CFQueryReader v1.2 - Critical Update Supporting ExtJS 3.x

@Andrew,

Are you using Ext.Direct? That line is only if you're using Ext.Direct, and have included the DirectCFM stuff. Best guess is that you only require the stuff from examples 2 and 3.

In CFQueryReader.js there is this line (at 161):

fi[b].mapArrLoc = cols.indexOf(Ext.util.Format.uppercase(fMap));

This means, that cfqueryreader *always* searches for capitalised column names in the receiving JSON string, even when they are not..

so a Json string like :

{"COLUMNS":["ID","afield"],"DATA":[[1,"1"]]}

will fail to load data because CFQueryReader will look for 'AFIELD' instead of 'afield'...

You are right, that would fail. I've never seen a ColdFusion return that didn't uppercase the column names, even in manually built queries using QueryNew. However, I can see the point.

The issue at hand is that JavaScript is case sensitive, whereas ColdFusion is not. We have to convert the return, in some fashion, to standardize the content prior to processing. For this, I would uppercase the values of the COLUMNS array (around line 140) prior to processing:
<code>
var cols = this.getQueryRoot(o).COLUMNS;
      
   for (var i = 0;i < cols.length;i++){
      cols[i] = cols[i].toUpperCase();
   }
</code>
Give that a try, and see if it resolves the issue. Let me know if that works out...

@Azadi,

Sorry, I just got back from vacation. Use the "Contact Blog Owner" link, at the bottom of the page, and send me a copy of your config, as well as a copy of the JSON return you are getting (Firebug will show that), and I'll see if I can't help you figure it out.

@Tony,

Send me the details through the "Contact Blog Owner" link at the bottom of the page, and I'll see what I can do to help out.

@Azadi,

Thanks Azadi. That is exactly the info I need to move forward with this. I'll see what I can dig up today.

The CFQueryReader.js code at RIAForge is still the same as the old code

Excellent... everything works now. Many thanks to Azadi who helped me on Experts Exchange and especially to Cutter for creating such a great reader.

Cutter... since you are a ColdFusion guy would you have any idea why a column that is defined in my 2008 sql server database as a varchar(5) and contains county FIPS codes would come through as numeric? If I do a cfdump of the query in the cfc the value will show as 06017 but when I see it in the console it is 6017.0 numeric. Here is the record model mapping:
      {name:'FIPS',mapping:'FIPS',type:'string'}

Hey Cutter, I got hooked on ExtJS when i attended your talk back at webmaniacs in 2008 and have your book. Awesome btw. Now i'm starting to play w/ 3.0 and the awesome new features like the Adobe Chart. Ran into a problem though and was wondering if you've encountered it before?

When using cfqueryreader to pull a recordset to feed a series to the chart component, my series fail to take up the whole width of the chart. If there are 2 series it uses half of the x width, if there are 3, it uses 1/3 and so on. Almost as if the chart is expecting to use the total count of the data points. I changed the data reader to the old school cfjsonreader that I used to use and it works just fine. Any ideas on what's going on? Could it possibly be a bug?

Again, thanks for getting me hooked on this awesome framework, and all of your contributions.

CFQueryReader is just what I needed. I've built my site around Spry and CF Ajax functionality, but there are some UI things I need that Ext has and I didn't want to toss my CFCs since they work for everything else.

I use two ways to produce JSON from CF:
1) Return a query object and specify returnFormat="json" in the function definition
2) Return a string from SerializeJSON(qry, false) and specify returnFormat="plain" in the function definition

Most of my coding uses the last one, since I can use a custom JSON prefix scheme (I store a prefix in the Client scope that gets changed with every page request and check it on the client when data arrives).

Spry will consume the last one also, but you need to use SerializeJSON(qry, true) to make it work. This return
comes in the following format:

{"ROWCOUNT":8,"COLUMNS":["F1","F2"],"DATA":{"F1":[180,181,182,183,184,185,186,187],"F2":["Air Nigeria","Chrome Air Service","EAS Airlines","Falcon Airlines","Merchant Express Aviation","Nexus Aviation","Okada Airlines","RiteTime Aviation"]}}

As you can see it's very similar to the output from the CF QueryConvertForGrid() method, so i should be able to enhance it to accommodate it.

Again, thanks for a nice piece of work and for giving me the starting point I needed.