[fpc-devel] Current "state of opinion" on multiline strings? I still have this up-to-date locally with trunk FPC, despite having neglected pushing this anywhere online for a while now.

Mattias Gaertner nc-gaertnma at netcologne.de
Sun Dec 22 10:26:04 CET 2024



On 12/22/24 03:29, Ben Grasset via fpc-devel wrote:
> Facts:
> - Michael VC did not originally like the entire idea at all, if I recall
> correctly. However, he eventually added an implementation of it to Pas2JS
> that works literally the exact same as my main FPC compiler one does.
> - Delphi (in the time since, as of version 12 I guess) has added this
> themselves. Their version is not quite the same (frankly in my opinion,
> theirs is overall worse than mine, for reasons such as the fact you HAVE to
> put the closing backticks of a multiline string on their own entire line in
> Delphi, whereas in my version the closing backticks could appear anywhere
> and be parsed correctly. And so on and so forth).

Delphi does not support backticks. I guess you mean the odd apostrophs?


> I'm as ready as I was several years ago to put this in main FPC if it'll be
> accepted, however, would people want me to now imitate the Delphi
> version (which again did not exist when I originally implemented the
> functionality for FPC, and again is functionally different from both the
> unmerged main compiler impl of mine and Michael VCs identical
> actually-existing impl for Pas2JS).
> 
> I'd certainly don't really WANT to adapt to the (like I said, worse in my
> opinion) Delphi version, but I would if I "had" to. So that's why I figured
> I'd check in to see what people's general thoughts are on all of this
> nowadays.

Are there cases where ported Delphi code creates a different string literal?

Mattias



More information about the fpc-devel mailing list