Using catalog_uri for pagination

[tdipisa]

Possible fix for #68 using the catalog_uri() method from utils.
Unit tests reviewed.

[tdipisa]

Hello @amercader. Please, can you take a look at this pull request?

[metaodi]

I really like re-using the existing utils method. But it's important to note the difference between a URL and URI. The goal of the catalog_uri() method is to create a unique URI for the catalog, so if both the ckanext.dcat.base_uri and the ckan.site_url are not present it will either use the app uuid or simply generate a UUID to get a unique URI.

One could argue that in these cases the pagination is not important, but we would intentionally break it. I think it makes sense to build the pagination on top of ckan.site_url and use the current request URL as a fallback (just like I proposed in #69). I should probably finish my work on #69, so we finally have a working solution for issue #68.

[tdipisa]

Thank you @metaodi for your reply. We urgently need this fix to allow RDF harvesting. @amercader what do you think about that?

[amercader]

@tdipisa I agree with @metaodi in that the site URL and the catalog URI may or may not be the same. Now that #69 is merged please check if that solves your issues with pagination.

[tdipisa]

Now it seems to work fine, thank you @amercader and @metaodi ;)