Subject: Re: [xsl] question on EXSLT data partitioning From: Hermann Stamm-Wilbrandt <STAMMW@xxxxxxxxxx> Date: Thu, 21 Oct 2010 12:56:13 +0200 |
Erwan, I just tested your new solution, and it works for xsltproc as well as saxon. I added -t -replay:1000 to both executions (as Michael suggested, see below) and got the same result for the solution provided by Martin Honnen and your xsl:key solution: *** Average execution time over 1000 runs: 2ms Probably this is to be expected since the underlying algorithm is the same for both apporaches (iterate over input and process those elements of the current "block"), there is just the difference on how the elements in current "block" are determined. I doubt that bigger input files/bigger block sizes will show significant differences since the input/output processing will dominate the block determination time ... I also converted my recursive func:function() recursive EXSLT solution to xsl:function XSLT 2.0 and got this worse result: *** Average execution time over 1000 runs: 3ms [stammw@br8ggx73 Erwan]$ java -jar saxon9he.jar -t -xsl:part.xsl -s:data.xml -versionmsg:off Saxon-HE 9.2.0.3J from Saxonica Java version 1.6.0 Stylesheet compilation time: 537 milliseconds Processing file:/Erwan/data.xml Using parser org.apache.xerces.jaxp.SAXParserImpl$JAXPSAXParser Building tree for file:/Erwan/data.xml using class net.sf.saxon.tinytree.TinyBuilder Tree built in 22 milliseconds Tree size: 30 nodes, 16 characters, 0 attributes Loading net.sf.saxon.event.MessageEmitter <?xml version="1.0" encoding="UTF-8"?> <r> <c> <b>11</b> <b>12</b> <b>13</b> </c> <c> <b>14</b> <b>15</b> <b>16</b> </c> <c> <b>17</b> <b>18</b> </c> </r>Execution time: 63 milliseconds Memory used: 4562936 NamePool contents: 15 entries in 15 chains. 6 prefixes, 6 URIs [stammw@br8ggx73 Erwan]$ java -jar saxon9he.jar -t -xsl:ent2.xsl -s:ent.xml -versionmsg:off Saxon-HE 9.2.0.3J from Saxonica Java version 1.6.0 Stylesheet compilation time: 513 milliseconds Processing file:/Erwan/ent.xml Using parser org.apache.xerces.jaxp.SAXParserImpl$JAXPSAXParser Building tree for file:/Erwan/ent.xml using class net.sf.saxon.tinytree.TinyBuilder Tree built in 18 milliseconds Tree size: 41 nodes, 13 characters, 0 attributes Loading net.sf.saxon.event.MessageEmitter <all><arguments><argument>1</argument><argument>2</argument><argument>3</argu ment></arguments><arguments><argument>4</argument><argument>5</argument><argu ment>6</argument></arguments><arguments><argument>7</argument><argument>8</ar gument><argument>9</argument></arguments><arguments><argument>10</argument><a rgument>11</argument></arguments></all> Execution time: 54 milliseconds Memory used: 4621896 NamePool contents: 13 entries in 13 chains. 6 prefixes, 6 URIs [stammw@br8ggx73 Erwan]$ Mit besten Gruessen / Best wishes, Hermann Stamm-Wilbrandt Developer, XML Compiler, L3 Fixpack team lead WebSphere DataPower SOA Appliances ---------------------------------------------------------------------- IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter Geschaeftsfuehrung: Dirk Wittkopp Sitz der Gesellschaft: Boeblingen Registergericht: Amtsgericht Stuttgart, HRB 243294 From: Erwan AMOUREUX <erwan.amoureux@xxxxxxxxxxxxxxxxxxxxxxx> To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx Date: 10/21/2010 11:35 AM Subject: Re: [xsl] question on EXSLT data partitioning I had a good look at my xslt sheet and , i only use position in match key or with other context in use. Sorry for that, my only excuse it's a very bad cold... But could you test that, please ? My xml example <?xml version="1.0" encoding="UTF-8"?> <?xml-stylesheet type="text/xsl" href="D:\Mes documents\testxml \positioncle.xslt"?> <a> <b>11</b> <b>12</b> <b>13</b> <b>14</b> <b>15</b> <b>16</b> <b>17</b> <b>18</b> </a> My XSLT example <?xml version="1.0" encoding="UTF-8"?> <xsl:stylesheet version="1.0" xmlns:xsl=" http://www.w3.org/1999/XSL/Transform"> <xsl:output method="xml" version="1.0" encoding="UTF-8" indent="yes"/> <xsl:key name="part" match="b[(position() - 1) mod 3!=0]" use="generate-id (../b[(position() - 1) mod 3=0 and ((following-sibling::b)[position ()<3])[generate-id(.)=generate-id(current())]])"/> <xsl:template match="/"> <r> <xsl:for-each select="*/b[(position() - 1) mod 3=0]"> <c> <xsl:apply-templates select=".|key('part',generate-id(.))"/> </c> </xsl:for-each> </r> </xsl:template> <xsl:template match="b"> <xsl:copy-of select="."/> </xsl:template> </xsl:stylesheet> I think it's too complicated and maybe it can be false but i'm curious to know what you think about idea ----- Message d'origine ----- De : "Erwan AMOUREUX" <erwan.amoureux@xxxxxxxxxxxxxxxxxxxxxxx> @ : <xsl-list@xxxxxxxxxxxxxxxxxxxxxx> Envoyi : jeudi 21 octobre 2010 09:36 Objet : Re: [xsl] question on EXSLT data partitioning My mistake! sorry ----- Message d'origine ----- De : "Michael Kay" <mike@xxxxxxxxxxxx> @ : <xsl-list@xxxxxxxxxxxxxxxxxxxxxx> Envoyi : mercredi 20 octobre 2010 16:24 Objet : Re: [xsl] question on EXSLT data partitioning > >> For me it's logical >> When you use global variable, current node is root, so with the key ,it's the same thing and position is calculated with the root as initial context >> > > Whether it's logical or not, the specification is quite clear. In XSLT 1.0: > > the expression specified in the use attribute of the xsl:key element is > evaluated with x as the current node and with a node list containing > just x as the current node list > > So position() in the use expression returns 1. > > Michael Kay > Saxonica
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: [xsl] question on EXSLT data pa, Erwan AMOUREUX | Thread | Re: [xsl] question on EXSLT data pa, Erwan AMOUREUX |
Re: [xsl] question on EXSLT data pa, Erwan AMOUREUX | Date | [xsl] XSLT2.0 Processors, Neil Owens |
Month |