Can EditBot replace URLs that go to Nasjonalbiblioteket?

Question

The National Library of Norway site Nasjonalbiblioteket (nb.no) is used pretty extensively on WikiTree, mainly because the site hosts Norway farm and family books (Bygdebøker) that are very useful for genealogy. There is now an announcement at their nb.no/nbsok site that says:

"nbsok som eget nettsted legges ned i løpet av 2022. Alle dokumenter i nettstedet kan fortsatt søkes opp i Nettbiblioteket"

(nbsok as a separate website will be shut down during 2022. All documents on the website can still be searched in the Nettbiblioteket)

It sounds like any links that previously went to nb.no/nbsok/nb will stop working by the end of this year. Links will have to change to the new format nb.no/items instead and the page indicator will work differently too. For example, 

https://www.nb.no/nbsok/nb/3e2cda93c583944806de76f26e5d1d8d#145

will now be:

https://www.nb.no/items/3e2cda93c583944806de76f26e5d1d8d?page=145

(Just a note: many of the books on the site must be accessed using a Norwegian IP address)

The old format displayed the page number in a few different ways too. For example:

https://www.nb.no/nbsok/nb/3e2cda93c583944806de76f26e5d1d8d#147

or 

https://www.nb.no/nbsok/nb/3e2cda93c583944806de76f26e5d1d8d?page=1#147

or 

https://www.nb.no/nbsok/nb/3e2cda93c583944806de76f26e5d1d8d?index=5#147

Would all return the same page number. Now going forward it will only be formatted like this:

https://www.nb.no/items/3e2cda93c583944806de76f26e5d1d8d?page=147

I looked in WikiTree+ and there are currently 17,732 WikiTree profiles that link to nb.no. But I cannot tell how many of these links go to the nb.no/nbsok address. Would EditBot be able to change the old format url to the new format to prevent all the links from breaking?

Answer 1

EditBOT would be able to do it.

Usually we don't change the links until they work. So I can do it next year if they will actually stop working.

I can check all linkd to domain also by path. There seems to be another case of url

http://www.nb.no/nbsok/nb/361fb9acb74fecc79bd37c51eb98aea6.nbdigital?lang=no

There are actually 4127 links with /nbsok/nb/ on 3958 profiles but they link to only 340 unique URLs. It will be asome work for editbot, but not too much.

Here is the list of all links to the domain. https://www.softdata.si/wt/nb_no.xlsx

Comments

Thank you Ales! Yes that makes sense if we wait until the links stop working.

I'm a little concerned that the list of links does not seem to be picking up the page #s in the old url. In the old url format, the page number was always indicated with a # sign then the page number. 

For example one of the URLs with the 926 usage count is one of the books I've been referencing for my one-place study. But the links do not all go to the same page, and there should actually be a lot more unique URLs than that. Will it be possible for EditBot to preserve the page number in the URL?

Here is an example profile that references the book: Jonsen-765

---

In my url extraction I don't include the # sign. It represents the internal link on the page and has no affect with my URL validation. EditBOT can be programmed to replace also #323 as a page number like ?page=323

---

Ok awesome thanks!

---

The link without a page number and ?lang=no at the end gives a bibliographical description of the work, and an image of the front page, where you can open the book itself. The "lang=no" gives the information in Norwegian. I tested with "lang=en" which gives a description where some of the keywords are in English. If you omit the "lang" entirely, it's still intelligible - at least for a bilingual Norwegian.