We're getting a little far from my real question, since caching the examples is
IMHO a hack to cache the most expensive part of the computation, while the whole
'doc ought to be cachable, but....
I forgot what my own code was doing.
That's basically what I'm doing:
(define (make-cached-eval name . rest)
(define ev
(make-log-based-eval
(cachefile name)
(if (file-exists? (cachefile name))
'replay
'record)))
(for ([r rest])
(ev r))
ev)
I'm using with-cache's `cachefile` just to generate and manage the file path, so
it ends up in have a good place.
I was trying to do something more complex earlier, but it didn't work.
On Wed, Mar 25, 2020 at 09:59:15PM +0100, Ryan Culpepper wrote:
> I'm not clear on what `with-cache` is doing in this setup, but it seems
> like a potential source of errors. If the goal is to automatically use
> `'replay` if the log file exists and `'record` otherwise, why not do the
> following?
>
> (make-log-based-eval the-log-file (if (file-exists? the-log-file)
> 'replay 'record))
>
> Ryan
>
>
> On Wed, Mar 25, 2020 at 8:57 PM William J. Bowman <w...@williamjbowman.com>
> wrote:
>
> > On Wed, Mar 25, 2020 at 08:51:18PM +0100, Ryan Culpepper wrote:
> > > You can use `raco make` (or `raco setup` for docs of installed packages)
> > to
> > > compile the Scribble files, but that won't compile the examples. Those
> > are
> > > dynamically evaluated when the Scribble documents are run.
> > Yeah, I was thinking of "compilation" as in caching the output document
> > from
> > each module.
> >
> > > For `make-log-based-eval`, are you using a separate evaluator (and
> > separate
> > > log file) for each Scribble file?
> > Yes. However, I'm using `with-cache` and a wrapper to detect whether a
> > cache
> > file exists (separate cache file each evaluator), and use 'replay mode if
> > the
> > cache file exists, so I don't have to manually switch to 'replay mode, or
> > manually re-record if I alter an example (instead, just clear the cache).
> >
--
You received this message because you are subscribed to the Google Groups
"Racket Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to racket-users+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/racket-users/20200325220252.GA31619%40williamjbowman.com.
