File searching in pocket pc

I've asked similar things months ago but couldn't get
reply... so I'd like to ask in another way.

I'm writing a English-Chinese dictionary for pocket pc.
But the problem is the it's very time consuming and also
seems impossible to load all data (word and meaning) to
an array before perform searching.

So shall I use some file-based searching techniques? If
yes, can anyone give me some suggestions? Or there's some
other methods for pocket pc actually? I'm writing the
program by evc.

I've searching for informations for few months already
and really feel so hopeless.

Thanks alot!

Tashkant
Fri Jan 02 07:35:37 CST 2004

this is not really a pocketpc question...it's an algorithm
question. anyways...

you don't want to load the entire dictionary for obvious reasons
that you found out.

what you want to do is a fast file based search...but searching
the disk for a word will take a very long time.

so you have to use something called an index table (which
could be part of the dictionary file or a separate file)..it
basically allows you to do quick searches.

you want to be able to narrow down quickly to a small subset.
...do a search in the internet for "hash tables".

every word can be represented by a number - for example a
simple one is the sum of all it's ascii letters. it won't be unique
number but the resulting arrow will be small enough.

then save all the numbers with the same number (hash value) in
a separate array in the index table along with a pointer to the
real word and chinese text in the dictionary.

----

so the user types in a english word...this is what your program
does:
- convert the word to it's hash value
- search the index table for the array of similar hash values
- for each value in the index table
search in dictionary
if word is same as english
found - display chinese translation.


as i wrote before, do a search in the internet for "hash tables"
or "hash algorithm"


"Caprice" <caprice_1982@hotmail.com> wrote in message
news:04c901c3d0fc$23c00a70$a101280a@phx.gbl...
> I've asked similar things months ago but couldn't get
> reply... so I'd like to ask in another way.
>
> I'm writing a English-Chinese dictionary for pocket pc.
> But the problem is the it's very time consuming and also
> seems impossible to load all data (word and meaning) to
> an array before perform searching.
>
> So shall I use some file-based searching techniques? If
> yes, can anyone give me some suggestions? Or there's some
> other methods for pocket pc actually? I'm writing the
> program by evc.
>
> I've searching for informations for few months already
> and really feel so hopeless.
>
> Thanks alot!

caprice
Fri Jan 02 09:13:37 CST 2004

Sorry for posting in a wrong section...
but anyway, thanks very much.
It runs quite smoothly with hashing now.


>-----Original Message-----
>this is not really a pocketpc question...it's an
algorithm
>question. anyways...
>
>you don't want to load the entire dictionary for obvious
reasons
>that you found out.
>
>what you want to do is a fast file based search...but
searching
>the disk for a word will take a very long time.
>
>so you have to use something called an index table (which
>could be part of the dictionary file or a separate
file)..it
>basically allows you to do quick searches.
>
>you want to be able to narrow down quickly to a small
subset.
>....do a search in the internet for "hash tables".
>
>every word can be represented by a number - for example a
>simple one is the sum of all it's ascii letters. it
won't be unique
>number but the resulting arrow will be small enough.
>
>then save all the numbers with the same number (hash
value) in
>a separate array in the index table along with a pointer
to the
>real word and chinese text in the dictionary.
>
>----
>
>so the user types in a english word...this is what your
program
>does:
> - convert the word to it's hash value
> - search the index table for the array of similar
hash values
> - for each value in the index table
> search in dictionary
> if word is same as english
> found - display chinese
translation.
>
>
>as i wrote before, do a search in the internet for "hash
tables"
>or "hash algorithm"
>
>
>"Caprice" <caprice_1982@hotmail.com> wrote in message
>news:04c901c3d0fc$23c00a70$a101280a@phx.gbl...
>> I've asked similar things months ago but couldn't get
>> reply... so I'd like to ask in another way.
>>
>> I'm writing a English-Chinese dictionary for pocket pc.
>> But the problem is the it's very time consuming and
also
>> seems impossible to load all data (word and meaning) to
>> an array before perform searching.
>>
>> So shall I use some file-based searching techniques? If
>> yes, can anyone give me some suggestions? Or there's
some
>> other methods for pocket pc actually? I'm writing the
>> program by evc.
>>
>> I've searching for informations for few months already
>> and really feel so hopeless.
>>
>> Thanks alot!
>
>
>.
>