Bo Peng wrote:
> This may be the most troublesome step (need to figure out which files
> are needed). After that, it is only a replacement of gzstream by
> filtering_ostream( gzip_compressor() | file_sink(...)).
>
> Bo
I've used this in the past:
for file in `grep -r '^ *# *include *<boost' . | \
cut -d'<' -f2 | \
cut -d'>' -f1 | \
sort -u`
do
ls $file > /dev/null
done
It's still iterative, in that you have to bring over those files that are
reported as missing, but it wouldn't be very hard to automate the whole
thing by parsing the "ls" errors and creating a loop.
Anyway, perhaps it helps...
--
Angus
