Return to Answer

included da move for completeness

edited Jul 30, 2011 at 20:41

10.8k
3
37
59

for dateNfile in $(grep dateishere *html | sed -r 's/^([^:]*):(.*)dateishere/\2:\1/' | sort -n) do echo name=${dateNfile/*:/}   # echo date=${dateNfile/:*/} mv $name /other/location done

So I guess there is just one line per file, containing 'dateishere', and that the line starts with the date, but the sed command could be modified, to cut the date even in other cases.

If you have blanks in the filenames, or linebreaks and such, things get a bit more complicated, so I would like to ignore these, if possible. But from your assertion, this isn't the case.

for dateNfile in $(grep dateishere *html | sed -r 's/^([^:]*):(.*)dateishere/\2:\1/' | sort -n) do echo name=${dateNfile/*:/} echo date=${dateNfile/:*/} done

So I guess there is just one line per file, containing 'dateishere', and that the line starts with the date, but the sed command could be modified, to cut the date even in other cases.

for dateNfile in $(grep dateishere *html | sed -r 's/^([^:]*):(.*)dateishere/\2:\1/' | sort -n) do name=${dateNfile/*:/}   # echo date=${dateNfile/:*/} mv $name /other/location done

So I guess there is just one line per file, containing 'dateishere', and that the line starts with the date, but the sed command could be modified, to cut the date even in other cases.

If you have blanks in the filenames, or linebreaks and such, things get a bit more complicated, so I would like to ignore these, if possible. But from your assertion, this isn't the case.

Source Link

answered Jul 30, 2011 at 19:58

user unknown

10.8k
3
37
59

for dateNfile in $(grep dateishere *html | sed -r 's/^([^:]*):(.*)dateishere/\2:\1/' | sort -n) do echo name=${dateNfile/*:/} echo date=${dateNfile/:*/} done

So I guess there is just one line per file, containing 'dateishere', and that the line starts with the date, but the sed command could be modified, to cut the date even in other cases.