I'm trying to use a jQuery UI Dialog for a sort of "soft" validation--if the form looks suspicious, warn the user, but allow them to continue the submission anyway:
// multiple submit buttons...
var whichButton;
$("#myForm").find("[type=submit]").click(function()
{
whichButton = this;
}
var userOkWithIt = false;
$("#myForm").submit(function()
{
if (dataLooksFishy() && !userOkWithIt)
{
$("Are you sure you want to do this?").dialog(
{
buttons:
{
"Yes": function()
{
$(this).dialog("close");
// override check and resubmit form
userOkWithIt = true;
// save submit action on form
$("#myForm").append("<input type='hidden' name='" +
$(whichSubmit).attr("name") + "' value='" +
$(whichSubmit).val() + "'>");
$("#myForm").submit(); /****** Problem *********/
},
"No": function() { $(this).dialog("close"); }
}
});
return false; // always prevent form submission here
} // end data looks fishy
return true; // allow form submission
});
I've checked this out with a bunch of debugging alert statements. The control flow is exactly what I expect. If I first fail dataLooksFishy(), I am presented with the dialog and the method returns false asynchronously.
Clicking "yes" does re-trigger this method, and this time, the method returns true, however, the form does not actually submit...
I'm sure I'm missing a better methodology here, but the main target is to be able to simulate the behavior of the synchronous confirm() with the asynchronous dialog().