Joseph, It is almost identical in performance metrics (CPU and memory) to bare XML pull parsing on Android.
About 6 months ago I was looking at different parsing techniques, assuming that SAX was the fastest, but I actually found pull-parsing to be the fastest. Here are some STAX-spec compliant parsing numbers: http://www.xml.com/lpt/a/1702 and specifically some XPP3 vs SAX numbers: http://www.ximpleware.com/benchmark1.html >From what I understand of the Android platform, XPP3/MXP1 was chosen as the implementation backing the org.xmlpull APIs and includes heavy customizations for parsing the binary-XML format used a lot by Android. So any of the XPP3 numbers you see will be the ones to identify with. SJXP is the result of months of me working with hand-written STAX parsers for parsing real-world, spec-molesting feeds in RSS1, 2, ATOM and RDF format. Just saw an opportunity to simplify the entire XML parsing approach while maintaining the tight performance of pull parsing. I thought others might find it handy. Best, Riyad On Mar 10, 4:35 pm, Joseph Earl <[email protected]> wrote: > What's this like in terms of speed/memory compared to say a > SaxParser ? > > On Mar 10, 10:02 pm, Riyad <[email protected]> wrote: > > > > > > > > > Simple Java XML Parser (SJXP) 2.0 is released. > > > SJXP is a very light weight (4 class) abstraction to XML Pull Parsing > > on Android (and Java in general) that allows you to use XPath-like > > expressions in parsing without needing to worry about while loops, > > event handling, try-catch blocks, etc. > > > A quick example of parsing story links from an RSS feed would look > > like this: > > ================= > > IRule linkRule = new DefaultRule(Type.CHARACTER, "/rss/channel/item/ > > link") { > > @Override > > public void handleParsedCharacters(XMLParser parser, String > > text, Object userObject) { > > // Also store the link, or something equivalently > > fancy > > } > > > } > > > XMLParser parser = new XMLParser(linkRule); > > parser.parse(xmlInputStream); > > ================= > > > You just specify a type for the rule, provide a path in the XML you > > want parsed and then override 1 of 3 default no-op handlers and then > > you can just sit around waiting for data to show up at your doorstep. > > Namespaces are also supported, please see the project page for a lot > > more documentation, examples and benchmarks. > > > Performance + Ease of Use is the big win and the design focus of SJXP. > > > New and Noteworthy > > =============== > > * Memory usage and CPU usage are a magnitude times smaller than what > > they were in 1.x series. Spent some quality time with HPROF finding > > and removing hotspots. > > * A new TAG type of rule is supported if you are just analyzing XML > > metrics and don't want the overhead of parsing data out of it. > > * Support for user-object passthrough was added to make it easier to > > get at data stores inside of the handlers. > > > Project:http://www.thebuzzmedia.com/software/simple-java-xml-parser-sjxp/ > > Download:https://github.com/downloads/thebuzzmedia/simple-java-xml-parser/sjxp... > > > SJXP is licensed under the Apache 2 license. > > > Good or bad feedback is appreciated! -- You received this message because you are subscribed to the Google Groups "Android Developers" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/android-developers?hl=en
