Arc Forumnew | comments | leaders | submitlogin
12 points by zachbeane 4155 days ago | link | parent

Arc's version of CL's (incf (gethash k table 0)) is gross.


5 points by randallsquared 4155 days ago | link

Seems like it wouldn't be hard to add defaults to the hash referencing (and string referencing), to allow (++ (table k 0)) which seems much nicer than the CL version, in my opinion.

Edit: Actually, not string and array referencing, since defaults don't matter for that, and there's a much more useful meaning for that: slices.

-----

6 points by pg 4155 days ago | link

I might do it that way, but it seems cleaner to do it when hash tables are created.

-----

5 points by randallsquared 4155 days ago | link

Only if every key is a similar type. The default that's most appropriate is often peculiar to how you're using the value or location at the use site.

-----

2 points by lg 4155 days ago | link

speaking of hash tables, I remember in ACL you explained why CL has two return values for gethash, to differentiate between the nil meaning "X is stored as nil in my table" and the nil meaning "X is not stored in my table". So why not in Arc?

-----

2 points by pg 4154 days ago | link

Because it turns out that in practice it's rarely to never an issue. If you have nil as a val in a hash table, you usually don't care whether that's because it was never set, or explicitly set to nil.

-----

1 point by dr_drake 4150 days ago | link

Dear Paul, I can not believe you would make such a statement. Either you're living in a vastly different programming universe than the one I am living in, or you really haven't done that much programming at all. In any case, there are many situations where one stores types of values that may include nil in a hash table, and in most of these there is a very significant difference between 'value is nil' and 'value is not stored'. I understand that Arc isn't trying to all 'enterprisey', but these are fundamental concepts that, I thought, only complete amateurs did not understand. Sincerely, Dr. Drake

-----

2 points by pg 4150 days ago | link

You know, I do actually understand the difference between the two cases. What I'm saying is that in my experience hash tables that actually need to contain nil as a value are many times less common than those that don't.

In situations where the values you're storing might be nil, you just enclose all the values in lists.

My goal in Arc is to have elegant solutions for the cases that actually happen, at the expense of elegance in solutions for rare edge cases.

-----

1 point by mschw 4151 days ago | link

Python's defaultdict takes a factory function at construction time.

http://docs.python.org/lib/defaultdict-objects.html

-----

12 points by pg 4155 days ago | link

That's true. Fixing that is one of the top priorities.

-----

3 points by ryantmulligan 4155 days ago | link

Could you please elaborate on what Arc's version is? Personally I don't even understand what your CL code is doing, being a Lisp Newb.

-----

5 points by jimbokun 4155 days ago | link

(incf (gethash k table 0))

table is a hash table, k is a key, gethash returns the value in table for k or 0 if no value for k is found. Think of incf as ++. It will increment the value by 1 and set that as the value for k in table.

-----

7 points by pg 4155 days ago | link

And the problem with Arc is that currently the default value for an entry in a hash table is nil, rather than zero. If h is a hash table and you know (h 'foo) is 1, you can safely say

  (++ (h 'foo))  
But if you don't know whether (h 'foo) has a value yet you have to check explicitly:

  (= (h 'foo) (+ 1 (or (h 'foo) 0)))

-----

1 point by metageek 4155 days ago | link

How about if <code>h</code> takes an optional second argument, which is the default, and the macros are smart enough that you can do <code>(++ (h 'foo 0))</code>?

-----

1 point by greatness 4151 days ago | link

I agree, this is probably the best solution.

-----

1 point by reitzensteinm 4155 days ago | link

Perhaps (++ containsnil) should result in 1 anyway? Is there any case where that would break anything?

-----

2 points by simonb 4155 days ago | link

For one it breaks the expectation of a strongly typed language.

If something goes wrong and you want to fail as soon as possible not propagate the defect through the system.

-----

1 point by reitzensteinm 4155 days ago | link

Oh, it definitely throws strong typing right out of the window.

The reason I suggested it is because it would seem that almost all of the time where you go to do an increment on a nil value, you're working with an uninitialized element (not necessarily in a hash map) and treating that as 0 (as you're doing an increment) would in a certain sense be reasonable behaviour.

But I guess you're right, in the case where nil does represent an error, it'll be two steps backwards when you go to debug the thing.

-----

1 point by william42 4148 days ago | link

Or perhaps just set containsnil to 0 when you do that. (Knowing pg, this would probably work.)

-----

1 point by Tichy 4155 days ago | link

What is the usage scenario for that? I have never written such a code (incrementing values in hashtables) - maybe it is more common in LISP?

-----

1 point by bOR_ 4155 days ago | link

Happens when you want to categorize the frequency of items in a list, and I've been doing that all the time (categorizing gene frequencies in an agent-based model).

In ruby I'd extend the array class with this code

  class Array
    def categorize
      hash = Hash.new(0)
      self.each {|item| hash[item] += 1}
      return hash
    end
  end
although the other day I saw someone achieve the same thing using a hack on inject (the `; hash' part is only there because inject demands that, the work is done earlier.)

  array.inject(Hash.new(0)) {|hash,key| hash[key] += 1 ; hash}
Noticing that lisp / arc is more concise indeed. I'll have fun learning it.

-----

1 point by smallpaul 4152 days ago | link

Why would you extend rather than subclass the Array class? It kind of confirms all of my worst fears about Ruby's too-easy class reopening. (what happens when someone else defines an Array method called "categorize" for a totally unrelated purposes?)

I think that the Python syntax for this is

h[x] = h.get(x, 0) + 1

It isn't quite as concise as the Common Lisp but more so than Arc. I'd be curious to see what the Common Lisp looks like if you are doing something more complicated than an in-place increment. E.g. the equivalent of:

h[x] = h.get(x, 1) * 2

-----

2 points by bOR_ 4149 days ago | link

I'm a phd, working alone on projects, and the scripts I write a generally < 300 lines + 6 functions from a library I wrote. The agent-based models i write are ~ 200 lines, no libraries.

For me there's not much risk in redefining things.

-----

1 point by jsg 4152 days ago | link

(setf (gethash x h) (* (gethash x h 1) 2))

-----

1 point by ijoshua 4152 days ago | link

A hashtable containing integer values is a common implementation for the collection data structure known as a Bag or Counted Set. The value indicates how many instances of the key appear in the collection. Incrementing the value would be equivalent to adding a member instance. Giving a zero default is a shortcut to avoid having to check for membership.

-----