Video scan doesn't seem to recognize IMDB number (1 Viewer)

[2BitSculptor]

Trying to scan Cinderella (2021), tt10155932. When I get the IMDB results, the last digit is missing. So it is named "unknown"

BTW, This is for MyVideos, so it problably should go to a more appropriate section.

 

[ajs]

When I get the IMDB results, the last digit is missing. So it is named "unknown"

IMDB result? Grabber?
Update grabbers, set TI_MDB and try again, if not work, post log ....

 

[2BitSculptor]

Update grabbers, set TI_MDB and try again, if not work, post log ....

Thankyou. It is working now.

 

[CSMR]

TI_IMDB not working for me. Doesn't detect the tt number as it should. "No IMDB info found" error popup. Now the source code is on github, maybe I can add a test case?

 

[catavolt]

Doesn't detect the tt number as it should. "No IMDB info found" error

Same here, but only when trying to actualize the respective movie

 

[CSMR]

I found that the TI_IMDB script doesn't recognize the MovieName (date) [tt123...] format with square brackets, which either it or other grabbers in the past did recognize. Tests will make this clear and I started work on this, but blocked by problems in the repo preventing a build (described here).

 

[ajs]

MovieName (date) [tt123...]

Try:

MovieName (date) - tt123...

 

[CSMR]

The repo is not in good shape but I'll try for minimal change to get things to work without doing massive cleanup at this point.
I was able to get the test project to build via a nuget.config change.
The test projects didn't run and I found they had nunit dlls checked into the repo! So updated to nunit 3 so now they run (but often fail) in VS.

Ideally there would be:

• an automated yml build and run of test projects in github
• consistency of handling of video title formats between title formats confirmed by tests
• an md file explaining this format and deployed to the website
• removed need for individual scripts to parse title formats, with this being done centrally and the script interface modified
• replacement of .csscript files with .cs files to get intellisense on them, centralize creation of these files, remove the need for interpreters

For my PR I'll just include:

• fix the nuget.config file
• update nunit so that the tests can be run (with many existing ones failing)
• make a small md file documenting the format based on the consolidated regexes across the various script files
• add tests checking this format and checking that data is retrieved from imdb/tmdb, fixing the failing scripts

For me these script files have failed many many times over the years so after this is done my life is going to get better, at least when the next mediaportal version is released. I hope that's the same for others too.

@ajs can someone enable github issues?

 

[ajs]

The repo is not in good shape but I'll try for minimal change to get things to work without doing massive cleanup at this point.

Everything is working fine—what problems are you talking about?

I was able to get the test project to build via a nuget.config change.

The box already has the correct configuration; everything is set up and nothing needs to be changed.

The test projects didn't run and I found they had nunit dlls checked into the repo! So updated to nunit 3 so now they run (but often fail) in VS.

They haven't been used for about 10 years and don't cover all the functions of the MP.

an automated yml build and run of test projects in github

This is how it works, and this is exactly how all releases are built and pull requests are tested.

• consistency of handling of video title formats between title formats confirmed by tests
• an md file explaining this format and deployed to the website

I don't see the point of that. We won't be able to verify whether the image is displayed correctly on the screen anyway.

replacement of .csscript files with .cs files to get intellisense on them, centralize creation of these files, remove the need for interpreters

This would undermine the entire logic of custom grabbers, as well as the flexibility of updating and modifying them. You shouldn't do that.

fix the nuget.config file

No need—it's still relevant.

update nunit so that the tests can be run (with many existing ones failing)

I don't see the point, since the tests don't cover the current functionality, but if you'd like to, that's fine.

• make a small md file documenting the format based on the consolidated regexes across the various script files
• add tests checking this format and checking that data is retrieved from imdb/tmdb, fixing the failing scripts

That’s a dead end. The TMDB API is well-documented, but IMDb’s regular expressions change multiple times a year, as does the site’s structure. Regular expressions are a dead end. The internal IMDb API, on the other hand, isn’t documented, but it’s quite stable.

We have a wiki for documentation.

For my PR I'll just include:

One issue — one Jira ticket — one branch — one pull request. This is the Way.

@ajs can someone enable github issues?

What for? All issues are discussed here on the forum.

 

[CSMR]

Everything is working fine—what problems are you talking about?

OK. It sounds from your reply that some of the dev team is committed to maintaining the repo in a non-working state based on the fact it's been like that for decades. That's a shame because mediaportal cannot progress if so.

The problems were:

• I cloned the repo and went to build the test projects. It did not build without a nuget.config change. That was problem no. 1. I can give you screenshots to prove it.
• After that, visual studio was not able to run the tests. The tests had a copy of nunit as a dll. mediaportal/MediaPortal.Tests/nunit.framework.dll That's "working fine"? Updating to nuget 3 fixed that and I was able to delete the dll.
• After getting tests to run, they fail. Who cares because you "don't see the point in tests"... is that a joke?

This is how it works, and this is exactly how all releases are built and pull requests are tested.

Searching the repo for "yml" brings up no relevant results. I don't see any build definition.

We have a wiki for documentation.

Is that in source control? A code search brings up nothing. I assume it's there but not in the repo. That's not ideal but is manageable - someone with access just needs to make the change once the definition is clear.

This would undermine the entire logic of custom grabbers, as well as the flexibility of updating and modifying them. You shouldn't do that.

In theory I understand your point, but this wonderful flexible system of scripts is completely broken. Of the 12 scripts, only TI_MDB is close to getting data successfully, in the tests in my draft PR Add grabber script tests and get some working by charlesroddie · Pull Request #458 · MediaPortal/MediaPortal-1 .