Received: from hostbureau.hostbureau.net (hostbureau.hostbureau.net [209.150.128.194] (may be forged)) by mail.postgresql.org (8.11.1/8.11.1) with ESMTP id eBFCF1096274 for ; Fri, 15 Dec 2000 07:15:01 -0500 (EST) (envelope-from mig@utdt.edu) Received: from ant.utdt (acad.utdt.edu [200.41.62.131]) by hostbureau.hostbureau.net (8.8.5/8.8.5) with SMTP id GAA15127; Fri, 15 Dec 2000 06:14:12 -0600 From: miguel sofer Date: Fri, 15 Dec 2000 12:13:08 GMT Message-ID: <20001215.12130800@ant.utdt> Subject: Re: How to represent a tree-structure in a relationaldatabase To: Alvar Freude CC: pgsql-sql@postgresql.org In-Reply-To: <3A398526.3C491E91@merz-akademie.de> References: <3A379A5F.97FFAF01@joerdens.de> <20001213.19440200@ant.utdt> <3A398526.3C491E91@merz-akademie.de> X-Mailer: Mozilla/3.0 (compatible; StarOffice/5.2;Linux) X-Priority: 3 (Normal) MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by mail.postgresql.org id eBFCF2096275 X-Archive-Number: 200012/162 X-Sequence-Number: 496 > > I once started writing a small paper on this subject; it is still in a > > rather preliminary state. > > > > You can download the draft (and some ill documented code, 53kB) from > > http://www.utdt.edu/~mig/sql-trees > ah, this looks very, very nice! Thanks. > on page 5ff you describe the Postgres implementation, but the URL (page > 5 bottom) is't complete -- can i find the files somewhere? > Included is a "tree_general.sql", but this seems not to be complete and > not the same version as the ps-file (First draft, may 6, 2000): in the > draft there is written about an base 160 encoding, tree_general.sql uses > base 159 encoding ;) Sorry, I never got around to completing this, or thinking any further. My other files are definitely not in a usable state right now. I hope to be able to improve things over the (southern) summer holidays, so there may be news soon - but do not hold your breadth! I can't remember why I switched from base 160 to base 159; my guess now is that I got confused at coding time between the base and the maximal number (base-1): ie, it may be a mistake. > What's against using all characters >= 32, excluding special characters > with special meaning in LIKE and regexps? With base 208 encoding it's > possible to have 43264 elements on each level. Nothing, I guess. I probably got some kind of "start counting at zero" blockage when I started, and never really looked back on it, my shame. Hey, I told you it was rather preliminary ... Thanks for pointing it out. > i guess, with base 160 encoding there might be a problem: if postgres is > compiled with --enable-locale (e.g. for german umlauts), the ordering > isn't according to the ASCII number of the character, so for this > purpose it's needed to build the encoding table according to the locate > settings. Or simply sort it according the locale settings. Yes indeed; never thought about that one. Cheers Miguel