nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

TS6:FormAPI - 552 code not working?

Adam Stoller

I've got some FormAPI code that works in 5.5.2 SP2b, but doesn't seem to be working in TS6.0L (both running on Solaris)

The FormAPI code includes the following within the init() function: IWEventRegistry.addItemHandler("/Preview", "onCallout", checkIfSaved);(/Preview is a cgi-callout button within the DCT which runs a deployment to a "preview server" and then provides URL links to that preview server for previewing the rendered DCR data)
The code for checkIfSaved is as follows - with the two lines I supsect of having a problem being shown in red:function checkIfSaved(item) {
if (item.getName() == "/Preview") {
errMsg = "...Preview...\nWould you like to save now?";
}
else {
errMsg = "...Set Metadata...\nWould you like to save now?";
}
var title;
title = IWDatacapture.getItem("/General Info/Title").getValue();

if(IWDatacapture.isModified() || title == ""){
if(confirm(errMsg)){
parent.frames[1].saveClicked(1, null,true);
if(parent.frames[1].datacapture.isFormValid()){
return true;
}
else{return false;}
}
else{return false;}
}
else{return true;}
}However some sprinkling of alerts show that 'parent.frames[1]' and 'parent.frames[1].datacapture' are objects - but I'm getting an error in IE that says something to the effect of:

Error: Object doesn't support this property or method

This would be a lot more helpful if it told be which object didn't support which property or method.
Anyone able to pinpoint something in this code the changed from 5.5.2 to 6.0L ?

--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com

Find more posts tagged with

Comments

JonathonG

Hmm..don't have an answer for you. But, those 2 lines of code are the kinds of code I would *expect* to break from release to release, as they are not part of the documented API. I assume you're not using IWDatacapture.save() because it runs without validation, right? So, instead you're trying to simulate a button click. Do I at least have the intent of what you're doing right?

Jonathon
Independent Interwoven Contractor

Sri

Fish,

The FormAPI code that your FormAPI script has is not a public API. (those two lines where you access internal functions through parent.frames...). Any such methods are subjected to change without prior notice. Is there any reason why you can not use IWDatacapture.save()?

Sri

Adam Stoller

In response to both yours and JG's post - I have no idea. This is code I "inherited" and haven't spent much time looking through to understand and "fix". I guess I no longer have that luxory and will now have to spend some time trying to make sure I understand what the intended logic was and then figure out how to reproduce the same logic within.
I had a feeling it was those two lines ...
The recollection from one of the original authors was that IWDatacapture.save() didn't do any error checking but that the saveClicked() function did ... Can anyone verify whether or not this is an issue or whether IWDatacapture.save() will make sure that all the verification steps (both IWOV and custom checks) are run through

--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com

JonathonG

I can verify that IWDatacapture.save will not do all the error checks. Quoting from http://YOURTEAMSITESERVER/iw/help/tst/formAPI/formapi_28.html#SEC28:

IWDatacapture.save ()
Saves the current DCR if there is a name to save it to. Does not fire onSave, onSaveValid, or onSaveNameSpecified events. Does invoke onSaveDone event.
If no name is specified, either because it is a new DCR or because setDCRName(null) was called, save() does nothing.

Arguments:
None.
Returns:
Nothing.

Jonathon
Independent Interwoven Contractor

navneet

Please see the following post by me:
http://devnet.interwoven.com/forums/cgi-bin/showflat.pl?Cat=&Board=PRODUCTS_TEMPLATING&Number=30643&page=0&view=collapsed&sb=5&o=&part=

The reason I bring it up is because there seems to be some discrepancy in the init() method behavior going to 6.x. Support recommended not using the init() call. I don't know if this solves your issue, but may be worth looking at..

Migrateduser

Fish,

This is true; IWDatacapture.save() does not validate the form before saving. I'm afraid that you will have to validate the form within the script.

That being said, I see two lines of code that goes behind the formAPI (and thus is not supported.)

parent.frames[1].saveClicked(1, null,true);
and
if(parent.frames[1].datacapture.isFormValid())

In 6.1, the frame structure has changed, and thus parent.frames[1] probably accesses something completely different. On top of this, these functions may have changed (I know saveClicked probably has.)

--Daniel

Sri

Yes, Jonathon is correct. IWDatacapture.save() does a "blind" save of the DCR without any checks. This was as designed and hence not a bug. I can think of two options:
* One is to traverse the datacapture item tree (through getChildren()) and call isValid on each item
* Call getItem() to get hold of specific items that you may think can be invalid and call isValid() on those.

In any case, it'd be good idea for FormAPI to include an explicit method to validate the entire form. Can you file a FR for this? If you don't have access, let me know, I can do it.

Sri

Adam Stoller

I will file the feature request (via support case) - thanks for the confirmations - now I'll figure out how to code things so that the validation is performed... shouldn't be too difficult - right?

--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com

JonathonG

Based on my limited knowledge, it may be easier to try to figure out the 6.x "hack" that corresponds to the 5.5.2 "hack" that you already have. Otherwise, you have to write the whole save event tree logic from scratch (or at least those parts that you care about). That seems rather daunting to me. I think the FR should include the ability to initiate a Save as if the user clicked the save button. That way, not only does validation happen, but thinks like popping up a save as dialog (if necessary), and all of the event handlers getting called.

Just my $0.02 worth,

Jonathon
Independent Interwoven Contractor

Adam Stoller

FYI - this is the wording of my support case:

--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com

Adam Stoller

It's going to take me a bit of work because right now our validation routine includes a server callout with a callback function and I need to make sure that the place where I call it doesn't proceed until the callback determines whether to return 'true' or 'false'. I may end up having to create an alternate validation routine without the server callout and deal with the odd-cases where the piece that was being validated via the callout isn't valid later on (sigh)

--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com

Migrateduser

When you get an FR number, can you post it? I'd like to be on the FR list too.

- Jason

Adam Stoller

Will do - (I was planning on doing it anyway)

--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com

JonathonG

I think that's the point I was making. You're probably better off (from a purely amount-of-time perspective) digging into the backend stuff and figuring out how to simulate the Save button click, just like you were doing in 5.5.2. Obviously, that has the potential of breaking at the next upgrade, so you have to make the trade-off. But, trying to simulate the "Save" sequence seems to me an overly daunting task.

But maybe I'm just lazy Smiley Wink

Jonathon
Independent Interwoven Contractor

Adam Stoller

The # for this feature request is: 52759

--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com