unofficial mirror of emacs-devel@gnu.org 
 help / color / mirror / code / Atom feed
* Stream implementation of seq-mapn
@ 2017-12-18 11:58 Michael Heerdegen
  2017-12-18 12:32 ` Nicolas Petton
  0 siblings, 1 reply; 9+ messages in thread
From: Michael Heerdegen @ 2017-12-18 11:58 UTC (permalink / raw)
  To: Emacs Development; +Cc: Nicolas Petton, Stefan Monnier

[-- Attachment #1: Type: text/plain, Size: 73 bytes --]

Hello,

I want to define `seq-mapn' for streams.  This is what I have:



[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Implement-seq-mapn-for-streams.patch --]
[-- Type: text/x-diff, Size: 1520 bytes --]

From 46caedf7b85f4285e83f72646fa1d6940bd7ea00 Mon Sep 17 00:00:00 2001
From: Michael Heerdegen <michael_heerdegen@web.de>
Date: Mon, 18 Dec 2017 12:49:38 +0100
Subject: [PATCH] Implement `seq-mapn' for streams

---
 packages/stream/stream.el | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/packages/stream/stream.el b/packages/stream/stream.el
index 810adf90d..a0e851015 100644
--- a/packages/stream/stream.el
+++ b/packages/stream/stream.el
@@ -318,6 +318,27 @@ applications of FUNCTION on each element of STREAM in succession."
      (cons (funcall function (stream-first stream))
            (seq-map function (stream-rest stream))))))
 
+(cl-defmethod seq-mapn (function (stream stream) &rest streams)
+  "Map FUNCTION over the STREAMS.
+
+Example: this prints the first ten Fibonacci numbers:
+
+  (letrec ((fibs (stream-cons
+                  1
+                  (stream-cons
+                   1
+                   (seq-mapn #'+ fibs (stream-rest fibs))))))
+    (seq-do #'print (seq-take fibs 10)))
+
+\(fn FUNCTION STREAMS...)"
+  (if (not (cl-every #'streamp streams))
+      (cl-call-next-method)
+    (cl-callf2 cons stream streams)
+    (stream-make
+     (unless (cl-some #'seq-empty-p streams)
+       (cons (apply function (mapcar #'stream-first streams))
+             (apply #'seq-mapn function (mapcar #'stream-rest streams)))))))
+
 (cl-defmethod seq-do (function (stream stream))
   "Evaluate FUNCTION for each element of STREAM eagerly, and return nil.
 
-- 
2.15.1


[-- Attachment #3: Type: text/plain, Size: 229 bytes --]



Questions:

(1) Is it ok to implement it with `cl-call-next-method' this way?

(2) When I do C-h f seq-mapn, the signature of this new method is
printed like

| (arg2 (arg3 stream) &rest rest)

Why is that?


Thanks,

Michael.

^ permalink raw reply related	[flat|nested] 9+ messages in thread

* Re: Stream implementation of seq-mapn
  2017-12-18 11:58 Stream implementation of seq-mapn Michael Heerdegen
@ 2017-12-18 12:32 ` Nicolas Petton
  2017-12-18 12:51   ` Michael Heerdegen
  0 siblings, 1 reply; 9+ messages in thread
From: Nicolas Petton @ 2017-12-18 12:32 UTC (permalink / raw)
  To: Michael Heerdegen, Emacs Development; +Cc: Stefan Monnier

[-- Attachment #1: Type: text/plain, Size: 1152 bytes --]

Michael Heerdegen <michael_heerdegen@web.de> writes:

Hi Michael,

> [...]
> +(cl-defmethod seq-mapn (function (stream stream) &rest streams)
> +  "Map FUNCTION over the STREAMS.

I would add to the docstring that all elements of `streams' should be
streams, otherwise it will default to the generic implementation.

> +
> +Example: this prints the first ten Fibonacci numbers:
> +
> +  (letrec ((fibs (stream-cons
> +                  1
> +                  (stream-cons
> +                   1
> +                   (seq-mapn #'+ fibs (stream-rest fibs))))))
> +    (seq-do #'print (seq-take fibs 10)))
> +
> +\(fn FUNCTION STREAMS...)"
> +  (if (not (cl-every #'streamp streams))
               ^^^^^^^^
               Since this is for stream.el, why not use `seq-every-p'?
> +      (cl-call-next-method)
> +    (cl-callf2 cons stream streams)
> +    (stream-make
> +     (unless (cl-some #'seq-empty-p streams)
                 ^^^^^^^
                 Same question :-)
> [...]


> (1) Is it ok to implement it with `cl-call-next-method' this way?

I guess it is.  Have you tried running `seq-mapn' with a mix of streams
and lists?

Cheers,
Nico

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 487 bytes --]

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: Stream implementation of seq-mapn
  2017-12-18 12:32 ` Nicolas Petton
@ 2017-12-18 12:51   ` Michael Heerdegen
  2017-12-18 13:50     ` Nicolas Petton
  0 siblings, 1 reply; 9+ messages in thread
From: Michael Heerdegen @ 2017-12-18 12:51 UTC (permalink / raw)
  To: Nicolas Petton; +Cc: Stefan Monnier, Emacs Development

Nicolas Petton <nicolas@petton.fr> writes:

> > [...]
> > +(cl-defmethod seq-mapn (function (stream stream) &rest streams)
> > +  "Map FUNCTION over the STREAMS.
> I would add to the docstring that all elements of `streams' should be
> streams, otherwise it will default to the generic implementation.

I already had "all the STREAMS must be streams", but that looked weird.


> > +  (if (not (cl-every #'streamp streams))
>                ^^^^^^^^
>                Since this is for stream.el, why not use `seq-every-p'?
> > +      (cl-call-next-method)
> > +    (cl-callf2 cons stream streams)
> > +    (stream-make
> > +     (unless (cl-some #'seq-empty-p streams)
>                  ^^^^^^^
>                  Same question :-)

Ok, done.


> > (1) Is it ok to implement it with `cl-call-next-method' this way?
>
> I guess it is.  Have you tried running `seq-mapn' with a mix of streams
> and lists?

Sure.  I does what I want: the default method is invoked.


Michael.



^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: Stream implementation of seq-mapn
  2017-12-18 12:51   ` Michael Heerdegen
@ 2017-12-18 13:50     ` Nicolas Petton
  2017-12-19 13:05       ` Michael Heerdegen
  0 siblings, 1 reply; 9+ messages in thread
From: Nicolas Petton @ 2017-12-18 13:50 UTC (permalink / raw)
  To: Michael Heerdegen; +Cc: Stefan Monnier, Emacs Development

[-- Attachment #1: Type: text/plain, Size: 134 bytes --]

Michael Heerdegen <michael_heerdegen@web.de> writes:

> Sure.  I does what I want: the default method is invoked.

That's fine.

Nico

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 487 bytes --]

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: Stream implementation of seq-mapn
  2017-12-18 13:50     ` Nicolas Petton
@ 2017-12-19 13:05       ` Michael Heerdegen
  2017-12-19 14:08         ` Nicolas Petton
  0 siblings, 1 reply; 9+ messages in thread
From: Michael Heerdegen @ 2017-12-19 13:05 UTC (permalink / raw)
  To: Nicolas Petton; +Cc: Stefan Monnier, Emacs Development

Nicolas Petton <nicolas@petton.fr> writes:

> That's fine.

Great.

While preparing the commit I noticed these two problems:

(1) Why does (seq-empty-p (stream-empty)) return nil?  That's at least a
pitfall, or a bug.  Why is there no implementation of `seq-empty-p' in
stream.el (I know stream-empty-p exists)?

(2) In stream-tests.el, there is now an undefined function
`seq-take-until'.


Thanks,

Michael.



^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: Stream implementation of seq-mapn
  2017-12-19 13:05       ` Michael Heerdegen
@ 2017-12-19 14:08         ` Nicolas Petton
  2017-12-26 22:09           ` Michael Heerdegen
  0 siblings, 1 reply; 9+ messages in thread
From: Nicolas Petton @ 2017-12-19 14:08 UTC (permalink / raw)
  To: Michael Heerdegen; +Cc: Stefan Monnier, Emacs Development

[-- Attachment #1: Type: text/plain, Size: 477 bytes --]

Michael Heerdegen <michael_heerdegen@web.de> writes:

> While preparing the commit I noticed these two problems:
>
> (1) Why does (seq-empty-p (stream-empty)) return nil?  That's at least a
> pitfall, or a bug.  Why is there no implementation of `seq-empty-p' in
> stream.el (I know stream-empty-p exists)?

That's a bug, I'll post a fix :-)

> (2) In stream-tests.el, there is now an undefined function
> `seq-take-until'.

Ok, I'll post a fix for this as well.

Thanks!
Nico

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 487 bytes --]

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: Stream implementation of seq-mapn
  2017-12-19 14:08         ` Nicolas Petton
@ 2017-12-26 22:09           ` Michael Heerdegen
  2017-12-29  9:04             ` Nicolas Petton
  0 siblings, 1 reply; 9+ messages in thread
From: Michael Heerdegen @ 2017-12-26 22:09 UTC (permalink / raw)
  To: Nicolas Petton; +Cc: Stefan Monnier, Emacs Development

[-- Attachment #1: Type: text/plain, Size: 343 bytes --]

Hello,

I uploaded a slightly tweaked version and added regression tests (but
didn't bump the package version).  I also reported the method signature
problem as "bug#29786: 27.0.50; About the argument list of methods".
Unanswered until now, so the generated help for the method still looks
weird.

For completeness, here is what I committed:


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Implement-seq-mapn-method-for-streams.patch --]
[-- Type: text/x-diff, Size: 3673 bytes --]

From 93e9009c827bc5b0f67c683a4e1748c90c6e4686 Mon Sep 17 00:00:00 2001
From: Michael Heerdegen <michael_heerdegen@web.de>
Date: Mon, 18 Dec 2017 12:49:38 +0100
Subject: [PATCH] Implement `seq-mapn' method for streams

* stream/stream.el (seq-mapn): New method.
* stream/tests/stream-tests.el (stream-seq-mapn-test): New test.
And add a `deftest-for-delayed-evaluation'.
---
 packages/stream/stream.el             | 22 ++++++++++++++++++++++
 packages/stream/tests/stream-tests.el | 15 +++++++++++++++
 2 files changed, 37 insertions(+)

diff --git a/packages/stream/stream.el b/packages/stream/stream.el
index b412807e2..3f6bc4b5b 100644
--- a/packages/stream/stream.el
+++ b/packages/stream/stream.el
@@ -321,6 +321,28 @@ applications of FUNCTION on each element of STREAM in succession."
      (cons (funcall function (stream-first stream))
            (seq-map function (stream-rest stream))))))
 
+(cl-defmethod seq-mapn (function (stream stream) &rest streams)
+  "Map FUNCTION over the STREAMS.
+
+Example: this prints the first ten Fibonacci numbers:
+
+  (letrec ((fibs (stream-cons
+                  1
+                  (stream-cons
+                   1
+                   (seq-mapn #'+ fibs (stream-rest fibs))))))
+    (seq-do #'print (seq-take fibs 10)))
+
+\(fn FUNCTION STREAMS...)"
+  (if (not (seq-every-p #'streamp streams))
+      (cl-call-next-method)
+    (cl-labels ((do-mapn (f streams)
+                         (stream-make
+                          (unless (seq-some #'stream-empty-p streams)
+                            (cons (apply f (mapcar #'stream-first streams))
+                                  (do-mapn f (mapcar #'stream-rest streams)))))))
+      (do-mapn function (cons stream streams)))))
+
 (cl-defmethod seq-do (function (stream stream))
   "Evaluate FUNCTION for each element of STREAM eagerly, and return nil.
 
diff --git a/packages/stream/tests/stream-tests.el b/packages/stream/tests/stream-tests.el
index decf3ad47..896c72993 100644
--- a/packages/stream/tests/stream-tests.el
+++ b/packages/stream/tests/stream-tests.el
@@ -166,6 +166,18 @@
   (should (= -1 (stream-first (seq-map #'- (stream-range 1)))))
   (should (= -2 (stream-first (stream-rest (seq-map #'- (stream-range 1)))))))
 
+(ert-deftest stream-seq-mapn-test ()
+  (should (streamp (seq-mapn #'+ (stream (list 1 2 3)) (stream (list 4 5 6)))))
+  (should (not (streamp (seq-mapn #'+ (stream (list 1 2 3)) (stream (list 4 5 6)) (list 7 8 9)))))
+  (should (= 2 (seq-length (seq-mapn #'+ (stream (list 1 2 3)) (stream (list 4 5))))))
+  (should (equal (letrec ((fibs (stream-cons
+                                 1
+                                 (stream-cons
+                                  1
+                                  (seq-mapn #'+ fibs (stream-rest fibs))))))
+                   (seq-into (seq-take fibs 10) 'list))
+                 '(1 1 2 3 5 8 13 21 34 55))))
+
 (ert-deftest stream-seq-do-test ()
   (let ((result '()))
     (seq-do
@@ -292,6 +304,9 @@
 (deftest-for-delayed-evaluation (seq-drop (make-delayed-test-stream) 2))
 (deftest-for-delayed-evaluation (seq-take-while #'numberp (make-delayed-test-stream)))
 (deftest-for-delayed-evaluation (seq-map #'identity (make-delayed-test-stream)))
+(deftest-for-delayed-evaluation (seq-mapn #'cons
+                                          (make-delayed-test-stream)
+                                          (make-delayed-test-stream)))
 (deftest-for-delayed-evaluation (seq-filter #'cl-evenp (make-delayed-test-stream)))
 (deftest-for-delayed-evaluation (stream-delay (make-delayed-test-stream)))
 (deftest-for-delayed-evaluation (seq-copy (make-delayed-test-stream)))
-- 
2.15.1


[-- Attachment #3: Type: text/plain, Size: 20 bytes --]



Thanks,

Michael.

^ permalink raw reply related	[flat|nested] 9+ messages in thread

* Re: Stream implementation of seq-mapn
  2017-12-26 22:09           ` Michael Heerdegen
@ 2017-12-29  9:04             ` Nicolas Petton
  2017-12-30 14:28               ` Michael Heerdegen
  0 siblings, 1 reply; 9+ messages in thread
From: Nicolas Petton @ 2017-12-29  9:04 UTC (permalink / raw)
  To: Michael Heerdegen; +Cc: Stefan Monnier, Emacs Development

[-- Attachment #1: Type: text/plain, Size: 570 bytes --]

Michael Heerdegen <michael_heerdegen@web.de> writes:

> Hello,

Hi Michael,

> I uploaded a slightly tweaked version and added regression tests (but
> didn't bump the package version).  I also reported the method signature
> problem as "bug#29786: 27.0.50; About the argument list of methods".
> Unanswered until now, so the generated help for the method still looks
> weird.
>
> For completeness, here is what I committed:

Your additions have improved thunk.el/stream.el quite a lot, thank
you!

I'd be curious to see your use cases for these libraries.

Cheers,
Nico

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 487 bytes --]

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: Stream implementation of seq-mapn
  2017-12-29  9:04             ` Nicolas Petton
@ 2017-12-30 14:28               ` Michael Heerdegen
  0 siblings, 0 replies; 9+ messages in thread
From: Michael Heerdegen @ 2017-12-30 14:28 UTC (permalink / raw)
  To: Nicolas Petton; +Cc: Stefan Monnier, Emacs Development

Nicolas Petton <nicolas@petton.fr> writes:

> I'd be curious to see your use cases for these libraries.

Currently, I mainly use them in el-search.  I also have used streams to
implement "helm-browse' which is something like "swiper" but uses
streams in the search engine to make it fast and extensible.  It's quite
cool, but I haven't published it so far, mainly because I don't use Helm
that much in the meantime, and swiper itself is quite nice.

I use streams here and there in my init files instead of lists where it
makes sense, e.g. for traversing directory hierarchies.

So, not too much cool stuff...


Michael.



^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~2017-12-30 14:28 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2017-12-18 11:58 Stream implementation of seq-mapn Michael Heerdegen
2017-12-18 12:32 ` Nicolas Petton
2017-12-18 12:51   ` Michael Heerdegen
2017-12-18 13:50     ` Nicolas Petton
2017-12-19 13:05       ` Michael Heerdegen
2017-12-19 14:08         ` Nicolas Petton
2017-12-26 22:09           ` Michael Heerdegen
2017-12-29  9:04             ` Nicolas Petton
2017-12-30 14:28               ` Michael Heerdegen

Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).