read pointer in a script read from the
standard input to preserve this kind of synchronization between parent
and child processes.

One way to accomplish this is to read the script one character at a
time.  Bash uses a set of buffered input functions (found in input.c)
to handle its own buffering and input stream rewinding.

Chet
--
``The lyf so short, the craft so long to lerne.'' - Chaucer

Chet Ramey, ITS, CWRU    chet@case.edu    http://cnswww.cns.cwru.edu/~chet/