From: Jody G. <jga...@re...> - 2007-07-23 16:22:19
|
Thanks for bringing this to our attentions, I echo Adrian's words - great work and welcome! Here is the list... > > * The parser uses XppReader, which is part of the XStream^ > <http://xstream.codehaus.org/>project, and which in turn uses > XPP^ <http://www.extreme.indiana.edu/xgws/xsoap/xpp/>, a pull > parser. [/pro: have experience in it, con: external dependency/] > Justin is very interested in the pull parser idea; you may also be interested in his binding framework thing - it streams data (so you are not limited by size). He has considered switching from SAX to a pull parser in the past. If you are familiar with XppReader you guys should probably grab an IRC session and talk shop. The best I can do is point you towards these docs (http://docs.codehaus.org/display/GEOTDOC/XML+Developers+Guide). > > * The data container bean currently stores the geometry, name, > description, comment and date values from the file. [/these were > just enough for the other project, and as I saw, thees would > almost always be enough/] > If it is what you got then it is fine. You may want to consider a "bounds" for the entire file ... if the information is in a header? > > * The DataStore when created, creates two FeatureClasses, which > describe the same features every time: a point class, and a > track class. > Okay. > > * The DataStore parses the GPX file in the constructor, so if a > parser error occurs, the creation of the DataStore itself fails. > [/pro: prevents deferred errors/] > Sounds fine; sometimes we just parse the header in the constructor - and put off parsing the rest until the user actually calls the iterator for the first time. > > * The FeatureReader implementations convert the data beans to > Feature instances. A waypoint's geometry becomes a Point with > lat/lon/elevation, and a track's becomes a MultiLineString with > lat/lon/elevation coordinates. > Good. > > * The structure of the module is minimalistic in the sense that it > depends highly on the default implementations of the interfaces. > [/for the sake of fast development, and that is what default > implementations is there for. As this module assumes to work > with small data, this should cause no performance issues./] > Great. All the best, Jody Adrian Custer wrote: > Hey Peter, > > It looks good. > > Justin, Jody, anyone else, > > The bottom of the page has some design considerations. Perhaps you could > take 10 minutes to add some comments and pointers. It seems like Peter > has gotten amazingly far in getting things hooked up and now he could > use your help. > > --adrian > > On Fri, 2007-07-20 at 23:43 -0500, Bolla Péter wrote: > >> Hello, >> >> Thank you, for the cleanup, it certainly looks better. :) >> >> Also I reviewed it again, and separated the current state of the module >> and the planned developments, and wrote some words on design considerations. >> >> Comments and ideas still welcome! >> >> Best, >> Peter >> >> >> Adrian Custer wrote: >> >>> Hello, >>> >>> Glad to see you are getting started. I will be cool to have your module >>> working in Geotools. >>> >>> I made some edits to the layout to make the page, I hope, a bit clearer. >>> >>> I would encourage you to re-visit the last part where you discuss the >>> design decisions you are considering. Specifically, perhaps you can talk >>> about the tradeoffs or lack of knowledge that led to your decisions. >>> Also drop the "I was thinking" kind of language in favour of what you >>> were thinking about. >>> >>> cheers and, if I haven't said so before, welcome to the project, >>> >>> --adrian >>> >>> >>> On Thu, 2007-07-19 at 20:05 -0500, Bolla Péter wrote: >>> >>>> Hi, >>>> >>>> I've created the RnD page on my GPX module. You can read it here: >>>> http://docs.codehaus.org/display/GEOTOOLS/GPX+module >>>> >>>> Any comments and ideas are welcome! >>>> >>>> Best, >>>> Peter >>>> >>>> >>>> ------------------------------------------------------------------------- >>>> This SF.net email is sponsored by: Microsoft >>>> Defy all challenges. Microsoft(R) Visual Studio 2005. >>>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >>>> _______________________________________________ >>>> Geotools-devel mailing list >>>> Geo...@li... >>>> https://lists.sourceforge.net/lists/listinfo/geotools-devel >>>> >>> > > |