Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Nov 28 00:07
    gilch opened #138
  • Nov 27 23:56
    gilch edited #137
  • Nov 27 23:56
    gilch commented #137
  • Nov 27 23:48
    gilch opened #137
  • Nov 27 23:41

    gilch on update-contributing

    Update for current practice (compare)

  • Nov 27 22:01
    majidaldo starred gilch/hissp
  • Nov 26 04:12
    sguzman starred gilch/hissp
  • Nov 25 22:01

    gilch on gilch-patch-quickstart

    (compare)

  • Nov 25 22:00

    gilch on master

    Reorder to fix issue with untes… Merge pull request #136 from gi… (compare)

  • Nov 25 22:00
    gilch closed #136
  • Nov 25 22:00
    gilch opened #136
  • Nov 25 21:54

    gilch on gilch-patch-quickstart

    Reorder to fix issue with untes… (compare)

  • Nov 19 14:29
    matt-hughes starred gilch/hissp
  • Nov 18 22:41
    ihavenonickname starred gilch/hissp
  • Nov 15 05:00
    codecov[bot] commented #135
  • Nov 15 04:59
    gilch opened #135
  • Nov 15 04:59

    gilch on gilch-patch-1

    typo (compare)

  • Nov 14 13:21
    i4get2code starred gilch/hissp
  • Nov 13 04:57
    boticello starred gilch/hissp
  • Nov 08 12:37
    italy starred gilch/hissp
Kyle I S Harrington
@kephale
i understand your reasoning, and maybe it lines up with the forced indenting style of python
but since tooling isnt provided for hissp or default configurations
it does seem like an extra hurdle for new users
gilch
@gilch
Looks like Parinfer is configurable, so it might depend on which plugin.
gilch
@gilch
I agree it's a hurdle for new users. I still don't think that allowing tabs is worth it, as it is just a config change. Once and done. Once Lissp is more stable, maybe it can have dedicated tooling that handles this. Then they just install the plugin for their editor and it configures that for them.
gilch
@gilch
I'm hoping to get the next release out some time this weekend. At the rate I'm going, I feel optimistic.
Kyle I S Harrington
@kephale
huzzah!
gilch
@gilch
I've accomplished my issues for the 0.2.0 release. I still need to review and merge the copy edit, then I just have to build, manually test, and release the new package.
I can probably get that done tomorrow.
Kyle I S Harrington
@kephale
:confetti_ball:
gilch
@gilch
pip install hissp
@/all Hissp 0.2.0 is up on PyPI.
Kyle I S Harrington
@kephale
:thumbsup: awesome
gilch
@gilch
Hissp has officially hit 100 stars on GitHub!
Kyle I S Harrington
@kephale
nice!
in the last reddit about Hy i mentioned that Hissp turned out to be the better choice for me ;)
gilch
@gilch
I had upvoted your comment ;)
That did drive some traffic to the GitHub page.
mgrasko
@mgrasko
I am one of those from Kyle I S Harrington Reddit post. I have to say I am very impressed with Hissp. Next week I plan to write some actual usefull code and will let you know. Right now everything seems to work fine and it seems like something that will let me stay sane using python ecosystem after reading SICP
also there is really no exposure, if not for this single reddit mention I would never stumble upon hissp
gilch
@gilch
Welcome.
Daehwan Nam
@daehwannam
I'm new to Hissp and feel it's cool! I need to run lisp code which is dynamically created during run time. At first, I tried Hy's eval function but It's too slow. Now I'm trying Hissp and found that 'readerless' and Lissp has no big overhead and it's roughly 40 times faster than Hy's eval!
gilch
@gilch
I'm always happy to hear that Hissp is finding use! Hissp's implementation is much simpler than Hy's. That's probably why it's faster, but I had no idea that the difference was that dramatic. Hissp and Hy have very different architectures. There may be something else about Hissp's approach that accounts for the difference.
Mike Pelletier
@salixbipedis:matrix.org
[m]
Thanks for inviting me into your brain. If I want to unit test a scrap of Lissp, I suppose I can just use any Python unit testing framework? Any helpful best practices?
Mike Pelletier
@salixbipedis:matrix.org
[m]
(mod "%s" '(hello)) feels so wrong and makes me never want to overload another operator in my life
2 replies
salt rock lamp
@elementio:salt-rock-lamp.ems.host
[m]
pytest would probably work on the compiled python, and unittest almost certainly would
salt rock lamp
@elementio:salt-rock-lamp.ems.host
[m]
Mike Pelletier: also you can do (define % mod) to paper over the ickiness. but this is why i (personally) like f-strings anyway, or .format if you want to save a "template" to reuse later.
that said, i dont think hissp has f-strings?
salt rock lamp
@elementio:salt-rock-lamp.ems.host
[m]
hm, im reading the docs and i feel like i still have no idea how args and *kwargs work
ah, the : comes before :* args
salt rock lamp
@elementio:salt-rock-lamp.ems.host
[m]
(define fmt (lambda (template : :* args :** kwargs) (.format template : :* args :** kwargs))) is this the right unpacking/packing syntax?
gilch
@gilch
@salixbipedis:matrix.org I've been using the standard library for unit tests, and it's working fine. You can see the tests in the git repository. Some of them are written in Lissp: https://github.com/gilch/hissp/blob/master/tests/test_basic.lissp
1 reply
Pytest does some AST metaprogramming to rewrite assert statements. Hissp doesn't have those (since it has no statements to speak of), so that part wouldn't work, but I expect the rest of it would.
Hissp doesn't have operators, so there's no need to overload them. The %s string formatting was borderline deprecated anyway, so I usually use (.format ...)in Lissp, like I would in Python.
gilch
@gilch
There's nothing stopping you from giving multiple aliases to operators either, so you can continue to use mod on numbers, but you can also (define format mod) (or whatever) and then use (format ...) on strings.
But at that point, why not just use (.format ...)?
While the f-strings technically don't exist in Hissp, they do in Python. You can inject f-strings with .#"..." like any other Python expression. E.g. .#"f'foo{bar}'".
I don't particularly want to encourage the overuse of injections, but the option is there.
gilch
@gilch
If I needed something like an f-string in Clojure, I'd probably just use (str ...) most of the time.
user=> (def person "Bob")
#'user/person
user=> (str "Hello, " person "! It's nice to meet you!")
"Hello, Bob! It's nice to meet you!"
gilch
@gilch
I had a similar analog in the entuple group (some of which you can see in the quickstart).
I think it was something like
def enstr(*args):
    return "".join(map(str, args))
You should be able to use that in about the same way as Clojure's str.
Another problem with injecting f-strings, besides the usual problem of macros becoming unable to read the subexpressions, is that the subexpressions for f-strings have to be written in Python, so you can't use your macros.
gilch
@gilch
In the same way that Lissp creates a bytestring with the b# reader macro, you could theoretically create an f# reader macro in the same way, but this way the subexpressions could be written in Lissp. You'd use a regex to extract them, and then rewrite it to a normal (.format ...) call.
gilch
@gilch
@elementio:salt-rock-lamp.ems.host that's the right syntax, although that particular example could be simplified to (define fmt str.format), assuming you're always using a str type for the template argument. Unbound methods are first-class in Python.
salt rock lamp
@elementio:salt-rock-lamp.ems.host
[m]
first-class, but not equivalent :) it breaks if you try to use partial (and probably some other stuff that I can't think of them right now)
gilch
@gilch
>>> fmt = str.format
>>> from functools import partial
>>> partial(fmt, "Welcome, {}!")
functools.partial(<method 'format' of 'str' objects>, 'Welcome, {}!')
>>> _('Bob')
'Welcome, Bob!'
partial(), at least, worked that time. But there is one case I can think of where they're not equivalent.
Mike Pelletier
@salixbipedis:matrix.org
[m]
exactly what I wanted