[Info-vax] Pascal question

[John Reagan]

On Tuesday, January 22, 2019 at 2:58:47 AM UTC-5, Brian_R wrote:
> On Tuesday, January 22, 2019 at 1:44:04 AM UTC, John Reagan wrote:
> > On Monday, January 21, 2019 at 8:34:32 PM UTC-5, Arne Vajhøj wrote:
> > > On 1/21/2019 8:23 PM, Bill Gunshannon wrote:
> > > > On 1/21/19 6:56 PM, Dave Froble wrote:
> > > >> Actually, while at the risk of wasting your valuable time, I'd ask, 
> > > >> why would Pascal as implemented by you not already have such a 
> > > >> function? Understand, I don't use Pascal.  Was string handling not a 
> > > >> featured part of the language?
> > > > 
> > > > The original Pascal as defined by Niklaus Wirth didn't have
> > > > a string type at all.
> > > 
> > > First version of his paper actually had an alfa type that seems
> > > very similar to a string.
> > > 
> > > In the 1973 version and forward it was all packed array of char.
> > > 
> > > Arne
> > 
> > ALFA was a PACKED ARRAY [1..10] OF CHAR.  60-bits to fit in a CDC-6400 register with 6-bit characters.
> > 
> > Extended Pascal added a variable string STRING() type that is almost identical to our VARYING OF CHAR type (which itself was modeled on PL/1's VARYING OF CHAR).
> > 
> > As for string manipulation, like upcase, it hasn't been a popular request.
> 
> Why would it be, its handled quite nicely by the OS libraries. I use VMS Pascal on a daily basis and we use STR$xxxxx for standard string handling. 
> 
> Now additional routines (RTL preferred) for the manipulation of JSON structures might be a plan.

LLVM has a pretty robust YAML library inside of it.  It is used to serialize bitcode files between frontends, optimizers, code generator, etc.  I suppose providing that with perhaps a VMS-centric interface would be helpful.