As Jorn said (and he'd be THE definitive source as the author of that
plugin) - the MAX parameter is not used by the server. Otherwise, I
think you have the gist of it.
The routine regarding the MAX goes like so:
- user enters some text
- that text is passed to the server side code via the Q parameter
- the server side code does something that may or may not make use of
the Q parameter to generate results. (The point is that what happens on
the server doesn't really matter for the plugin - as long as the results
are in the proper format)
- the server then sends those results back. ALL of them that matched
the server side processing. If you had 100 entries that started with
"A" and the user had typed "A" then you would be getting 100 entries
returned to your plugin (assuming you did the filter correctly on the
server)
- The plugin checks if it has a MAX value - if so then it takes the
first X results and displays them. (where X is the MAX value)
- If there is no MAX value (or it is set to zero - I think), then all
the results are rendered.
So, if the server side filtering results in 1000 values, then 1000
values are returned to the script. If a MAX value is not specified,
then 1000 values are displayed. So, you may want to set a height and
"overflow: auto" on the result div so you get a nice scrollbar.
Don't forget that MAX may have a default value. I think the docs say it
defaults to 10, but I remember getting eradicate issues with this. But
that was probably my specific code. Anyways - I found it better to set
an explicit value for MAX to get the behavior I wanted (which was to NOT
restrict the number of output rows...).
Not sure if it helps, but I blogged about the autocomplete plugin back
in December: http://grover.open2space.com/node/190 More to do with
making use of database IDs with the results though. Maybe this is a
little advanced for what you need, or maybe it'll trigger something
that'll help... (otherwise, my apologies if this is a shameless plug)
Shawn
dineshv wrote:
Hi Shawn.
Thank-you very much for the clear explanation. I mean "remote" as in
"the same domain". Let's see if I understand your explanation:
A large number (>10,000) of string items are held in a database (or
array) at the server. We want to make the string items available to
the autocomplete plugin at the browser. However, we only want a
(relevant and small) subset of the string items sent to the browser at
a time dependent on what the user enters into the input box.
Text (well letters, really) entered into the input box are sent to the
server as the autocomplete 'q' parameter. The server queries the
database (or array) to find MAX items around the user-entered text and
sends these items to the browser (as a JSON string). The MAX
parameter can be set within autocomplete or within the query to the
database (or array).
Is this the gist of it? Btw, my backend is written in Python.
Dinesh
On Apr 1, 4:09 pm, Shawn <[EMAIL PROTECTED]> wrote:
When you say remote, do you mean another domain?? If so, then yes there
are some problems there, but your server side page for getting your
results can handle that.
If you mean remote as in "not on the same web page, but from a
file/database accessible in the same domain", then I suspect a slight
misunderstanding is taking place here. In this case, you feed the
autocomplete plugin a file to request your data from. That file can be
server side processing code that will dynamically create your results
list for you. How that server side page goes about the process is
irrelevant to the autocomplete plugin - it only cares about what data it
is given. Sooooo, the sample you need is something like this:
$("#mytext").autocomplete("mydata.php");
Then the mydata.php file does the filtering you need:
<?php
$sql = "select name from employee where name like '". $_GET["q"] ."'";
//... database connection code ....
$result = $db->query($sql);
//... code to generate the autocomplete input based on the data
// in the $result variable.
writeoutput($resultText);
?>
Notice that the $sql will automatically filter the result list based on
what data is passed to the PHP file (via the Q parameter - which is what
the person has typed in the text box). So the results being passed back
to the autocomplete plugin are already filtered to be a smaller subset
of data.
Now, that subset of data can still be a large number of items. You can
address that by either changing the SQL statement, the way the server
side code works, or by setting the "max" parameter of the autocomplete
plugin. Setting the max parameter will only show X number of items -
but all the results are still returned to your script. So if you are
simply getting too much data back, the ideal solution is to change the
server side code to something more suitable.
Hopefully I have properly understood what you were asking for, and that
this helps you out. G'Luck.
Shawn
dineshv wrote:
Thanks Hamish but as far as I can tell this is a known problem ie.
retrieving remoteJSON-format data using the autocomplete plugin.
Be great to hear from folks who know if this has been fixed in the
latest autocomplete plugin code and if example code exists to show how
it is done. Cheers!
Dinesh
On Apr 1, 3:10 pm, Hamish Campbell <[EMAIL PROTECTED]> wrote:
TheJSONexample on the autocomplete
site doesn't work as it sends the entire
remote data to the browser instead of
returning just a limited number of items.
Probably becuase it requires a backend page to provide dynamic results
based on the search query.
Note:
When the user starts typing, a request
is send to the specified backend
("my_autocomplete_backend.php"),
with a GET parameter named q that contains
the current value of the input box and a paremeter
"limit" with the value specified for the max option.
A value of "foo" would result in this request url:
my_autocomplete_backend.php?q=foo&limit=10
The result must return with one value on each line.
The result is presented in the order
the backend sends it.
