iterate-devel

iterate-devel@common-lisp.net

271 discussions

Re: [iterate-devel] Iterate download error
by Hoehle, Joerg-Cyril 05 Jan '06

05 Jan '06

Marc Battyani wrote: >Oops sorry. I really thought it was up to date. I will change it. Thanks. >I have an email from you dated 20 december 2005. Is that the >latest one? That day, I just forwarded the June version. >Also I added some patches sent by Fare in the cl-pdf version. They are not in what I sent you (bug was not known in June). Feel free to change the one delete -> (remove nil temps) It's clearly a bug. During holidays I revisited the two other changes. In remove-clause: (delete entry index) -> (setq index (delete entry index)) I decided *not* to leave that in the code (but add a comment) because (setq index ...) is a noop: index is local variable, and the function is just about to exit, so nobody cares about it's value. The code depends on non-guaranted behaviour of DELETE, basically that it implements the naive / canonical operation: RPLACD the previous CDR. Adding (setq index ...) would gain nothing in a Lisp which would provide another implementation of DELETE. E.g. (setf (fdefinition 'delete) #'remove) -- REMOVE is AFAIK a legal implementation of DELETE. Remove-clause would break. As I have not yet seen such other implementation, I've not bothered to write MY-DELETE to guarantee the behaviour needed by Iterate, or implement a different approach. If somebody yells that's not acceptable, please do so. I might change my opinion. Faré, since you reported that it's sbcl who pointed you to delete in Iterate, I'd like to know which version you used. I could not reproduce a warning with sbcl-0.9.2 (as found in Ubuntu/Breezy). Regards, Jörg Höhle.

2 1

[iterate-devel] Iterate download error
by Surendra Singhi 05 Jan '06

05 Jan '06

The following message is a courtesy copy of an article that has been posted to gmane.lisp.web.uncommonweb.devel as well. Hi, The download link for iterate both on the home page and cliki gives error and doesn't work. http://common-lisp.net/project/iterate/releases/iterate-current.tar.gz Thanks. -- Surendra Singhi http://www.public.asu.edu/~sksinghi/index.html ,---- | WHY SHOULD WE SAVE TIGER? | Ans: Saving the tiger means saving mankind.. | | Help http://pudang.tripod.com/ | or https://secure.worldwildlife.org/forms/tiger_appeal_1.cfm `----

2 1

Re: [iterate-devel] Iterate download error
by Hoehle, Joerg-Cyril 05 Jan '06

05 Jan '06

Marc Battyani wrote: >This is for this kind of problem that I keep an up to date >ITERATE version in cl-pdf. ;-) I'm sorry, Marc, it is not up to date :-( I recently looked at your online cl-pdf repository only to discover that it contains the Iterate version as of February/March 2005. I'm sure I sent you the June 2005 version half a year ago. readme-pdf says: "This version is known to work well with cl-pdf and cl-typesetting". This can also be read as "don't trust the current or more recent versions" or "I haven't tested anything more recent so I can't tell nor make guarantees". I don't like the taste of that phrase, but I can fully understand that you provide a full package with all dependencies instead of having people fetch a half-dozen utility packages themselves. Hopefully, a Debian distribution will split things up and permit independent evolution of packages. I bet that my most current versions works equally well, if not better. I believe you just didn't get to update your copy of iterate in cl-pdf, because cl-pdf is essentially stable and you've been working on other things since then. >Interested people can get it from there until the official >site is OK again. I cannot recommend that until you update to what I sent you last June. I'll check whether the mailing list archives (or allows) attachments. Then I should send updates to the list, instead of to a few individual persons like I did back on 23rd of June 2005 (to Marco Baringer, you, Peter van Eynde, and CC: Kevin Rosenberg and Mark Williams). Of course, the best solution would be for me to get write access to common-lisp.net. I'll see how far I can get there. BTW, I'm slowly preparing an update to the June version - don't hold your breath! Interested people can get an interim version for compatibility testing from me (very little changes so far, I mostly increased testing coverage). Regards, Jörg

2 1

Re: [iterate-devel] Iterate download error
by Hoehle, Joerg-Cyril 05 Jan '06

05 Jan '06

Surendra Singhi wrote: > The download link for iterate both on the home page and cliki >gives error and doesn't work. >http://common-lisp.net/project/iterate/releases/iterate-current.tar.gz I suppose this is related to the bug reported by Rune Nesheim on 29th of December and analysed by Andreas Fuchs: please wait for Marco Baringer to come back from holidays (I suppose) and change some file permissions on the UNIX server. I'm sorry for the inconvenience. Regards, Jorg Hohle.

2 1

[iterate-devel] 403 FORBIDDEN
by Rune Nesheim 26 Dec '05

26 Dec '05

http://common-lisp.net/project/iterate/releases/iterate_1.4.1.tar.gz 403 FORBIDDEN, this also means asdf:install is broken for iterate.

1 0

Re: [iterate-devel]/[series-users] collect-tail ?
by Hoehle, Joerg-Cyril 20 Dec '05

20 Dec '05

Chris Dean wrote: >[1] The anatomy of a loop: a story of scope and control. > Olin Shivers. > http://www.cc.gatech.edu/~shivers/citations.html#loop > >[2] Eager Comprehensions in Scheme: The design of SRFI-42 > Sebastian Egner > http://www.deinprogramm.de/scheme-2005/program.html Thank you very much for these pointers. I'm very happy to see research in iteration in a way that is potentially useable for CL (unlike iteration constructs in many other languages). I'm still looking for the ideal construct myself. :-) >This is my standard looping question in any language: Read the last N >lines of a file, transform them, filter them, and return the unique >elements. Not trivial, even with SRFI-42. It's interesting how Iterate and SRFI-42 appear somewhat inside-out: one has COLLECT buried somewhere inside the body, the other right upfront (list-ec). OTOH, one has the form to be collected buried deep inside, the other right next to COLLECT. Good reading. Regards, Jorg Hohle

2 1

Re: [iterate-devel]/[series-users] collect-tail ?
by Hoehle, Joerg-Cyril 19 Dec '05

19 Dec '05

Chris, since you're asking similar questions in series-users and iterate-devel, I'm starting to wonder whether you're going to write a paper on comparisons of some looping frameworks? :-) >how do I go about collecting the last N elements of anything? > (iter (for line :in-file my-file :using #'read-line) > (collect-tail line :how-many 10)) If 10 is known at macro-expansion time, you can always write a macro that expands to a series of 10 uses of PREVIOUS in Iterate, or 10 consecutive latches in Series. I.e., IMHO there are always two ways, either with series of iterate: a) Either take a code generation approach and create code that latches (= delay by one) N times, b) Or take a functional approach, implement a ring datastructure and use that. The former looks like a better use of Series: it could be a series of the 10 last values (as a list, or 10 series of values farther and farther behind). In/After the last iteration, you'll have the 10 last values. As to which approach is better... Probably you wouldn't want to generate code for the last 100 values. BTW, you'll have to defined what happens when less than 10 iterations were taken. Straight-forward code is likely to return (val1 val2 nil nil ...6xnil), whereas one could wish for (val1 val2) with only 2 iterations. Regards, Jorg Hohle.

2 1

[iterate-devel] collect-tail ?
by Chris Dean 15 Dec '05

15 Dec '05

How would I go about reading the last 10 lines in a files? Or more generally, how do I go about collecting the last N elements of anything? I'm looking for advice on the "right way" to do this using iterate. On way would be to write collect-tail that just kept the last N elements: (iter (for line :in-file my-file :using #'read-line) (collect-tail line :how-many 10)) Is this a good way? Regards, Chris Dean

1 0

[iterate-devel] Between-each, FORMAT's ~^ and code-walking
by Hoehle, Joerg-Cyril 09 Dec '05

09 Dec '05

Mészáros Levente wrote: >> Would you like to show more elaborate loop examples? >Let's say one has a list and one wants to transform it into a >string by transforming each element to a string and putting a >comma between them. >(concatenate-as-string > (iter (for e in l) > (except-first-time (collect ", ")) > (collect (as-string e)))) I recently came across that article about Lispy FORMAT replacements and started wondering how to nicely implement FORMAT's "~^" idiom. http://www.cs.yale.edu/homes/dvm/format-stinks.html BTW, I haven't seen the code for McDermott's out macro, but I expect it to be subject to code-walking problems similar to Iterate, series and other packages that depend on traversing code. Why? consider one example from the above article: (out "#<Foo " (Foo-label f1) (:q ((eq (Foo-status f1) ':abnormal) "?")) (:e (dolist (x (Foo-contents f1)) (:o " " x))) ">") Do you see the :o keyword inside the dolist? I think keywords are constants and cannot be locally bound (using MACROLET). Therefore, I believe the OUT macro needs to code-walk its body to track down all uses of :o and insert the replacememt code. Had a non-constant symbol been used therein, then MACROLET could have been used and the implementation of the macro out would (presumably) be clean. In any case, ~^ is quite different from *-first-time. It's more like an early test of loop termination. The macro out does not propose an idiom for that. The examples just repeat the termination test, as in (defun print-xapping ...) ... (do ((d ... (cdr d))) ... (:q ((not (null (cdr d))) " "))))) Repeating the test makes it unable to use the much more readable DOLIST instead of DO, and therefore must also use (CAR d) to access the elements. Reading the full paper (ytools) instead of the short FORMAT critique, it appears to me that OUT is no silver bullet either. It has its own set of controls: :a, :e, :t, :i>, :i<, :f, :v, :pp-block etc., so it's still complex to grok. I found the difference more appealing between e.g. the sexp-based regular expression notation (IIRC there's a Scheme SRFI about it) that the error-prone flat string syntax. BTW, McDermott's diver&snorkeler macro design pattern is interesting. Regards, Jörg Höhle

2 1

Re: [iterate-devel] Between-each
by Hoehle, Joerg-Cyril 02 Dec '05

02 Dec '05

Hi, Mészáros Levente wrote: >I may be wrong about this, but I could not find support for a >quite commonly used loop pattern, namely when you want >something to be evaluated between each loop cycles. Please accept my apology for responding that late. I felt a little uncomfortable (see below) and did not quite know how to reply. >(def-special-clause BETWEEN-EACH (between) > (return-code :body (list (if-1st-time nil (list (walk-expr between)))))) Correct code (like implementation of IF-FIRST-TIME). >Maybe I'm wrong, could somebody tell how to do this with >iterate? If there's no support for this perhaps you could put >something like this into the main branch. You could use a macro-substitution and expand to the exported and documented IF-FIRST-TIME and DEFMACRO-CLAUSE, instead of using package internals like return-code. Your implementation shows that there's the same relationship between IF-FIRST-TIME and your BETWEEN-EACH like there's between WHEN and UNLESS. So why not add it, even it would be redundant to purists? However, given that analogy, I find the given name quite misleading, as it's very different from AFTER-EACH. >And it works like: >(iter (for i from 1 to 3) (between-each (collect (+ i 100))) > (collect i)) >(1 102 2 103 3) Hmm, I feel uncomfortable with the name, given the functionality you show. "Between" is something I'd expect to happen after one iteration, while you use it within the body. I.e., my wording shows I'm thinking in terms of AFTER-EACH. Well, there's AFTER-EACH, which is subject to code movement, but wouldn't it work for you? Another question is: how does this extension behave when used outside the top-level of the loop. Does the given name adequately reflect the functionality? E.g. (iter (for i...) (if (oddp i) (between-each (collect i)))) What would a reader expect as value from such a form? What about BUT-FIRST[-TIME] as a name? Would you like to show more elaborate loop examples? Regards, Jörg Höhle.

1 0

← Newer
1
...
18
19
20
21
22
23
24
...
28
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

iterate-devel