Aspell and

Mike C. Fletcher mcfletch at rogers.com
Tue Oct 22 01:38:28 CEST 2002


Kevin Atkinson wrote:
...

>You misunderstand.  word-list-compress simply compresses and decompresses 
>a sorted word list to save space.  It is not used in any way by Aspell 
>itself.
>
I understood that word-list-compress was an offline utility, but I'm not 
sure what you're using to search the compressed word-list on disk.  That 
is, how does _Aspell_ make use of the data once it's been compressed? 
 Is it something like a bisect algo with an index somewhere on disk to 
the items in the file?  (I gather by your reaction that the entire 
word-list is not all loaded into memory at run-time).

>leditdistance != typo edit distance.  leditdistance uses a different algo. 
>than the normal edit distance algorithm.  The normal edit distance algorithm 
>"editdist.cpp" and the typo edit distance algorithm "typo_editdist.cpp" 
>are basically the same except for the weights.
>
Ah, I had thought those were older implementations that had been 
superceded by the leditdistance algorithm (mostly because of the greater 
number of comments in that one).  I think for the moment I'll just leave 
the typo-distance code, and concentrate on the overall evironment.

>>It would seem that you'd need 
>>each "swap" to be a lookup into the typo table.  
>>    
>>
>
>Yes that is correct.  But it is not a "swap" but a replacement.  A swap is 
>when the adjective letters are interchanged, "teh" vs. "the".
>
Thanks for the clarification,
Mike

_______________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://members.rogers.com/mcfletch/







More information about the Python-list mailing list