On Wed, 15 Jun 2022, at 05:35, Jeremy Nicoll wrote: > That implies that the parser looks for possible ending > delimiters before it parses the contents of the delimited > string. > > It's not how I'd have designed it! I'd have gone with > a strict left to right character by character lexical scan.
Oops, must think more carefully. This is what comes of writing a reply at 5am... I assumed, without engaging my brain, that the repeated single quote in your example meant that ispf edit would regard that as an escaped single quote, but apparently it doesn't. So if it's not considering escaped chars in a delimited string the simpler parsing that looks first for possible end delimiters is easier to understand. Also, would CHANGE PARM ""Hello," says O'Reilly." fail because the space after the double quote after the comma would also be taken to imply that the delimited string ends there? That is "says O'Reilly." is unexpected surplus? -- Jeremy Nicoll - my opinions are my own. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
