Hindi searches

Do you know how Mediawiki currently handles this? We obviously don't want all spaces to be normalized to that character, and we don't want that character to be normalized to a space either (or terrible matches would be made).

No, I don’t know how it works. &# x200C; and &# x200B; definitely should not be removed or changed, but &# x200C; should be seen by the software as the equivalent of a space, and &# x200B; should be seen as the equivalent of nothing at all.