On 24/02/2014, at 4:24 PM, Alan Silverstein wrote:

> Bear in mind that JudyNL is "just" an array-of-array... of JudyL, where
> the top level keys are key lengths, and each successive subarray carries
> four successive bytes of the keys.

It had better be 8 bytes on a 32 bit machine!

> 
> Cheers,
> Alan Silverstein
> 
> ------------
> 
> ...Doug was not happy about several things, if I recall right:
> 
> * JudyNL sorts indexes by length first rather than lexicographically
>  (left to right).

Yep, that sucks.

> * JudyNL is possibly overkill for a given application because it allows
>  a variety of index sizes in a single array, at a cost in complexity
>  and overhead, when perhaps any given user or application might want
>  only one or a set of single-size index arrays (fixed-depth trees),
>  like Judy2L.

Yes, but the user can more or less easily implement that.
At least, given JudyL I can implement it easily :)

> * And ultimately, no user had requested this feature, there was no
>  "pull", unlike JudySL.  We created JudyNL on my initiative, after I
>  recognized the difference between length-terminated and
>  length-associated variable-size indexes, I decided to implement both
>  (and internally they are quite different).


Doug's right on the ordering. The whole point is to get lexicographically
ordered C++ strings which might include nul bytes. 


--
john skaller
[email protected]
http://felix-lang.org




------------------------------------------------------------------------------
Flow-based real-time traffic analytics software. Cisco certified tool.
Monitor traffic, SLAs, QoS, Medianet, WAAS etc. with NetFlow Analyzer
Customize your own dashboards, set traffic alerts and generate reports.
Network behavioral analysis & security monitoring. All-in-one tool.
http://pubads.g.doubleclick.net/gampad/clk?id=126839071&iu=/4140/ostg.clktrk
_______________________________________________
Judy-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/judy-devel

Reply via email to