Hi, Thanks for the warning. As far as the specs are, a semicolon is not allowed as data in the datastream (yet????) (of an electronic banking backend application). Some files however don't use the doublequote combined with the semicolon, but just the semicolon to identify fields.
So far the semicolon is probably the best way to split the data. So the question is more like, howmuch time to put in an possible change from some side in the future. Bernard On Tuesday 30 August 2005 23:40, Scott Taylor wrote: > Wiggins d'Anconia said: > > Scott Taylor wrote: > > I did not. > > >>>Hi, > >>> > >>>How can I get all the " characters out of a csv file. > >>> > >>>Input looks like > >>>"bla bla";"bla bla";"bla bla" > >>> > >>>and it has to look like > >>>bla bla;bla bla; bla bla > >>> > >>>I tried $text=~ tr(#\"##); > >>>but perl keeps complaining about "Might be a runaway multi-line ;;" > > > > On a more serious note. Simply removing *all* double quotes can be a > > dangerous proposition unless your data really does look like "bla bla" > > which I somewhat doubt. Usually a delimited file like this will quote > > specific fields because they may contain the delimiter, in this case a > > semi-colon. Are you sure "bla bla" can't be "blah; blah blah", and then > > in the future need to still be delimited by the semi-colon? > > > > Just checking... > > It is not uncommon to find CSV output in the format he describes. Often I > need to remove all the quotes out of simple data streams like that. > > -- > Scott -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] <http://learn.perl.org/> <http://learn.perl.org/first-response>
