Re: [Factor-talk] Iterating over arrays and string

2013-01-21 Thread John Benediktsson
Good idea. On Sun, Jan 20, 2013 at 8:51 PM, Alex Vondrak wrote: > What about adding to the list at > https://github.com/slavapestov/factor/blob/master/basis/compiler/tree/propagation/known-words/known-words.factor#L224as > a stop gap? > > --Alex Vondrak > > > On Sun, Jan 20, 2013 at 11:44 AM,

Re: [Factor-talk] Iterating over arrays and string

2013-01-20 Thread Alex Vondrak
What about adding to the list at https://github.com/slavapestov/factor/blob/master/basis/compiler/tree/propagation/known-words/known-words.factor#L224as a stop gap? --Alex Vondrak On Sun, Jan 20, 2013 at 11:44 AM, John Benediktsson wrote: > I noticed this before and made a patch: > > > https://g

Re: [Factor-talk] Iterating over arrays and string

2013-01-20 Thread John Benediktsson
I noticed this before and made a patch: https://github.com/slavapestov/factor/commit/cd6026e70efef3416d08160cd7cbaa5f67d8e0c7 But then reverted it, because of some test failures that asserted the image size was below a certain size. https://github.com/slavapestov/factor/commit/3b94eb5fb78d72af

Re: [Factor-talk] Iterating over arrays and string

2013-01-20 Thread Joe Groff
On Sun, Jan 20, 2013 at 8:20 AM, Alex Vondrak wrote: > and noticed the same types being inferred (object vs string). So, looking > at the definitions, > > IN: scratchpad \ >array see > USING: sequences ; > IN: arrays > : >array ( seq -- array ) { } clone-like ; > IN: scratchpad \ >stri

Re: [Factor-talk] Iterating over arrays and string

2013-01-20 Thread Alex Vondrak
> This is a bug. It should know that `M\ array like` returns an array. (Why > it's unable to infer that from the definition isn't immediately obvious to > meā€”from the definition of `M\ array like` it seems to me it ought to.) > Monkey-patching an `{ array } declare` onto the end of the method makes

Re: [Factor-talk] Iterating over arrays and string

2013-01-19 Thread Joe Groff
On Sat, Jan 19, 2013 at 5:35 PM, Jon Harper wrote: > Here are my questions: > 1) in bar1, "" like is enough to allow the compiler to infer type and > generate code using string-nth-fast and set-string-nth-fast instead of > nth-unsafe and set-nth-unsafe (just like using TYPED: in bar2); However, i

[Factor-talk] Iterating over arrays and string

2013-01-19 Thread Jon Harper
Hi list, after some experiments, I was surprised with how the optimizing compiler could infer types. Not sure if all those behaviors are expected.. If anyone has some explanations, that would be great. Please see this paste for code: https://paste.factorcode.org/paste?id=2836 Here are my questions