On Aug 05, 2006, at 8:02 AM, D Jungk wrote:
On Friday 04 August 2006 8:31 pm, Navdeep Bains wrote:
Hello,
Thanks for the input. Unfortunately, this isn't exactly a standard
use of the
database and the table names will be coming from file names. Each
table will
be a book (novel, textbook etc.) with one sentence per record. It
will be
used to search for phrases in context for people learning a
language.
For the description of the project above, it doesn't really sound
like you would need or want to create a new table for each book.
Could you go into more detail about what you're trying to accomplish?
Thanks,
Navdeep Bains
Bains Software
Sure, I'm always open to suggestions from this group.
I considered using a database or using a dictionary, but went with
the dtabase
because I think it will be faster to search. Basically the project
is this:
The user will drop text files into a folder. These files will be
text files of
books, articles, etc. in a foreign language. The files will show up
in a
listbox. You can then "load" or "unload" the files into the
database. This
will include detecting the encoding, splitting into sentences and
filling a
database table. I would anticipate maybe 15 to 50 files, each one
with 2,000
to 25,000 sentences. As they are loaded into the database, the user
will add
data such as Author, Year, Country of Origin, etc. I would imagine
that,
except for occasionally adding new sources, there will be very little
creating and dropping of tables.
Normally the user will go straight to the "Search" panel. Here, he
will have a
listbox with the loaded books along with the Title, Author name,
etc. He can
now check which books he wants to search. If he wants to find uses
of "No
manches", he is not going to search in books before 1975. If he's
looking for
"Quedáis con", he won't need to look in books from Mexico. (If he's
looking
for "boogerhead", he'll know only to look in books by Dave Barry.)
It's basically a tool for finding specific examples of usage of
phrases for
foreign language students, with the ability to look only in certain
countries
or time periods or certain authors.
So far I've tested searching one table with 10,000 rows and the
results appear
instantaneous, but with more and larger books, I thought I needed
to search
only in certain tables to keep it from being slow.
OK then
Try this
Table : Books
bookID integer
bookName string
+ maybe details like publisher etc but that depends on whether you want
Table : BookText
bookID integer
lineNumber integer
lineText string
Now you could find EVERY book that used boogerhead, or you could
restrict it to books written by Dave Barry
You may find things you had not expected by doing it this way and you
do not require your users to "know" to only search in certain kinds
of books or books before a certain date for results.
They can find any usage anywhere
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
