Change Details

Similar to T117763 there could be a super trivial guessing game going on in the siteselector (client-side JavaScript). When you paste a full URL (for example https://de.wikipedia.org/wiki/Heinrich_Kaminski) and this gets you an empty result set, a super-trivial regular expression like `/\W(\w+)/` could extract the subdomain and make a second guess based on that. This regex searches for the first word that is preceded by a non-word character and would find the word "en" in `(en)`, `//en.wikipedia.org` and `foo en bar`.

Similar to T117763 there could be a super trivial guessing game going on in the siteselector (client-side JavaScript). When you paste a full URL (for example https://de.wikipedia.org/wiki/Heinrich_Kaminski) and this gets you an empty result set, a super-trivial regular expression like `/\W(\w+)/` could extract the subdomain and make a second guess based on that. This regex searches for the first word that is preceded by a non-word character and would find the word "en" in `(en)`, `//en.wikipedia.org` and other strings like `foo en bar`. This could be a first step towards {T49983} (also see the parent task T112241).

Similar to T117763 there could be a super trivial guessing game going on in the siteselector (client-side JavaScript). When you paste a full URL (for example https://de.wikipedia.org/wiki/Heinrich_Kaminski) and this gets you an empty result set, a super-trivial regular expression like `/\W(\w+)/` could extract the subdomain and make a second guess based on that. This regex searches for the first word that is preceded by a non-word character and would find the word "en" in `(en)`, `//en.wikipedia.org` and other strings like `foo en bar`. This could be a first step towards {T49983} (also see the parent task T112241).