Pádraig Brady <[email protected]> wrote: > Pádraig Brady wrote: >> Well it's more awkward to output the warning for all cases, >> as currently we can't distinguish between short reads from >> a pipe, and a partial block at the end of the stream. >> Thus the error message ".. past end of input file" >> is not appropriate for both cases. >> Therefore I split out the 2 error cases in the patch below. >> It's more consistent but I'm a little worried it's too involved. > > It is too involved actually. I'll just do a more generic error message: > > if (us_blocks || (!input_offset_overflow && us_bytes)) > { > error (0, 0, > _("%s: unable to skip to requested offset"), > quote (input_file)); > }
Sounds good. Thanks. _______________________________________________ Bug-coreutils mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-coreutils
