Bug #223

source asteridex reported as broken

Added by tshif over 2 years ago. Updated over 2 years ago.

Status:Closed Start:10/30/2009
Priority:High Due date:11/02/2009
Assigned to:patrick_elx % Done:

100%

Category:-
Target version:Caller ID Superfecta Source Files Estimated time:2.00 hours

Description

(Posted by Ward Mundy)
AsteriDex module is hosed. Can't open DB and once I hard-coded it, the query always fails.

History

Updated by patrick_elx over 2 years ago

Ok two things I would like to know there:

first what should be the default value for the database login/password? (the default one is maybe wrong, but it should at least work after the proper info are entered in the option panel).

When the proper infos are in the option panel, what are the error report in the httpd log?

Again I do not have asteridex on my system and I based my update on the previous source query.

Updated by tshif over 2 years ago

  • Target version set to Caller ID Superfecta Source Files

Updated by tshif over 2 years ago

  • File source-AsteriDex.php added
  • Status changed from New to Assigned

More info has been provided by NerdUno:

Looks like the dsn[username] and dsn[password] are hard-coded to asteriskuser and amp109 despite what a user enters. We never use that MySQL user account for Asteridex, by the way. Database name also is being passed incorrectly. The wrong variables are also being passed to initialize MySQL. And then it looks like there's an error in the query string as well. Did someone actually try this??? The AsteriDex app includes sample airline entries so it's easy to test whether it's working.
In the meantime, here's a version of the file that hardcodes the usual entries which will work for 99.9% of folks.
--wm

Updated by tshif over 2 years ago

Additional info, from Ward:

Default entries should be....

username: root
password: passw0rd
dbname: asteridex
host: localhost

That query was a little much for me to tackle without a dozen or so cups of coffee, but mine works as long as entries in AsteriDex match the correct rightmost number of characters. :-)
--wm

Updated by tshif over 2 years ago

  • Due date set to 11/02/2009
  • Assigned to set to tshif
  • % Done changed from 0 to 70
  • Estimated time set to 2.00

Worked with Ward M to validate updated repaired datasource asteridex.

It looks like all problems have been corrected - but we are still waiting on validation of the default matching length for the data source.

Dev believes that 9 is the best choice - using the nine right-most numbers for typical 10 digit (USA) users, and the 9 (native) digits used by the EU, this default should work for most users without change.

We are awaiting user validation that 9 digits dies indeed match for both 10 digit and 9 digit users, then this updated source will be released.

Updated by patrick_elx over 2 years ago

We can maybe add in the flyby a comment saying that if you are in a NANPA country the best value would be 10.
You want to put this value at the max significant digits for your scheme. It can be different for different users. But starting out of the box with 9 is a one size fit all to start with.

Updated by tshif over 2 years ago

testing results from mr. Mundy:
Hi Tony,
Sorry for delay. Database match fails when set to 9. American Airlines = 8004337300. Setting to 9 and searching for 8004337300 fails. Also search for 004337300 fails. So I recommend you set 10 as default.
--wm

Updated by tshif over 2 years ago

More thoughts from Ward:

Just thinking out loud... Seems to me it would make more sense with AsteriDex and SugarCRM to match whatever number of characters are passed as the CallerID number, e.g. 12345 would search for all matches on 12345 and return the first one. 1234567890 would search for a match on that and return the first one. Because of MySQL indexing, right-to-left substrings are difficult to implement.

Updated by tshif over 2 years ago

  • File deleted (source-AsteriDex.php)

Updated by tshif over 2 years ago

  • Status changed from Assigned to QA Testing
  • Assigned to changed from tshif to patrick_elx
  • % Done changed from 70 to 80

This data source has been repaired to work as it is currently architected.
However, a change to internal functions has been suggested and needs to be evaluated.

Pending that possible change - I will update the Asteridex datasource in the live update repo now so folks with 2.2.2 can still use it.

Updated by tshif over 2 years ago

  • Status changed from QA Testing to Assigned

Updated by tshif over 2 years ago

  • Status changed from Assigned to Closed
  • % Done changed from 80 to 100

The part of this case related to the failure of the SQL query has been corrected, and the file posted to live update repo.

This case is being closed now to document the work done on this source for the failed sql.

A new case is now open to discuss the change in approach to the significant digits match. #227

Please make no further notations in this ticket unless it is related to the sql problem.

Also available in: Atom PDF