How about some (more) fun? Can I put a "Radical Mode" checkbox in CLTBoss? It will bring a chuckle only to the insiders. I mean, Tesla has "Ludicrous" mode, so why not? 

One reason for a more compact XML format is size. The 3000+ profiles XML file chokes everything (CookTop, Excel, Chrome). The Invelos XML includes everything, including all the credits for people not in the variants table. Maybe I'll do another run of Christopher Lee and send you a link to take a look and give suggestions. Meanwhile, I am going into debug mode and just output a one line text record to validate results.

Quoting GSyren:

Quote:

Quote:
The 3000+ profiles XML file chokes everything (CookTop, Excel, Chrome).

Sure, but is it really useful to open such a file that way, other than for debugging purposes? You need an analysis tool to get meaningful information from either format.

Quote:
The Invelos XML includes everything

Yes, and that's the point. It doesn't limit its use to just doing common name analysis. Think outside the box!

I think I'm doing a pretty good job at thinking outside the box, and trying to use all the resources at my disposal. But I am having trouble validating my results during my testing and looking for ideas. I will figure it out, as I have been doing all along. It was my decision to (1) load the CLTPlus XML and (2) Create Invelos XML output, long before being asked to do so - with the express purpose of enabling the activities you describe. So, yes, the point is in fact at this time, debugging.

Here is the next thing I could use help on:

Remember / recap: When I scrape the UPCs, I do not have localities, due mostly in part to the difficulty in scraping from the CLT. All the methods I have tried so far, take enormous amounts of time. Finally, I am scraping the UPCs reliably and relatively fast.

So, having the (formatted) UPC, which actually includes the discID style that starts with "I" and the variants style, which start with "V", I make a call to the online database "give me all profile IDs that use the given UPC as a base". A large percentage of the time, I get one profile ID (which includes the locality I was missing), and fine - just grab the XML.

But in other cases, I get more than one profileID. Usually, it is other localities for the same DVD, and the actor is credited in all of them. BY default, I grab the XML, so fine, no problem.

Now here is the sticky part: there are two major cases where I get profiles that were not included in the original CLT count:

- The extreme case that we saw recently where the same "UPC" was used as the base for a bunch of unrelated movies, only one of which contained the desired credit
- Cases where the actor is not credited in all localities for the same movie. I will get those profiles, but some were not counted by the CLT due to the "uncredited" flag.

There is only one way to resolve it. Search the credits.

My feeling is that I want to just blast them into the XML without doing a credit search to weed out the exceptions, but are the intended users of this tool's output willing to accept what is - by definition - not the exact result of the CLT? (in the interest of getting the output in the least time possible)

Quoting GSyren:

Quote:
Quoting mediadogg:

Quote:
I think I'm doing a pretty good job at thinking outside the box

Just for the record, I didn't mean to imply that you weren't! It was more me thinking out loud. "Yay, we're thinking outside the box!"

Hmm ... Can you think out loud in writing?

Thank you. On to the next one ... need help again. See my previous post.

Was really hoping you would say that. If have to do a search, we are back at square 1.

By the way, CLTInfo would be awfully handy for my testing. Don't know any faster way of loading and viewing XML info. So, if you don't mind, I would like to get access to the latest version.

Here is a small sample of actual variants output (dvds in debug format) - not sure you need it, but just in case.


<?xml version="1.0" encoding="windows-1252"?>
<!--DVD Profiler Collection Export-->
<Collection>
<!--Start CLTBoss Plugin Extension-->
<Variants>
  <Variant FirstName="Christopher Lee" MiddleName="" LastName="" BirthYear="0" HashName="Lee_Christopher_">Christopher Lee</Variant>
</Variants>
<!--End CLTBoss Plugin Extension-->
<DVD>001861921725</DVD>
<DVD>0044007826225.2</DVD>
</Collection>


Reminds me, I have to add the Encoding for codepage 1252. I forgot.

Quoting GSyren:

Quote:
Quoting GSyren:

Quote:
Quoting mediadogg:

Quote:
My feeling is that I want to just blast them into the XML without doing a credit search to weed out the exceptions, but are the intended users of this tool's output willing to accept what is - by definition - not the exact result of the CLT?

Works for me. I can just ignore the profiles that do not contains relevant name information.

On second thought ... I may have been thinking "inside the box" here.

Irrelevant profiles may throw a spanner in the works when it comes to using the XML for other types of queries. But maybe we can ignore that small problem for now?

I was hoping that you would offer to sanitize on the way in, then never have to worry about it for subsequent queries.

Quoting mediadogg:

Quote:
I was hoping that you would offer to sanitize on the way in, then never have to worry about it for subsequent queries.

That requires that the output is run through CltInfo before being used for other purposes. But maybe that's a limitation that we can live with.