Search Text in Multiple Documents

[shantanu]

Hi All,
I have a requirement to develop a search engine based on
some search criteria that will search for the string or statement in
all the documents uploaded in the website. The search result will be
displayed same as the Google Group search, with highlighted texts.

I have few questions:
1. What will be the logic to implement?
2. The documents that will be uploaded in the website will be saved in
the datbase or not?
3. If the document contents are not saved in the database and stored
as links of the files,
then how will the search take place
4. What will be the concept i should follow?

Kindly Have a discussion.
Regards and thanks

Shantanu

 

[Alberto Poblacion]

I have a requirement to develop a search engine based on
some search criteria that will search for the string or statement in
all the documents uploaded in the website. The search result will be
displayed same as the Google Group search, with highlighted texts.

I have few questions:
1. What will be the logic to implement?
2. The documents that will be uploaded in the website will be saved in
the datbase or not?
3. If the document contents are not saved in the database and stored
as links of the files,
then how will the search take place
4. What will be the concept i should follow?

Kindly Have a discussion.

You could use the Index Server that comes with Windows. Basically, you
set up Index Server on Windows pointing it to the physical directory where
your documents reside. Index Server will build an index that will later
enable it to do fast searches in the documents.
When you want to do a search, you send the query to Index Server in a
language that is similar to SQL. You can do this from your code by means of
OleDb and the OleDb driver for Index Server. You get back the list of
documents that satisfy the search criteria.

An alternative would be to save all documents as BLOBs in a SQL Server
database, and then use Full-Text Search to locate docuents that satisfy the
search criteria. If your database can handle the load, this will work very
nicely. If you don't want to duplicate the documents on the database and the
filesystem, you can write a single GetDocument.aspx page that will take a
document ID as a parameter and then write onto the Response the content of
the document extracted from the database. It could also highlight the
matches on the fly.

 

[Guest]

To add to Alberto:

SQL Server provides search filters for xls, doc, html, ppt and text files.

 

[shantanu]

You could use the Index Server that comes with Windows. Basically, you
set up Index Server on Windows pointing it to the physical directory where
your documents reside. Index Server will build an index that will later
enable it to do fast searches in the documents.
When you want to do a search, you send the query to Index Server in a
language that is similar to SQL. You can do this from your code by means of
OleDb and the OleDb driver for Index Server. You get back the list of
documents that satisfy the search criteria.

An alternative would be to save all documents as BLOBs in a SQL Server
database, and then use Full-Text Search to locate docuents that satisfy the
search criteria. If your database can handle the load, this will work very
nicely. If you don't want to duplicate the documents on the database and the
filesystem, you can write a single GetDocument.aspx page that will take a
document ID as a parameter and then write onto the Response the content of
the document extracted from the database. It could also highlight the
matches on the fly.

The second option is good, that was my idea too, i have to check the
feasability once. But if the database thing is not feasible and if the
documents are stored in a directory structure, then how will the
search criteria work. You told Index Server is a option, but how will
this technique interact with my website, where i am using ASP.net and
C#. Are there any help available to use the meathods of Index server
from C#.

regards
shantanu

 

[Guest]

As Alberto mentioned, Index Service is a separate OS component and you use
ADO.NET to interact with it. What this means is that you have to store the
uploaded documents in a physical folder on the server and have Index service
create a searchable index of that folder. A sample is here:
http://www.codeproject.com/aspnet/search.asp

 

[shantanu]

As Alberto mentioned, Index Service is a separate OS component and you use
ADO.NET to interact with it. What this means is that you have to store the
uploaded documents in a physical folder on the server and have Index service
create a searchable index of that folder. A sample is here:http://www.codeproject.com/aspnet/search.asp







- Show quoted text -

Thanx Siva
I am going tru the link.
regards
shantanu

 

[shantanu]

Thanx Siva
I am going tru the link.
regards
shantanu- Hide quoted text -

- Show quoted text -

I got a very good link that tells everything
http://support.microsoft.com/default.aspx?scid=kb;en-us;820105
Regards
shantanu