Issue: Parsing wikimedia image paths (with multiple paths)

[matadorhernan]

Test case:

https://en.wikipedia.org/wiki/Facebook

Issue

The page returns infobox.image() as one path for two files and the current solution is redirecting as a combined entity as opposed to an array of entities. You can fix that changing lines: 4 to 11 but I don't know how would that affect your code

Fix

//src/image/Image.js

//line 4
const encodeTitle = function (path: string) :string[] {

  //separate files using split() join()
  let paths = path.split(".svg").join(".svg{}") //repeat for all kinds of media
                    .split("{}");

//remove all falsey values I usually use lodash and I just noticed I don't know
//which method is best if not using lodash but here, (whenever the image path only contains one file the second index is falsey)

  paths.filter(file =>{
   return file != undefined && file != '' && file != null;
  });
  
  paths.map(file =>{
     let title = file.replace(/^(image|file?)\:/i, '');
     //titlecase it
     title = title.charAt(0).toUpperCase() + title.substring(1);
     //spaces to underscores
     title = title.trim().replace(/ /g, '_');
     return title;
  }) 

  return paths;
 
}

I know you probably can code this with a regex but I don't know how, so whenever you do Ill see it.

[spencermountain]

thanks Miguel - yeah, good find. This is the first case I've seen of two images in one infobox property - we are grabbing the logo text without a test, or validation.

For reference:

{{Infobox website
| name = Facebook
| logo = [[File:Facebook f logo (2019).svg |100px]]<br/><br/>[[File:Facebook Logo (2019).svg|196px]]
}}}

We're stuck behind a breaking change on dev branch, but i'm happy to add this to the next release, which may be in a few weeks.
cheers