On Thu, 2008-09-18 at 13:53 +0000, Tzafrir Cohen wrote: > On Thu, Sep 18, 2008 at 02:29:29PM +0100, michael wrote: > > I've been struggling to get this to work but I think 'sed' should be > > able to do it if I could just get some help with the correct > > incantation... > > > > > > given a file with many strings, include many of the form > > <a href="some url or other">www</a> > > > > I wish each to be transformed to > > <a href="some url or other">some url or other</a> > > > > ie so literal 'www' is replaced with the actual URL for each > > occurance... > > Maybe: > > sed -e '\(<a href="\([^"]*\)">\)www\(</a>\)|\1\2\3|g' >
yes, I've just worked out the subexpressions, thanks to most useful page http://rute.2038bug.com/node11.html.gz#SECTION001170000000000000000 so for my particular case I can use sed -e 's%\(<a\ href\=\)\(\"\)\([^ ]*\)\(\"\)\(>www\)%\1\2\3\4\>\3%g' which is very sim to yours (NB my example here may not be precise since I have copied it from a script which is doing something slightly differently) > -- > Tzafrir Cohen | [EMAIL PROTECTED] | VIM is > http://tzafrir.org.il | | a Mutt's > [EMAIL PROTECTED] | | best > ICQ# 16849754 | | friend > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
