is this feasible?

Anon

hi.
i had an idea about data retrieval and presentation. i don't know if it's feasible.
at my level, i could only do one-step data presentation. after query, i get the result. however, i have too many fields in my result page that if i get 10 results, my page becomes too long.
what if i just wanted one row (like lastname for example) to show up but make it hyperlinked to a page that contains the rest of the result for that particular row. is this feasible at all? and how?
thank you again.

heliumflash01

Manny -

I'm not quite following your entire train of thought, but if I'm reading correctly, there's no problem with that.

Just get all the fields you want to display, iterate through a FOR or WHILE loop to output the rows, and make each instance of whichever field (e.g. lastname) a link in your PHP-outputted HTML.

As for the display of the rest of the row result, just have the link point to another PHP page, sending along the primary key of that row (it's ideal if you have an auto_increment field to uniquely identify each row). Like this:

Then on that next page, just do a query that grabs all fields from the database where row_id = $row_id (which is passed in on the querystring).

-Jeremy

Anon

jeremy,
thank you! yup, that's what i was asking about, and i think that will work. i didn't know i could pass a variable to another page that way. thanks again. i'l update you once i get it to work.

e42078

I've pondered your situation myself. Breaking large result sets into many pages can be tricky on a site that must handle many users. I don't know exactly how search sites do it, but here's what I've dreamed up:

Create a results table to store the primary keys of a user's search results along with some sequence number (if needed).
Setup your web page to pull chunks of rows from this results table by sequence number ranges. MySql, as I recall, has an option to select based on row position. Oracle sort of does (ROWNUM), but a column sequence would work better in this situation.
Use session id's in the results table so that you can differenciate between many users' results and also have some process "age" the data out.

Anon

jeremy,
i used your code above, configured it, but i was getting parse errors. i put it into my do-while loop and appended echo in front of it. i even tried putting in semicolons but i still couldnt get it. im using dreamweaver, and it's coloring the </a> yellow.
what's wrong?

Anon

thanks david. i know this is a good idea, but its way above my head. i don't quite reach your level of explanation. simpler please?

e42078

Let's say you have some relational tables that store employee information. Your main table could be setup with SSN as a primary key (doesn't always work in the real world, but good for explanation). You then have a web page that allows a search on all employee attributes (position, salary, name, address, etc.). So user enters their search criteria and submits. The big question, as I understand it, is how to deliver the search results if there were, say, 10000 employees that matched the criteria? Well, here's what I would do:

Create a SEARCH_RESULTS table with three things. (1) SSN, (2) SESSION_ID (From PHP session), and (3) SEQ_NO (starts from 1 and iterates).
Put the 10,000 hits for the user's search into the SEARCH_RESULTS table. If you have 10 users each with 10,000 hits, you'd end up with 100,000 rows in that table. Each user's result set is differenciated by their SESSION_ID. You can deliver the results to their browser in blocks by specifying a SEQ_NO range (start with 1-20, next page 21-40, etc.). That delivery to the browser will be a SQL statement that joins the SEARCH_RESULTS table with the employee table(s) via SSN for them to see all those employee attributes mentioned above.
Because this SEARCH_RESULTS table will experience exponential growth, you'll need to manage data left over from previous searches. I'm not sure exactly how to do this, but you would probably want to set SESSION_ID's to expire somehow and coordinate that with, perhaps, a DATETIME field in the SEARCH_RESULTS table (although that wouldn't be normalized. A normalized model would be a SESSIONS table with the SESSION_ID and EXPIRE_DATETIME field and SEARCH_RESULTS as a child table). Then you could update the EXPIRE_DATETIME for each new search that user (SESSION_ID) executes.

Hope this helps.

e42078

Manny,

You know what? I re-read your initial message and realized that my other long-winded thread about results tables solves a slightly different problem. I've actually implemented something just like what Jeremy suggested. Within my <?php ?> tags, inside my loop, I have:

echo "<a href=\"doc_detail.php?row_id=$row_id\" target=\"doc_detail\">";

That'll be your starting anchor tag, which you'll have to terminate (i.e. </a>). Basically, this delivers the primary key (in this example, stored in the $row_id variable) directly to the doc_detail.php web page, which will then be able to use it to fetch the rest of your columns from the database. Also, it will do this in a popup window because of the "target" option.

Anon

thanks david. i got it as far as passing to the next.php page. then i did this:

$result= "select * from mytable WHERE row_id LIKE '$row_id'";

then my next.php outputs "sorry no records found!".
if i remove the where clause i get all the data in my table.
i guess my next question is: how does next.php page catch the passed row_id and insert it to its query?

i think you should be writing an article on this already.

thanks again.
manny

Anon

i finally got it! this is how it looked like:

echo "<a href=\"doc_detail.php?row_id=$row[row_id]\"
target=\"doc_detail\">";

the difference was :
$row[row_id] : it was just row_id before.

and in the next.php my select statement looked like:

$result="select * from mytable WHERE row_id LIKE '$row_id'";

and it worked !

thank you so very much for your guidance!