From: Verachten Bruno (Bruno.Verachten@atosorigin.com)
Date: 06/07/2001 - 08:12
Bonjour,
je rencontre quelques difficultés pour produire un fichier CSV à partir d'un
log en XML.
Les logs me donnent des temps de parsing, etc, de différents parseurs XML,
et
j'aimerais réussir à produire un fichier à importer sous Excel pour faire de
jolies
courbes... Oui, j'aurais peut-être pu les faire en SVG :-(
En attendant, j'arrive à trier le XML par parseur, mais j'aimerais réussir à
faire une moyenne des différents temps rencontrés avant de produire le
fichier
CSV.
Exemple:
source:
<REPORT>
<JVM JAVAVERSION="1.3.0_01" NAME="Java HotSpot(TM) Client VM"
VMVERSION="1.3.0_01" VENDOR="Sun Microsystems Inc." />
<ALERT MESSAGE="**Memory checking enabled, timing results may be
effected**" />
<ALERT MESSAGE="**Validation checking enabled, timing results may be
effected**" />
<PARSER TYPE="DOM">
<NAME>JDOM</NAME>
<MEMSIZE TYPE="INITIAL">225928</MEMSIZE>
<EVENT NAME="Built from parse">
<TIME_ELAPSED SCALE="ms">1973</TIME_ELAPSED>
<MEMSIZE TYPE="AVERAGE">145563</MEMSIZE>
</EVENT>
<EVENT NAME="Walked tree">
<TIME_ELAPSED SCALE="ms">201</TIME_ELAPSED>
<MEMSIZE TYPE="AVERAGE">155223</MEMSIZE>
</EVENT>
<EVENT NAME="Generated text">
<TIME_ELAPSED SCALE="ms">1052</TIME_ELAPSED>
</EVENT>
<EVENT NAME="Serialized">
<TIME_ELAPSED SCALE="ms">1191</TIME_ELAPSED>
</EVENT>
<EVENT NAME="Unserialized">
<TIME_ELAPSED SCALE="ms">2504</TIME_ELAPSED>
</EVENT>
<DESCRIPTION TYPE="DOM">
<ELEMENTS>277</ELEMENTS>
<ATTRIBUTES>443</ATTRIBUTES>
<TEXT_CHARACTERS>2272</TEXT_CHARACTERS>
<TEXT_SIZE SOURCE="21615" OUTPUT="20659" SERIALIZED="43843" />
</DESCRIPTION>
</PARSER>
<PARSER TYPE="DOM">
<NAME>dom4j</NAME>[...]
</PARSER> <PARSER TYPE="DOM">
<NAME>JDOM</NAME>
<MEMSIZE TYPE="INITIAL">593672</MEMSIZE>
<EVENT NAME="Built from parse">
<TIME_ELAPSED SCALE="ms">440</TIME_ELAPSED>
<MEMSIZE TYPE="AVERAGE">150510</MEMSIZE>
</EVENT>
<EVENT NAME="Walked tree">
<TIME_ELAPSED SCALE="ms">110</TIME_ELAPSED>
<MEMSIZE TYPE="AVERAGE">158478</MEMSIZE>
</EVENT>
<EVENT NAME="Generated text">
<TIME_ELAPSED SCALE="ms">311</TIME_ELAPSED>
</EVENT>
<EVENT NAME="Serialized">
<TIME_ELAPSED SCALE="ms">731</TIME_ELAPSED>
</EVENT>
<EVENT NAME="Unserialized">
<TIME_ELAPSED SCALE="ms">741</TIME_ELAPSED>
</EVENT>
<DESCRIPTION TYPE="DOM">
<ELEMENTS>277</ELEMENTS>
<ATTRIBUTES>443</ATTRIBUTES>
<TEXT_CHARACTERS>2272</TEXT_CHARACTERS>
<TEXT_SIZE SOURCE="21615" OUTPUT="20659" SERIALIZED="43843" />
</DESCRIPTION>
</PARSER>
</REPORT>
Comme on peut le voir, il y a plusieurs fois le même parseurs
(itération différents). J'aimerais réussir à faire la moyenne des
différents temps, tailles mémoires, etc, pour chaque parseur...
mais je bloque.
XSL actuelle:
<xsl:template match="/REPORT">
<xsl:text>Parser:</xsl:text>
<xsl:text>Size:</xsl:text>
<xsl:text>Built:</xsl:text>
<xsl:text>Walked:</xsl:text>
<xsl:text>Generated:</xsl:text>
<xsl:text>Serialized:</xsl:text>
<xsl:text>Unserialized:</xsl:text>
<xsl:apply-templates>
<xsl:sort select="NAME"/>
</xsl:apply-templates>
</xsl:template>
<xsl:template match="PARSER">
<!--xsl:value-of select="@TYPE" />
<xsl:text>,</xsl:text-->
<xsl:apply-templates select="NAME"/>
<xsl:text>,</xsl:text>
<xsl:apply-templates select="MEMSIZE"/>
<xsl:text>,</xsl:text>
<xsl:apply-templates/>
<!--xsl:apply-templates select="EVENT"/-->
<xsl:text>
</xsl:text>
</xsl:template>
<xsl:template match="PARSER/MEMSIZE">
<xsl:apply-templates/>
</xsl:template>
<xsl:template match="EVENT">
<xsl:text>,</xsl:text>
<xsl:apply-templates select="TIME_ELAPSED"/>
<xsl:text>,</xsl:text>
</xsl:template>
Ce qui me génère quelque chose du genre:
Parser:Size:Built:Walked:Generated:Serialized:Unserialized:
Crimson DOM,223424,1302,20,520,
Crimson DOM,569424,460,10,320,
Crimson DOM,574584,430,10,340,
Crimson DOM,594224,421,10,351,
Crimson DOM,598000,440,20,330,
Crimson DOM,580456,421,10,321,
Crimson DOM,618368,441,10,321,
Crimson DOM,578672,430,10,340,
Crimson DOM,590344,430,10,320,
Crimson DOM,578800,451,10,371,
dom4j,216496,912,31,521,
dom4j,585488,401,10,410,
dom4j,581944,400,20,360,
dom4j,595864,390,20,370,
dom4j,601432,371,20,1322,
dom4j,580600,391,10,381,
dom4j,589648,381,20,401,
dom4j,584176,371,20,440,
dom4j,567040,430,10,441,
dom4j,583176,391,10,451,
Electric XML,578296,541,40,300,561,611,
Electric XML,567504,280,10,201,550,551,
Electric XML,568552,450,10,331,621,671,
Electric XML,570592,291,0,200,581,581,
Electric XML,568144,270,10,210,521,551,
Electric XML,562896,280,0,190,521,561,
Electric XML,562960,290,10,211,511,560,
Electric XML,582480,270,10,190,531,541,
Electric XML,562888,290,10,210,511,551,
Electric XML,565328,281,0,190,561,561,
JDOM,225928,1973,201,1052,1191,2504,
JDOM,593672,440,110,311,731,741,
JDOM,604344,451,151,331,701,761,
JDOM,564512,520,100,361,831,781,
JDOM,591400,561,120,320,781,821,
JDOM,570128,461,120,330,681,761,
JDOM,585072,501,120,350,791,801,
JDOM,578288,461,140,311,721,1041,
JDOM,569984,460,100,341,771,811,
JDOM,586520,611,151,331,701,871,
Xerces DOM base,223288,2724,30,491,2333,1753,
Xerces DOM base,577544,491,10,190,1302,1202,
Xerces DOM base,590552,491,10,160,1332,1152,
Xerces DOM base,574136,481,10,181,1362,1231,
Xerces DOM base,574520,1773,10,681,5027,1132,
Xerces DOM base,618000,471,10,180,1262,1151,
Xerces DOM base,587872,450,10,170,1422,1222,
Xerces DOM base,592832,1523,21,160,1282,1202,
Xerces DOM base,589360,490,10,170,1282,1132,
Xerces DOM base,567912,490,10,320,4367,4196,
Xerces DOM deferred,417216,821,290,430,1332,1172,
Xerces DOM deferred,444296,330,190,290,1352,1192,
Xerces DOM deferred,428672,300,230,300,1933,2013,
Xerces DOM deferred,437384,300,180,291,1432,1211,
Xerces DOM deferred,418600,321,181,321,1322,1131,
Xerces DOM deferred,436496,351,180,261,1372,1101,
Xerces DOM deferred,413368,301,180,280,1522,1172,
Xerces DOM deferred,413536,421,181,291,1332,1131,
Xerces DOM deferred,423192,331,210,370,1603,1141,
Xerces DOM deferred,423280,942,191,271,1352,1121,
Je me répète, mais je voudrais donc faire la moyenne pour
chaque parseur de chaque colonne.
Par où commencer?
Merci.
B.Verachten.
--
Devenez redacteur <XML>fr et contribuez au developpement du
xml francophone (http://xmlfr.org/infos/redacteurs/) !
Liste de diffusion "xml-tech@xmlfr.org" (http://xmlfr.org).
Cette liste est a votre disposition pour discuter en francais de
tout sujet technique lie a XML.
Pour resilier votre abonnement, envoyez un message contenant
la commande "unsubscribe" a xml-tech-request@xmlfr.org
(mailto:xml-tech-request@xmlfr.org?Subject=unsubscribe)
Archive générée par hypermail 2.1.3 le 28/06/2004 - 11:06 UTC
webmaster@xmlfr.org
|
