Alex and I got the speed of guile-json up to about 19megs a second under guile 3 on a 4 year old intel i5, which may or may not be adequate for your needs, Tim.
I have some code I wrote during that spurt that preforms slightly better than guile-json in general (about 3-4% on my machine. Alex could not repeat those numbers) an on files with many large strings in particular (10% in extreme cases). If that is interesting I could put it online somewhere. The functioning code is a derivate work of guile-json, so the GPL applies. -- Linus Björnstam On Sat, 26 Dec 2020, at 05:28, John Cowan wrote: > I think that's the best bet, although JSON is also a possibility. I'm > working on a more general text-serialization solution, but it will be quite > some time before I have a chance to work on it. > > On Fri, Dec 25, 2020 at 10:28 PM Tim Meehan <btmee...@gmail.com> wrote: > > > I have a big-ish blob of key-value pairs that I would like to not have to > > store as text and then convert to a hashtable when I am filtering my data. > > > > Right now I am saving it as an alist and then converting back to a > > hashtable using "alist->hashtable" from SRFI 69. I was just hoping that > > there was something more clever ... perhaps it is plenty clever enough. > > >
