On 11/30/05, Tim <ignored_mailbox@xxxxxxxxxxxx> wrote: > Tim: > > >> I used to know of something that did that (something with various > >> different reformatting options for massaging text files), but I can't > >> think what it was. Quickly looking at the man file for the cat program, > >> you could do something like: > >> > >> cat --squeeze-blank inputfilename -> outputfilename > > Paul Smith: > > Thanks, Tim and Paul. Paul's method does not mysteriously work: > > > > $ more file1.txt > > word1 > > > > > > > > word2 > > > > word3 > > $ more -s file1.txt > file2.txt > > $ more file2.txt > > word1 > > > > > > > > word2 > > > > word3 > > Hmm, seems to work for me. Both with the "less" and "more" programs, as > well as the "cat" program. > > > Tim's way works partially, i.e., many blank lines are in effect > > erased, but some remain. I suspect that the left blank lines are not > > blank lines although they look like blank lines. Can one go further > > with deleting the left "false" blank lines? > > In what way do they remain? Can you provide an actual example? (Rather > than an explanation of what's happening.) > > What I see is that all consecutive blank lines are replaced by a single > blank line, on the file I tried it with. > > e.g. Tested on /etc/selinux/targeted/contexts/files/file_contexts > > If you want to remove all blank lines, then perhaps you could use grep. Thanks, Tim. Rodolfo's technique works fine for me. It reduced about 40Kb of a HTML document produced by NVU. I do not know why, but NVU seems to add blocks of blank lines, drastically increasing the size of the document. Paul
