Improved search

I suggest adding support for accentuated characters, such as most other search engines. That is, Martinez would search Martínez, Martinez, Martïnez, etc. I have a lot of articles from the same authors, but sometimes they are accentuated and others no. And it is a big problem when you have to use a keyboard layout with no easy access to accentuation combinations.

It would be a great improvement that the search would accept wildcards.

Thank you for this great product. Good work.
Felipe
  • I think that this request makes a lot of sense.
  • definitely agree too - I get caught out by this quite regularly as well.
  • agree - it's not the first time this has been brought up, either.
    I guess the way this should ideally work is that if you include diacrits in your search it's explicit - i.e. searching for Martínez only gives you Martínez - but using regular letters will return all relevant diacríts - i.e. search for Martinez will return Martínez, Martinez, Martïnez etc.

    If that's too complicated, ignoring diacrits entirely for searches would work, too. I don't think this should be controversial, anyone should feel free to add an issue over at github.
  • edited April 3, 2012
    This is an interesting problem. What of languages that use characters that when converted become two characters? (I'm using an iPad so cannot easily type those characters.) For example, the beta-like German symbol for doubled "s" or the conversion of a single umlaut-ed character to two vowels. Also, there are the authors with names that use non-Latin characters when converted can have several different Latin spellings.

    edit:
    I solve this search problem, in part, for my SafetyLit database search by allowing each auther name to have a name alias for each spelling varient. Keeping that current is a lot of hand-work -- 340,000 records with 1000+ added each week. With a Zotero library, even a very large one, less user effort would be required. Zotero would need to add db table to allow users to list author spelling equivalences. The db structural changes might be a lot of work for developers. This added complexity for the purpose of an author name search might not be welcome to most users.
  • I don't think we should over-think this - separate name versions make sense for a professionally maintained database, I don't think they make sense for Zotero users.

    Dan would have to say, but I imagine the way to go would be to just use some type of existing collection of conversions - Zotero already has the removeDiacrits (or whatever it's called exactly)function to handle this. If that only works for 98% of all cases so be it - it'd still be a huge improvement over the status quo.
  • @adamsmith I agree with both of your points. I was trying to point out that trying to implement a perfect name search system would be a great deal of trouble for users and developers alike.

    What the world needs is an author registry. But, if the existing attempts to do this are any indication, I don't think that is likely to happen in our lifetimes.
  • I agree with the suggestion, but a search and replace function (something that several people have wished for since a long time) would solve at least a large part of the problem. You would simply search for all Martínez and then replace them with Martinez (or the other way round) and you would have uniform names (and other terms).
Sign In or Register to comment.