collection('i_do_not_exist')/users/user[@name=()] is () even if i_do_not_exist does not exit
Hi, Either I found an interesting new XQuery feature or a bug: collection('i_do_not_exist')/some/path[@attr=()] retruns the empty sequence whereas collection('i_do_not_exist')/some/path throws an error err:FODC0002 so does collection('i_do_not_exist')/some/path[@attr="i_am_an_attr"] Am I missing something? Best regards, -- Mag. Ing. Omar Siam Austrian Center for Digital Humanities and Cultural Heritage Österreichische Akademie der Wissenschaften | Austrian Academy of Sciences Stellvertretende Behindertenvertrauensperson | Deputy representative for disabled persons Bäckerstraße 13, 1010 Wien, Österreich | Vienna, Austria T: +43 1 51581-7295 omar.siam@oeaw.ac.at | www.oeaw.ac.at/acdh
Hi again, Omar, The behavior may be surprising, but it's valid: If it can be detected at compile time that an expression will never yield results, it can be simplified and rewritten to an empty sequence, no matter if a referenced database may possibly exist at runtime. Hope this helps, Christian Omar Siam <Omar.Siam@oeaw.ac.at> schrieb am Do., 23. März 2023, 14:49:
Hi,
Either I found an interesting new XQuery feature or a bug:
collection('i_do_not_exist')/some/path[@attr=()] retruns the empty sequence whereas
collection('i_do_not_exist')/some/path throws an error err:FODC0002 so does collection('i_do_not_exist')/ome/path[@attr="i_am_an_attr"]
Am I missing something?
Best regards,
-- Mag. Ing. Omar Siam Austrian Center for Digital Humanities and Cultural Heritage Österreichische Akademie der Wissenschaften | Austrian Academy of Sciences Stellvertretende Behindertenvertrauensperson | Deputy representative for disabled persons Bäckerstraße 13, 1010 Wien, Österreich | Vienna, Austria T: +43 1 51581-7295 omar.siam@oeaw.ac.at | www.oeaw.ac.at/acdh
Well it is definitly an optimization that is new to BaseX 10. But I think I found a way to default to "" where I just passed () from a variable before. Am 23.03.2023 um 15:00 schrieb Christian Grün:
Hi again, Omar,
The behavior may be surprising, but it's valid: If it can be detected at compile time that an expression will never yield results, it can be simplified and rewritten to an empty sequence, no matter if a referenced database may possibly exist at runtime.
Hope this helps, Christian
Omar Siam <Omar.Siam@oeaw.ac.at> schrieb am Do., 23. März 2023, 14:49:
Hi,
Either I found an interesting new XQuery feature or a bug:
collection('i_do_not_exist')/some/path[@attr=()] retruns the empty sequence whereas
collection('i_do_not_exist')/some/path throws an error err:FODC0002 so does collection('i_do_not_exist')/ome/path[@attr="i_am_an_attr"]
Am I missing something?
Best regards,
-- Mag. Ing. Omar Siam Austrian Center for Digital Humanities and Cultural Heritage Österreichische Akademie der Wissenschaften | Austrian Academy of Sciences Stellvertretende Behindertenvertrauensperson | Deputy representative for disabled persons Bäckerstraße 13, 1010 Wien, Österreich | Vienna, Austria T: +43 1 51581-7295 omar.siam@oeaw.ac.at | www.oeaw.ac.at/acdh <http://www.oeaw.ac.at/acdh>
-- Mag. Ing. Omar Siam Austrian Center for Digital Humanities and Cultural Heritage Österreichische Akademie der Wissenschaften | Austrian Academy of Sciences Stellvertretende Behindertenvertrauensperson | Deputy representative for disabled persons Bäckerstraße 13, 1010 Wien, Österreich | Vienna, Austria T: +43 1 51581-7295 omar.siam@oeaw.ac.at |www.oeaw.ac.at/acdh
Well it is definitly an optimization that is new to BaseX 10. But I think I found a way to default to "" where I just passed () from a variable before.
True. With BaseX 10, XQuery compilation was split up into a static (logical) and a dynamic (physical) step to allow for a more fine granular locking. In the static compilation, your expression will be rewritten as follows: collection('i_do_not_exist')/users/user[@name=()] →collection('i_do_not_exist')/users/user[false()] →collection('i_do_not_exist')/users/() →collection('i_do_not_exist')/() →()
participants (2)
-
Christian Grün -
Omar Siam