Straight to the point of your question : it's an exception StackTrace, :-)
Now if you want some more question to be answered, you may well be a little
bit more explicit, as Berlin suggested too ! ;-)
--
Laurent
2009/2/21 Feng
>
> Exception java.lang.StackOverflowError:
> [1] clojure.lang.Pe
I have a horrible hack to do this, which uses even more appalling code
than [2] ripped off from a different forum, but which (in my limited
experience) seems to work OK. I'll email it to you privately, and to
anyone else who wants to use it (just ask).
-Jason
On Feb 20, 9:33 pm, Brian Carper w
One could argue that wildcard imports in Java (import package.*) are
evil, pollute your namespaces, create potential naming conflicts,
etc. One would probably be correct.
One could also argue that having to manually type a list of dozens of
classnames is pretty tedious, especially if all you wan
On Feb 20, 11:08 pm, Feng wrote:
> Exception java.lang.StackOverflowError:
> [1] clojure.lang.PersistentHashMap$LeafNode.nodeSeq
> (PersistentHashMap.java:567), pc = 2
> [2] clojure.lang.PersistentHashMap$BitmapIndexedNode$Seq.create
> (PersistentHashMap.java:503), pc = 21
> [3] clojure.l
Exception java.lang.StackOverflowError:
[1] clojure.lang.PersistentHashMap$LeafNode.nodeSeq
(PersistentHashMap.java:567), pc = 2
[2] clojure.lang.PersistentHashMap$BitmapIndexedNode$Seq.create
(PersistentHashMap.java:503), pc = 21
[3] clojure.lang.PersistentHashMap$BitmapIndexedNode.nodeSeq
Taking Chouser's suggestion of using a proxy that implements the
IDeref interface, and overrides deref to provide access to a closed-
over atom, I rewrote generic functions without resorting to gen-class
or external Java files. That makes me happy; good suggestion, Chouser.
I made a couple of oth
To me, the most incredible thing about Clojure is that this all
happened in about a year! Choosing to be hosted on an established
platform, though sometimes criticized by some people, was a very
effective way to get people to start writing useful programs quickly
with libraries they were used to.
That looks about right. I proposed adding this to clojure.contrib as
"lazy-get":
http://code.google.com/p/clojure-contrib/issues/detail?id=10
along with a corresponding "safe-get" function that errors if the key
is not found.
-Jason
On Feb 20, 4:42 pm, Rowdy Rednose wrote:
> What is the eas
What is the easiest way to make clojure's get function evaluate the
value for 'not-found' only if the key is not found? Do I have to write
a macro like
(defmacro my-get [map key not-found]
`(if (contains? ~map ~key)
(get ~map ~key)
~not-found))
--~--~-~--~~~--
"Stephen C. Gilardi" writes:
> I see that issue 79 against Clojure requesting this change is still
> open. If auto-wrapping of readers is no longer a change you'd like to
> see, would you please either withdraw it (if you as the originator
> have that capability) or add a comment to it requestin
On Feb 12, 2009, at 11:00 PM, Phil Hagelberg wrote:
Yikes. I hadn't thought of that--I was only using read in cases
where I
wanted to get a single object from the reader.
Taking things off the reader and then pushing them back on seems a
little odd to me--isn't there a way to just "peek" at
I filed it as an issue a few days ago, until then I wrote my own read-
line from the old read-line.
On Feb 20, 4:32 pm, "Stephen C. Gilardi" wrote:
> On Feb 20, 2009, at 12:58 PM, Perry Trolard wrote:
>
> > Hope I didn't imply by the above that I was suggesting a name change.
>
> > I know my pro
On Feb 20, 10:29 am, Laurent PETIT wrote:
> Hello Rich, thanks for answering.
>
> You may not have followed the following of the thread, where I precised my
> mind by saying what I really wanted to do (but did not at first) was using
> with-local-vars, since the value to be mutated is confined
On Feb 20, 2009, at 12:58 PM, Perry Trolard wrote:
Hope I didn't imply by the above that I was suggesting a name change.
I know my proposed version is ugly, but since BufferedReader &
LineNumberingPR can't be unified, special-case-ing for the latter
strikes me as an acceptable fix. But removin
Thanks for explanation, all!
Frantisek
On 20 Ún, 20:57, Jason Wolfe wrote:
> It probably does an "identical?" call on a pair before calling
> compare, for efficiency. In other words, it may "work" on non-
> comparable types, but only when passed n instances of the exact same
> object:
>
> user
For me, Clojure made programming exhilarating again.
Thank you, Rich and everyone else for making this happen.
On Feb 20, 9:59 pm, Rich Hickey wrote:
> There have been many new additions to the contributors list:
>
> http://clojure.org/contributing
>
> and many new donations:
>
> https://sourc
On Feb 20, 2:18 pm, Jason Wolfe wrote:
> Hmmm, that is weird ... I would expect the sort to throw an exception,
> but it seems to complete happily. (If you try it with a vector
> instead of a list, it will error).
>
Fixed in svn 1298 - thanks for the report.
Rich
--~--~-~--~~---
There have been many new additions to the contributors list:
http://clojure.org/contributing
and many new donations:
https://sourceforge.net/project/project_donations.php?group_id=137961
We're now at 1500+ members on the group, and still growing, all while
maintaining a friendly, helpful atmos
It probably does an "identical?" call on a pair before calling
compare, for efficiency. In other words, it may "work" on non-
comparable types, but only when passed n instances of the exact same
object:
user> (sorted-set '(1) '(1))
; Exception
user> (let [x '(1)] (sorted-set x x))
#{(1)}
This
It looks that it is more complicated than that:
user=> (sorted-set () ())
#{()}
user=> (sorted-set {} {})
#{{}}
user=> (sorted-set #{} #{})
#{#{}}
Frantisek
On 20 Ún, 20:33, Vincent Foley wrote:
> I'm pretty sure that sorted-set works only with values that are
> instances of a class that implem
I'm pretty sure that sorted-set works only with values that are
instances of a class that implements Comparable.
user=> (instance? Comparable [])
true
user=> (instance? Comparable {})
false
user=> (instance? Comparable ())
false
user=>
On Feb 20, 2:21 pm, Frantisek Sodomka wrote:
> sorted-set
On Fri, Feb 20, 2009 at 1:59 PM, Steffen Glückselig wrote:
>
> I was trying to use Clojure to verify the behavior some method in
> Java.
>
> Namely I wanted to quickly check whether Collections.sort does the
> sorting I need.
>
> So I came up with:
>
> (let [list '("1" "KB" "K6" "2" "EÜ" "EZ" "ES
sorted-set works for vectors, but doesn't work for lists, maps and
sets:
user=> (sorted-set [1 4])
#{[1 4]}
user=> (sorted-set [1 4] [1 4])
#{[1 4]}
user=> (sorted-set [4 1] [1 4])
#{[1 4] [4 1]}
user=> (sorted-set '(1 4))
#{(1 4)}
user=> (sorted-set '(1 4) '(1 4))
java.lang.ClassCastException: c
The Clojure collections are immutable, which is their entire reason for
existing. However, there is nothing stopping you from creating a plain old
Java collection in Clojure.
(let [list (ArrayList. '("Fred" "mary" "sue")]
(do
(java.util.Collections/sort list)
list))
Should work.
On Fri
Hmmm, that is weird ... I would expect the sort to throw an exception,
but it seems to complete happily. (If you try it with a vector
instead of a list, it will error).
Anyway, the typical way to do this would be to convert your Clojure
data structure to a mutable Java type first, then convert i
I was trying to use Clojure to verify the behavior some method in
Java.
Namely I wanted to quickly check whether Collections.sort does the
sorting I need.
So I came up with:
(let [list '("1" "KB" "K6" "2" "EÜ" "EZ" "ES")]
(do
(java.util.Collections/sort list)
list))
But since Collect
That is correct. There is no reason that you need to compile the clj
files before running them. Just include the clojure.jar package and
everything will work fine.
On Feb 20, 6:41 am, Jeffrey Straszheim
wrote:
> If your CLJ files are in the classpath, and you include clojure.jar, then
> your g
I haven't read _The Reasoned Schemer_, but I have read the logic programming
sections of SICP and Norvig .. as well as some non-lispy texts. I *suspect*
TRS's material is at a similar weight. What I'm trying to do now goes quite
a bit beyond those implementations. Plus, they are all use a top-do
There's a logic programming module in the files section. It
implements the system found in "The Reasoned Schemer".
On Feb 18, 2:59 pm, Jeffrey Straszheim
wrote:
> Did you cover logic programming? Any bottom up logic query techniques?
> (My motives are probably transparent.)
>
> On Wed, Feb 18,
On Feb 20, 10:35 am, Tom Ayerst wrote:
> You probably don't want to be doing this. Your function looks like it could
> use a lazy sequency and a filter.
>
> e.g. (doseq [e (filter odd? [1 2 3 4 5 6 7])] (prn e))
>
> You can get a long way with partition, for, filter and reduce; It is a pain
>
Hope I didn't imply by the above that I was suggesting a name change.
I know my proposed version is ugly, but since BufferedReader &
LineNumberingPR can't be unified, special-case-ing for the latter
strikes me as an acceptable fix. But removing the type hint so that
the function works out of the
On 20 Feb, 17:10, Jeffrey Straszheim
wrote:
> There have been some major changes in the last week or so.
> Seehttp://clojure.org/lazyfor a brief overview.
> Also:http://blog.n01se.net/?p=39
>
> On Fri, Feb 20, 2009 at 10:31 AM, Rock wrote:
>
> > After watching Rich's video presentations and
Based on the time frame, I'll look into doing the items Chouser
submitted. If I finish early, I may look into the Hindley Milner
optimization.
Joshua
On Feb 20, 11:51 am, chris wrote:
> Would you perhaps consider working on integrating some form of type
> inference into clojure to cut down on t
On Thu, Feb 19, 2009 at 4:51 PM, mikel wrote:
>
> If I can implement a custom printing function that gets called on
> generic functions, that presupposes that I have some way to tell that
> an object is a generic function. So, yes, if I solve the problem, then
> the problem is solved. :-)
Hm, per
Would you perhaps consider working on integrating some form of type
inference into clojure to cut down on the number of required type
declarations?
We know that clojure runs *much* faster when it isn't doing
reflection. It would be nice if I could annotate one function and
have the annotation au
You could also hide the vector inside your list model code, and require
modification through an api. This would let you know what updates to send
to the Swing object without having to walk the list.
On Fri, Feb 20, 2009 at 11:11 AM, Rowdy Rednose wrote:
>
> Thanks pmf and Jeffrey,
>
> so I guess
Thanks pmf and Jeffrey,
so I guess the idea is to have the TableModel hold on to the
(dereferenced) vector and when the agent fires, the model would
dereference the ref again and compare that value with the old one. It
would then have to do a diff on those 2.
That is one of the unelegant ideas I
There have been some major changes in the last week or so. See
http://clojure.org/lazy for a brief overview.
Also: http://blog.n01se.net/?p=39
On Fri, Feb 20, 2009 at 10:31 AM, Rock wrote:
>
> After watching Rich's video presentations and reading Stuart's fine
> book, I was absolutely convinced
After watching Rich's video presentations and reading Stuart's fine
book, I was absolutely convinced that if you take the *rest* of an
empty list, what you've got left is nil.
Instead, with the recent svn version of Clojure, I'm getting the empty
list (actually sequence of course) again:
user=>
As always, thanks for the quick fix! - Jeff
On Feb 20, 7:11 am, Rich Hickey wrote:
> On Feb 19, 8:18 pm, Jeffrey Chu wrote:
>
>
>
> > Hi,
>
> > Okay, I'm reasonably sure this is a bug with clojure's eval. Here's an
> > even more succinct version:
>
> > (defn lazy-identity [a]
> > (if (seq? a)
You probably don't want to be doing this. Your function looks like it could
use a lazy sequency and a filter.
e.g. (doseq [e (filter odd? [1 2 3 4 5 6 7])] (prn e))
You can get a long way with partition, for, filter and reduce; It is a pain
to get your head around it at first if your are not us
On Feb 20, 10:12 am, BerlinBrown wrote:
> This is a general termingology question. What is this idiom, called
> where you pass a function as an argument to another function and then
> use that function with in a loop.
I think this is sometimes called "inversion of control."
-Stuart Sierra
--~-
Hello Rich, thanks for answering.
You may not have followed the following of the thread, where I precised my
mind by saying what I really wanted to do (but did not at first) was using
with-local-vars, since the value to be mutated is confined in the function,
and the macro ensures (via the use of
Higher-order functions (HOFs), are functions that use "lower"
functions to perform a task. The map and reduce functions are both
HOFs.
On Feb 20, 2009, at 10:16 AM, Jeffrey Straszheim wrote:
> The OO folks call this an "internal iterator" or "visitor".
> However, I'd recommend against ad
Hi,
I'm new to clojure and lisp in general. I'm trying semi-porting a real
world application but at this time I lack patterns to reason about
clojure solutions to problems. I thought I'd ask the community about
one such problem I’m facing.
To give a simple example for the following discussion, s
On Feb 20, 2009, at 5:33 AM, Christophe Grand wrote:
>
> Laurent PETIT a écrit :
>> If I'm damn sure that the value will be set once from within the same
>> thread (here we are in the SWT UI Thread), is there a reason to
>> prefer
>> atoms ?
>> (This is a real question, not a disguised affirma
The OO folks call this an "internal iterator" or "visitor". However, I'd
recommend against adopting their point of view.
On Fri, Feb 20, 2009 at 10:12 AM, BerlinBrown wrote:
>
> This is a general termingology question. What is this idiom, called
> where you pass a function as an argument to ano
This is a general termingology question. What is this idiom, called
where you pass a function as an argument to another function and then
use that function with in a loop. I thought it reminded me of that
aspect oriented programming? cross cutting of concerns?
For example, I do that a lot, whe
On Feb 19, 8:18 pm, Jeffrey Chu wrote:
> Hi,
>
> Okay, I'm reasonably sure this is a bug with clojure's eval. Here's an
> even more succinct version:
>
> (defn lazy-identity [a]
> (if (seq? a)
> (map lazy-identity a)
> a))
>
> user=> (lazy-identity '(apply + '(1 2 3)))
> (apply + (quo
If your CLJ files are in the classpath, and you include clojure.jar, then
your good.
On Fri, Feb 20, 2009 at 9:30 AM, rob wrote:
>
> What do you mean when you say there is no need to compile your program
> to distribute it? Doesn't that require end users to set up a clojure
> environment? And
The tip on compilation is really useful. Thanks!
On Feb 16, 11:08 am, levand wrote:
> If you have a Clojure namespace that uses gen-class, and if there is a
> method called '-main' within that namespace, then the resultant
> *.class file is equivalent to a Java class with a 'main' method, and
It's true, responding to clicks should not necessarily require
mutation, but probably anything beyond that GUI-wise would require
it. So I guess it's a minor quibble that one or two of the basic
aspects are not as elegant as they could be (thanks to Java). But
I'll file this under "worse is bett
What do you mean when you say there is no need to compile your program
to distribute it? Doesn't that require end users to set up a clojure
environment? And how would you deploy a web-based application without
compiling it?
On Feb 19, 6:51 pm, Kevin Albrecht wrote:
> I can vouch for using SWT
As far as I can tell, futures are *not* agents, but wrap the
java.util.concurrent.Future class. However, they do run in the same thread
pool that the agents use, so your point still stands.
On Fri, Feb 20, 2009 at 9:09 AM, Mark Volkmann wrote:
>
> The doc string for the future function should p
It would be pretty easy to wrap an agent (as pmf suggests) to notify your
model class if a Vector changes. You could then do something like
(map = old_vec new_vec)
And then look for false results in the array and send a notification to
Swing.
On Fri, Feb 20, 2009 at 5:40 AM, Rowdy Rednose wrote
The doc string for the future function should probably mention that it
uses an Agent so shutdown-agents should be called at the end of
applications that use it.
--
R. Mark Volkmann
Object Computing, Inc.
--~--~-~--~~~---~--~~
You received this message because you
Ok,here's a small example that propagates changes to a ref's vector to
a watcher:
;; define your model
(def model (ref ["abc" "def" "ghi"]))
;; define a watcher
(def model-watcher (agent nil))
;; connect your model to the watcher
(add-watcher model :send model-watcher (fn [state source] (printl
One last question about clojure and derby: once I store a character
large object type (CLOB), does anyone know of a best-practices way of
converting it back into string from a resultset-seq for use in clojure
processing? Thanks in advance for any help in this area.
Brian
--~--~-~--~~
On Feb 20, 11:40 am, Rowdy Rednose wrote:
> Any elegant ideas or examples on how to do this when the underlying
> data structure is (a ref to) one of clojure's (immutable) collections,
> so that a change to that structure will fire the appropriate event?
You can use add-watcher to notify state c
Nice. It starts to work flawless. Did another installation test on a
fresh ubuntu.
1. mark emacs-snapshot and ant for installation in synaptic. You'll
automatically download the other files you need then.
2. get clojure-mode from http://github.com/technomancy/clojure-mode/ ,
and store the .el fil
2009/2/20 Christophe Grand
>
> Laurent PETIT a écrit :
> > If I'm damn sure that the value will be set once from within the same
> > thread (here we are in the SWT UI Thread), is there a reason to prefer
> > atoms ?
> > (This is a real question, not a disguised affirmation that I'm doing
> > the
All the clojure swing examples I've seen so far use JTables in a
static way, i.e. the data is not programmatically modified once the
table got created.
Has anyone actually tried to implement a TableModel that is backed by
a clojure Vector/Map and fires events to TableModelListeners when the
under
Laurent PETIT a écrit :
> If I'm damn sure that the value will be set once from within the same
> thread (here we are in the SWT UI Thread), is there a reason to prefer
> atoms ?
> (This is a real question, not a disguised affirmation that I'm doing
> the right thing, so please arguments welcom
Well, I had not particularly optimization concerns in mind.
I was just more thinking about what could be the right thing to do.
I feel that atoms are useful to allow atomic operations on a value from
several threads.
If I'm damn sure that the value will be set once from within the same thread
(here
Laurent PETIT a écrit :
> I haven't tried with something simpler that the array-like method, I
> suspect it won't work in clojure, too. And I didn't want to use
> clojure's mutable methods because :
> - we already are in an IO operation, the mutation is local to the
> operation
> - I don't want
65 matches
Mail list logo
