> Thanks a lot, the first one worked fine. > I still don't understand why \n[a-z didn't work with substitute.
Without your data and perhaps your vimrc, it's hard to tell. I tried this :%s/\n\s*\([a-z]\)/ \1/g on my vim and it worked fine using the following file: t...@rubbish:~$ xxd -c8 test.txt 0000000: 7468 6973 2069 7320 this is 0000008: 610a 7465 7374 206f a.test o 0000010: 660a 6d75 6c74 6970 f.multip 0000018: 6c65 206c 696e 6573 le lines 0000020: 0a You might do a check on your source data to make sure that 1) you accommodate any extra spaces that might fall at the beginning-of-line 2) you really do have lower-case letters (not some funky unicode look-alike) 3) your newlines aren't out of whack with your file-format (CR vs. NL vs CR/LF). Vim's *usually* pretty good about making a smart guess on this 4) try it in a vim with no startup file: vim -u NONE test.txt to see if some vimrc setting is screwing things up (perhaps something regarding "magic"ness) Just a few ideas, -tim --~--~---------~--~----~------------~-------~--~----~ You received this message from the "vim_use" maillist. For more information, visit http://www.vim.org/maillist.php -~----------~----~----~----~------~----~------~--~---
