Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
J, J.Pietschmann wrote: Clay Leeds wrote: Not trying to belabor a point, just trying to get understanding of how this all works. I would think that JVM only has to start once during a processing instance. This is correct. The start and warp-up overhead accounts for the difference between the 2s for page rendering and a few seconds more of total elapsed time. However, is the long delay you experienced with the commented file reproducible, or did you perform the experiment just once? On a multitasking system there can be quite a bit of jitter, in particular windows occasionally locks the whole machine for some non-trivial time (tens of seconds to minutes) when it feels like reorganizing its in-memory management structures or if it unloads no longer needed DLLs from memory in order to free swap space. J.Pietschmann I've found this to be re-producible with one file (it happens every time with that file). I haven't tried it with other files, although I did test w/ .4 & .5rc. You should be able to take a large XML INPUT file, comment out a large portion of the FLOW that gets processed with a FOR-EACH statement, process it with an XSL-FO file & get similar results. If you'd like, I could probably send you my FO & XML test files offline. They are 240 KB (246,218 bytes) & 109 KB (112,076 bytes), respectively. Web Maestro Clay -- Clay Leeds - [EMAIL PROTECTED] Web Developer - Medata, Inc. - http://www.medata.com PGP Public Key: https://mail.medata.com/pgp/cleeds.asc - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Clay Leeds wrote: Not trying to belabor a point, just trying to get understanding of how this all works. I would think that JVM only has to start once during a processing instance. This is correct. The start and warp-up overhead accounts for the difference between the 2s for page rendering and a few seconds more of total elapsed time. However, is the long delay you experienced with the commented file reproducible, or did you perform the experiment just once? On a multitasking system there can be quite a bit of jitter, in particular windows occasionally locks the whole machine for some non-trivial time (tens of seconds to minutes) when it feels like reorganizing its in-memory management structures or if it unloads no longer needed DLLs from memory in order to free swap space. J.Pietschmann - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
If you transform your XML to PDF using fop.bat you're starting one JVM. If you're calling the Xalan command line to generate the XSL-FO and then you invoke fop.bat with -fo to get the PDF, then you're starting two JVMs. It is very strange that it seems to be faster to start two VMs than to do the two transformations in one run (connected through SAX events). But this might be some strange side-effect from JVM-starting/class-loading. I'm sure this would look differently if you had a server app that can do multiple transformation runs without having to restart the JVM each time. The whole thing should (I'm getting cautious) be a lot faster by the way. Anyone there with Java knowledge and some time and curiosity who wants to analyze this a little more? On 24.01.2003 16:29:22 Clay Leeds wrote: > Jeremias Maerki wrote: > > As Jörg said, you have to take JVM startup into account. That makes the > > whole thing relatively unpredictable. Measuring times from batch files > > is ok to get a general idea but not for investigating details. For that > > you have to write a Java program and execute the transformations several > > times to make sure all classes are loaded (a lengthy process) and the > > JVM had a chance to compile some Java code into machine code for faster > > execution (HotSpot JIT compiler). > > > > On 23.01.2003 23:34:03 Clay Leeds wrote: > > > >>That's an intriguing idea. Assuming that it would help trouble-shoot > >>problems like this, that might be good for this type of testing. Also, I > >>noticed a difference in the amount of time, when FOP transformed & > >>rendered the PDF file from start to finish (01:05 for 0.20.4), versus > >>when xalan-2.0.1 transformed to XML, and then FOP transformed what was > >>essentially an FO file to PDF (47030ms+1843ms=48873ms). Any idea where > >>the additional 15+ seconds went? > > > > Jeremias Maerki > > Not trying to belabor a point, just trying to get understanding of how > this all works. I would think that JVM only has to start once during a > processing instance. I wouldn't think that it would matter if the intial > factor is 1) a *.bat file that initiates FOP (which runs XALAN, XERCES, > etc.); or 2) if it's a *.bat file that initiates FOP and then another > *.bat file that intiates FOP from the generated FO file. The way I > figure it, option 1 launches JVM once, whereas option 2 launches JVM > twice (one for each *.bat file). However, my results showed that I > would've saved 18 seconds if I used a *.bat file which launched my two > *.bat files. I realize that my results were anything but scientific, as > I eyeballed the Windows system clock instead of using cygwin's TIME()). Jeremias Maerki - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Jeremias, Jeremias Maerki wrote: As Jörg said, you have to take JVM startup into account. That makes the whole thing relatively unpredictable. Measuring times from batch files is ok to get a general idea but not for investigating details. For that you have to write a Java program and execute the transformations several times to make sure all classes are loaded (a lengthy process) and the JVM had a chance to compile some Java code into machine code for faster execution (HotSpot JIT compiler). On 23.01.2003 23:34:03 Clay Leeds wrote: That's an intriguing idea. Assuming that it would help trouble-shoot problems like this, that might be good for this type of testing. Also, I noticed a difference in the amount of time, when FOP transformed & rendered the PDF file from start to finish (01:05 for 0.20.4), versus when xalan-2.0.1 transformed to XML, and then FOP transformed what was essentially an FO file to PDF (47030ms+1843ms=48873ms). Any idea where the additional 15+ seconds went? Jeremias Maerki Not trying to belabor a point, just trying to get understanding of how this all works. I would think that JVM only has to start once during a processing instance. I wouldn't think that it would matter if the intial factor is 1) a *.bat file that initiates FOP (which runs XALAN, XERCES, etc.); or 2) if it's a *.bat file that initiates FOP and then another *.bat file that intiates FOP from the generated FO file. The way I figure it, option 1 launches JVM once, whereas option 2 launches JVM twice (one for each *.bat file). However, my results showed that I would've saved 18 seconds if I used a *.bat file which launched my two *.bat files. I realize that my results were anything but scientific, as I eyeballed the Windows system clock instead of using cygwin's TIME()). -- Clay Leeds - [EMAIL PROTECTED] Web Developer - Medata, Inc. - http://www.medata.com PGP Public Key: https://mail.medata.com/pgp/cleeds.asc - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
As Jörg said, you have to take JVM startup into account. That makes the whole thing relatively unpredictable. Measuring times from batch files is ok to get a general idea but not for investigating details. For that you have to write a Java program and execute the transformations several times to make sure all classes are loaded (a lengthy process) and the JVM had a chance to compile some Java code into machine code for faster execution (HotSpot JIT compiler). On 23.01.2003 23:34:03 Clay Leeds wrote: > That's an intriguing idea. Assuming that it would help trouble-shoot > problems like this, that might be good for this type of testing. Also, I > noticed a difference in the amount of time, when FOP transformed & > rendered the PDF file from start to finish (01:05 for 0.20.4), versus > when xalan-2.0.1 transformed to XML, and then FOP transformed what was > essentially an FO file to PDF (47030ms+1843ms=48873ms). Any idea where > the additional 15+ seconds went? Jeremias Maerki - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
RE: FOP 0.20.4 & 5rc timing w large COMMENT sections
What you and Oleg say makes perfect sense. I was not aware this could be an issue that would actually affect the rendered output, but it does. I won't be using that option anymore. Pat - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Patrick Dean Rusk wrote: Just today I've finally diagnosed a strangeness in my program to the fact that using FOP to transform gives a different result sometimes than doing the transform manually and following it with a FOP run on the .fo file. In my case, the difference was from having the following line in my XSL file: This is respected when doing the manual transform, but the FOP run seems to treat it as "no". You have to understand that XSLT can be a multiphase process 1. Parsing the source XML into a tree 2. Transforming the source tree into a result tree 3. Serialize the result tree. If you run the transformation as integral part of the FO processing, there is no serialization of the transformation result, the FO processor takes it input directly from the result tree of the transformation. Consequently all options regarding result serialization are ignored. This includes everything in the xsl:output element (you could have specified mathod="text" :-) ) as well as disabling output escaping. J.Pietschmann - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Patrick Dean Rusk wrote: I think it would be preferable to have a FOP option to output the intermediate FO file that it uses in its processing, similar to the way many compilers can output some of their intermediate files. It's well-known feature request, see http://nagoya.apache.org/bugzilla/show_bug.cgi?id=14351. Just today I've finally diagnosed a strangeness in my program to the fact that using FOP to transform gives a different result sometimes than doing the transform manually and following it with a FOP run on the .fo file. In my case, the difference was from having the following line in my XSL file: This is respected when doing the manual transform, but the FOP run seems to treat it as "no". Sure! There is no room for indentaion in SAX events stream and as there is no serialization altogether the whole element is ignored (rememeber - it's just a hint according to the spec). And after all it's not FOP, but a XSLT processor. -- Oleg Tkachenko eXperanto team Multiconn Technologies, Israel - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
RE: FOP 0.20.4 & 5rc timing w large COMMENT sections
I think it would be preferable to have a FOP option to output the intermediate FO file that it uses in its processing, similar to the way many compilers can output some of their intermediate files. Just today I've finally diagnosed a strangeness in my program to the fact that using FOP to transform gives a different result sometimes than doing the transform manually and following it with a FOP run on the .fo file. In my case, the difference was from having the following line in my XSL file: This is respected when doing the manual transform, but the FOP run seems to treat it as "no". I'll shortly be posting an email related to this latter issue. Patrick Rusk P.S. I think that XEP provides separate files for doing the transform followed by the rendering. I found them handy. Of course, I've created some for FOP, as I'm sure most do. -Original Message- From: J.Pietschmann [mailto:[EMAIL PROTECTED]] Sent: Thursday, January 23, 2003 5:25 PM To: [EMAIL PROTECTED] Subject: Re: FOP 0.20.4 & 5rc timing w large COMMENT sections Would you think we should put a Xalan.bat/Xalan.sh file into the FOP distribution? (or xslt.bar/xslt.sh) J.Pietschmann - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED] - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
J.Pietschmann wrote: Would you think we should put a Xalan.bat/Xalan.sh file into the FOP distribution? (or xslt.bar/xslt.sh) J.Pietschmann That's an intriguing idea. Assuming that it would help trouble-shoot problems like this, that might be good for this type of testing. Also, I noticed a difference in the amount of time, when FOP transformed & rendered the PDF file from start to finish (01:05 for 0.20.4), versus when xalan-2.0.1 transformed to XML, and then FOP transformed what was essentially an FO file to PDF (47030ms+1843ms=48873ms). Any idea where the additional 15+ seconds went? -- Clay Leeds - [EMAIL PROTECTED] Web Developer - Medata, Inc. - http://www.medata.com PGP Public Key: https://mail.medata.com/pgp/cleeds.asc - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
J.Pietschmann wrote: > Would you think we should put a Xalan.bat/Xalan.sh file > into the FOP distribution? (or xslt.bar/xslt.sh) Very good idea. Probably most people create these batchfiles/shellscripts anyway. Might also lead to more people trying the transformation step separately to see where the real problem is... For FOP beginners, it's a real hurdle writing the one-liner that does the transformation alone... BTW: In any case I'd prefer the naming xalan.bat(.sh) I actually have a saxon.sh(.bat), xalan.sh *and* xslt.sh. The latter calls whichever xslt processor my project is currently using. Just my 2 cents. -- Cappelino Informationstechnologie GmbH Arnd Beißner - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Clay Leeds wrote: It took 2 seconds of rendering time compared to 65 seconds of time taken from command line [Enter] to FOP output completion. This latter time includes JVM startup, loading all the code and whatnot. Occasionally the OS is busy and it takes longer to load an executable and the data. How often did you execute the experiment? If you want to have more accurate measurements you should install cygwin and use the time command. J.Pietschmann - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Would you think we should put a Xalan.bat/Xalan.sh file into the FOP distribution? (or xslt.bar/xslt.sh) J.Pietschmann - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
I ran the XSLT transform twice (just out of curiosity), and there was about 300ms difference between the two identicalk runs. However, it is more in line with the results we were expecting. However, it was still about 15-18 seconds shorter than my previous results (diligently watching the clock). This does indeed provide some food for thought. Keeping in mind the difference between the previous times I submitted (01:05 for 0.20.4--00:58 for 0.20.5rc) I see that xalan-2.4.1 & xerces-2.2.1 have some improvements. So, where does this leave us? Should we drop a note to the xalan folks? Should we make a note in the FOP docs? Here's the output from the DOS window: === C:\Program Files\Java\fop-0.20.4>java -cp lib/xml-apis.jar;lib/xercesImpl-2.0.1. jar;lib/xalan-2.3.1.jar org.apache.xalan.xslt.Process -IN C:\xml_med7\biglinux\t est_NCWC_sm_1.1.xml -XSL C:\xml_med7\fo\xml_med7_default.fo -OUT 030123_test_xml _med7_NCWC_sm_1.1.out.xml -DIAG Transform of C:\xml_med7\biglinux\test_NCWC_sm_1.1.xml via C:\xml_med7\fo\xml_me d7_default.fo took 46671 ms C:\Program Files\Java\fop-0.20.4>java -cp lib/xml-apis.jar;lib/xercesImpl-2.0.1. jar;lib/xalan-2.3.1.jar org.apache.xalan.xslt.Process -IN C:\xml_med7\biglinux\t est_NCWC_sm_1.1.xml -XSL C:\xml_med7\fo\xml_med7_default.fo -OUT 030123_test_xml _med7_NCWC_sm_1.1.out.xml -DIAG Transform of C:\xml_med7\biglinux\test_NCWC_sm_1.1.xml via C:\xml_med7\fo\xml_me d7_default.fo took 47030 ms C:\Program Files\Java\fop-0.20.4>fop -d -fo 030123_test_xml_med7_NCWC_sm_1.1.out .xml -pdf 030123_test_XSLT_002.pdf C:\Program Files\Java\fop-0.20.4>java -cp build\fop.jar;lib\batik.jar;lib\xalan- 2.3.1.jar;lib\xercesImpl-2.0.1.jar;lib\xml-apis.jar;lib\avalon-framework-cvs-200 20315.jar;lib\logkit-1.0.jar;lib\jimi-1.0.jar org.apache.fop.apps.Fop -d -fo 030 123_test_xml_med7_NCWC_sm_1.1.out.xml -pdf 030123_test_XSLT_002.pdf[DEBUG] Input mode: [DEBUG] FO [DEBUG] fo input file: 030123_test_xml_med7_NCWC_sm_1.1.out.xml [DEBUG] Output mode: [DEBUG] pdf [DEBUG] output file: 030123_test_XSLT_002.pdf [DEBUG] OPTIONS [DEBUG] no user configuration file is used [default] [DEBUG] debug mode on [DEBUG] dump configuration [DEBUG] quiet mode on [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [DEBUG] base directory: file:/C:/Program Files/Java/fop-0.20.4/ [INFO] FOP 0.20.4 [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [INFO] building formatting object tree [DEBUG] setting up fonts [INFO] [1] [DEBUG] Last page-sequence produced 1 pages. [INFO] Parsing of document complete, stopping renderer [DEBUG] Initial heap size: 726Kb [DEBUG] Current heap size: 2191Kb [DEBUG] Total memory used: 1464Kb [DEBUG] Memory use is indicative; no GC was performed [DEBUG] These figures should not be used comparatively [DEBUG] Total time used: 1843ms [DEBUG] Pages rendered: 1 [DEBUG] Avg render time: 1843ms/page -- Clay Leeds - [EMAIL PROTECTED] Web Developer - Medata, Inc. - http://www.medata.com PGP Public Key: https://mail.medata.com/pgp/cleeds.asc - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Jeremias Maerki wrote: On 23.01.2003 22:47:20 Clay Leeds wrote: That worked great (except I had to change the version numbers for xerces (xercesImpl-2.0.1.jar) & xalan(xalan-2.3.1.jar) (which was probably expcected :-): C:\Program Files\Java\fop-0.20.4>java -cp lib/xml-apis.jar;lib/xercesImpl-2.0.1.jar;lib/xalan-2.3.1.jar org.apache.xalan.xslt.Process -IN C:\xml_med7\biglinux\test_MIWC_1.1.xml -XSL C:\xml_med7\fo\xml_med7_MIWC.fo -OUT 030123_test_xml_med7_MIWC_1.1.out.xml -DIAG Transform of C:\xml_med7\biglinux\test_MIWC_1.1.xml via C:\xml_med7\fo\xml_med7_MIWC.fo took 2813 ms I'm confused. That's only the uncommented one, right? What about the time for the commented one? Jeremias Maerki Sorry. Brainfart. I just got back from lunch and I used the wrong file... I'll re-run the test. -- Clay Leeds - [EMAIL PROTECTED] Web Developer - Medata, Inc. - http://www.medata.com PGP Public Key: https://mail.medata.com/pgp/cleeds.asc - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
On 23.01.2003 22:47:20 Clay Leeds wrote: > That worked great (except I had to change the version numbers for xerces > (xercesImpl-2.0.1.jar) & xalan(xalan-2.3.1.jar) (which was probably > expcected :-): > > C:\Program Files\Java\fop-0.20.4>java -cp > lib/xml-apis.jar;lib/xercesImpl-2.0.1.jar;lib/xalan-2.3.1.jar > org.apache.xalan.xslt.Process -IN C:\xml_med7\biglinux\test_MIWC_1.1.xml > -XSL C:\xml_med7\fo\xml_med7_MIWC.fo -OUT > 030123_test_xml_med7_MIWC_1.1.out.xml -DIAG > > > Transform of C:\xml_med7\biglinux\test_MIWC_1.1.xml via > C:\xml_med7\fo\xml_med7_MIWC.fo took 2813 ms I'm confused. That's only the uncommented one, right? What about the time for the commented one? > In the end, the following took just over 4 seconds, which means I don't > think it was xalan which took so long. Jeremias Maerki - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Jeremias Maerki wrote: On 23.01.2003 22:15:17 Clay Leeds wrote: It helps. However, I only have fop-0.20.4 (&.5rc) installed. I also notice that in the lib/ folder I've got xalan-2.3.1.jar. Will this suffice, or do I need to install the full version of xalan. Also, the current version of xalan is 2.4.1... should I try it with that? Yep. In this case do: java -cp lib/xml-apis.jar;lib/xercesImpl-2.2.1.jar;lib/xalan-2.4.1.jar org.apache.xalan.xslt.Process -IN test2.xml -XSL test2.xsl -OUT test2.out.xml -DIAG That worked great (except I had to change the version numbers for xerces (xercesImpl-2.0.1.jar) & xalan(xalan-2.3.1.jar) (which was probably expcected :-): C:\Program Files\Java\fop-0.20.4>java -cp lib/xml-apis.jar;lib/xercesImpl-2.0.1.jar;lib/xalan-2.3.1.jar org.apache.xalan.xslt.Process -IN C:\xml_med7\biglinux\test_MIWC_1.1.xml -XSL C:\xml_med7\fo\xml_med7_MIWC.fo -OUT 030123_test_xml_med7_MIWC_1.1.out.xml -DIAG Transform of C:\xml_med7\biglinux\test_MIWC_1.1.xml via C:\xml_med7\fo\xml_med7_MIWC.fo took 2813 ms C:\Program Files\Java\fop-0.20.4> That will run the XSL transformation. It assumes you're in the root directory of your FOP distribution (location of fop.bat) and you're using Windows. You may have to adjust the names of the jar files depending on the FOP version. Lastly, I want my system to match--as closely as possible--that of my clients. What steps can I take to ensure that I can regain that "parity" for testing? Well, if you use fop.bat to do the testing, you're using the same jars I used above for the XSL transformation. Jeremias Maerki In the end, the following took just over 4 seconds, which means I don't think it was xalan which took so long. C:\Program Files\Java\fop-0.20.4>java -cp build\fop.jar;lib\batik.jar;lib\xalan-2.3.1.jar;lib\xercesImpl-2.0.1.jar;lib\xml-apis.jar;lib\avalon-framework-cvs-20020315.jar;lib\logkit-1.0.jar;lib\jimi-1.0.jar org.apache.fop.apps.Fop -d -fo 030123_test_xml_med7_MIWC_1.1.out.xml -pdf 030123_test_XSLT_001.pdf [DEBUG] Input mode: [DEBUG] FO [DEBUG] fo input file: 030123_test_xml_med7_MIWC_1.1.out.xml [DEBUG] Output mode: [DEBUG] pdf [DEBUG] output file: 030123_test_XSLT_001.pdf [DEBUG] OPTIONS [DEBUG] no user configuration file is used [default] [DEBUG] debug mode on [DEBUG] dump configuration [DEBUG] quiet mode on [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [DEBUG] base directory: file:/C:/Program Files/Java/fop-0.20.4/ [INFO] FOP 0.20.4 [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [INFO] building formatting object tree [DEBUG] setting up fonts [INFO] [1] [INFO] [2] [DEBUG] Last page-sequence produced 2 pages. [INFO] Parsing of document complete, stopping renderer [DEBUG] Initial heap size: 726Kb [DEBUG] Current heap size: 5396Kb [DEBUG] Total memory used: 4669Kb [DEBUG] Memory use is indicative; no GC was performed [DEBUG] These figures should not be used comparatively [DEBUG] Total time used: 4422ms [DEBUG] Pages rendered: 2 [DEBUG] Avg render time: 2211ms/page C:\Program Files\Java\fop-0.20.4> -- Clay Leeds - [EMAIL PROTECTED] Web Developer - Medata, Inc. - http://www.medata.com PGP Public Key: https://mail.medata.com/pgp/cleeds.asc - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
On 23.01.2003 22:15:17 Clay Leeds wrote: > It helps. However, I only have fop-0.20.4 (&.5rc) installed. I also > notice that in the lib/ folder I've got xalan-2.3.1.jar. Will this > suffice, or do I need to install the full version of xalan. Also, the > current version of xalan is 2.4.1... should I try it with that? Yep. In this case do: java -cp lib/xml-apis.jar;lib/xercesImpl-2.2.1.jar;lib/xalan-2.4.1.jar org.apache.xalan.xslt.Process -IN test2.xml -XSL test2.xsl -OUT test2.out.xml -DIAG That will run the XSL transformation. It assumes you're in the root directory of your FOP distribution (location of fop.bat) and you're using Windows. You may have to adjust the names of the jar files depending on the FOP version. > > Lastly, I want my system to match--as closely as possible--that of my > clients. What steps can I take to ensure that I can regain that "parity" > for testing? Well, if you use fop.bat to do the testing, you're using the same jars I used above for the XSL transformation. Jeremias Maerki - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Jeremias Maerki wrote: Run the XSL transformation alone. Using fop.bat you've done: XML > XSL-FO ---> PDF (XSLT) (FOP) The XSLT part of it is this: XML > XSL-FO (XSLT) (I hope the ASCII art comes out ok) Since the comments should in theory not reach FOP it is my guess that something's happening during the XSL transformation from XML to FO. Xalan does that, for example. I don't know what XSLT processor you use. If you use Xalan you can use its command line: http://xml.apache.org/xalan-j/commandline.html I think you can use the -DIAG option to get timings for the transformation. Does that clear it up? It helps. However, I only have fop-0.20.4 (&.5rc) installed. I also notice that in the lib/ folder I've got xalan-2.3.1.jar. Will this suffice, or do I need to install the full version of xalan. Also, the current version of xalan is 2.4.1... should I try it with that? Lastly, I want my system to match--as closely as possible--that of my clients. What steps can I take to ensure that I can regain that "parity" for testing? BTW, thanks for the quick replies! :-) -- Clay Leeds - [EMAIL PROTECTED] Web Developer - Medata, Inc. - http://www.medata.com PGP Public Key: https://mail.medata.com/pgp/cleeds.asc - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Run the XSL transformation alone. Using fop.bat you've done: XML > XSL-FO ---> PDF (XSLT) (FOP) The XSLT part of it is this: XML > XSL-FO (XSLT) (I hope the ASCII art comes out ok) Since the comments should in theory not reach FOP it is my guess that something's happening during the XSL transformation from XML to FO. Xalan does that, for example. I don't know what XSLT processor you use. If you use Xalan you can use its command line: http://xml.apache.org/xalan-j/commandline.html I think you can use the -DIAG option to get timings for the transformation. Does that clear it up? On 23.01.2003 21:47:09 Clay Leeds wrote: > Jeremias, > > Jeremias Maerki wrote: > > Hey Maestro! > > > > Have you tried to run only the XSLT part of it? Do that and compare > > execution times there. That should be interesting. > > > > > > Jeremias Maerki > > I don't know how to "run only the XSLT part of it"... Can you give any > more information as to what you're referring?__ Jeremias Maerki - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Jeremias, Jeremias Maerki wrote: Hey Maestro! Have you tried to run only the XSLT part of it? Do that and compare execution times there. That should be interesting. Jeremias Maerki I don't know how to "run only the XSLT part of it"... Can you give any more information as to what you're referring?__ -- Clay Leeds - [EMAIL PROTECTED] Web Developer - Medata, Inc. - http://www.medata.com PGP Public Key: https://mail.medata.com/pgp/cleeds.asc - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
Re: FOP 0.20.4 & 5rc timing w large COMMENT sections
Hey Maestro! Have you tried to run only the XSLT part of it? Do that and compare execution times there. That should be interesting. Jeremias Maerki - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
FOP 0.20.4 & 5rc timing w large COMMENT sections
Sorry for the re-post. I'd neglected to include some information on the XML file info summary for the un-commented version. I've included it, as well as a summary of the FOP .4 & .5rc rendering times. You can delete the previous post. :-) Howdy folks, I have a file where I've commented a fairly large section of my source XML file (commented areas affect FLOW content--so it won't print out so many darn pages! :-). I've found significant discrepancies involving the timing of outputting the commented version vs. the uncommented version. The intention of this message is to give a heads up (and hopefully initiate discussion) that there may be discrepancies in rendering time vs. actual output time when dealing with large comment sections. Respectfully, Web Maestro Clay === XML file & timing summary: COMMENTED version (1 page of output): FILENAME - test_NCWC_sm_1.1.XML FILESIZE - 240 KB (246,218 bytes) 0.20.4 TIME: ~01:05 (actual) 0.20.4 TIME: 2000ms (reported rendering time) 0.20.5rc TIME: ~00:58 (actual) 0.20.5rc TIME: 1687ms (reported rendering time) === UN-COMMENTED version (5 pages of output): FILENAME - test_NCWC_1.1.XML FILESIZE - 240 KB (246,213 bytes) 0.20.4 TIME: ~00:15 (actual) 0.20.4 TIME: 2000ms (reported rendering time) 0.20.5rc TIME: ~00:09 (actual) 0.20.5rc TIME: 8344ms (reported rendering time) When I processed the COMMENTED version of the file--which printed out a 1 page document--it took 01:05 (1 minute 5 seconds) watching the system clock, but the DOS output showed ~2000ms: C:\Program Files\Java\fop-0.20.4>fop -d -xml C:\xml_med7\biglinux\test_NCWC_sm_1.1.XML -xsl C:\xml_med7\fo\xml_med7_default.fo -pdf 030123_NCWC_sm_1.1.pdf [DEBUG] Input mode: [DEBUG] xslt transformation [DEBUG] xml input file: C:\xml_med7\biglinux\test_NCWC_sm_1.1.XML [DEBUG] xslt stylesheet: C:\xml_med7\fo\xml_med7_default.fo [DEBUG] Output mode: [DEBUG] pdf [DEBUG] output file: 030123_NCWC_sm_1.1.pdf [DEBUG] OPTIONS [DEBUG] no user configuration file is used [default] [DEBUG] debug mode on [DEBUG] dump configuration [DEBUG] quiet mode on [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [DEBUG] base directory: file:/C:/xml_med7/biglinux/ [INFO] FOP 0.20.4 [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [INFO] building formatting object tree [DEBUG] setting up fonts [INFO] [1] [DEBUG] Last page-sequence produced 1 pages. [INFO] Parsing of document complete, stopping renderer [DEBUG] Initial heap size: 2525Kb [DEBUG] Current heap size: 3641Kb [DEBUG] Total memory used: 1115Kb [DEBUG] Memory use is indicative; no GC was performed [DEBUG] These figures should not be used comparatively [DEBUG] Total time used: 2000ms [DEBUG] Pages rendered: 1 [DEBUG] Avg render time: 2000ms/page C:\Program Files\Java\fop-0.20.4> It took 2 seconds of rendering time compared to 65 seconds of time taken from command line [Enter] to FOP output completion. But when I run the normal un-COMMENTED version, it now took 15 seconds. However, the DOS window showed: C:\Program Files\Java\fop-0.20.4>fop -d -xml C:\xml_med7\biglinux\test_NCWC_1.1.XML -xsl C:\xml_med7\fo\xml_med7_default.fo -pdf 030123_NCWC_1.1.pdf [DEBUG] Input mode: [DEBUG] xslt transformation [DEBUG] xml input file: C:\xml_med7\biglinux\test_NCWC_1.1.XML [DEBUG] xslt stylesheet: C:\xml_med7\fo\xml_med7_default.fo [DEBUG] Output mode: [DEBUG] pdf [DEBUG] output file: 030123_NCWC_1.1.pdf [DEBUG] OPTIONS [DEBUG] no user configuration file is used [default] [DEBUG] debug mode on [DEBUG] dump configuration [DEBUG] quiet mode on [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [DEBUG] base directory: file:/C:/xml_med7/biglinux/ [INFO] FOP 0.20.4 [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [INFO] building formatting object tree [DEBUG] setting up fonts [INFO] [1] [INFO] [2] [INFO] [3] [INFO] [4] [INFO] [5] [DEBUG] Last page-sequence produced 5 pages. [INFO] Parsing of document complete, stopping renderer [DEBUG] Initial heap size: 2666Kb [DEBUG] Current heap size: 15561Kb [DEBUG] Total memory used: 12894Kb [DEBUG] Memory use is indicative; no GC was performed [DEBUG] These figures should not be used comparatively [DEBUG] Total time used: 9609ms [DEBUG] Pages rendered: 5 [DEBUG] Avg render time: 1921ms/page C:\Program Files\Java\fop-0.20.4> Both times it appeared to "hang" on "[DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser". In FOP's defense, I noticed the last line refers to "Avg render time:" BTW, I also tried this with 0.20.5rc and here're my results. COMMENTED version (00:58 seconds): C:\Program Files\Java\fop-0.20.5rc>fop -d -xml C:\xml_med7\biglinux\test_NCWC_sm_1.1.XML -xsl C:\xml_med7\fo\xml_med7_default.fo -pdf 030123_NCWC_sm_1.1.pdf [DEBUG] Input mode: [DEBUG] xslt transformation [DEBUG] xml input file: C:\xml_med7\biglinux\test_NCWC_sm_1.1.XML [DEBUG] xslt stylesheet: C:\xml_med7\fo\xml_med7_default.fo [DEBUG] Output mode: [DEBUG] pdf [DEBUG] output
FOP 0.20.4 & 5rc timing w large COMMENT sections
Howdy folks, I have a file where I've commented a fairly large section of my source XML file (commented areas affect FLOW content--so it won't print out so many darn pages! :-). I've found significant discrepancies involving the timing of outputting the commented version vs. the uncommented version. The intention of this message is to give a heads up (and hopefully initiate discussion) that there may be discrepancies in rendering time vs. actual output time when dealing with large comment sections. Respectfully, Web Maestro Clay === XML file info: COMMENTED version (1 page of output): FILENAME - test_NCWC_sm_1.1.XML FILESIZE - 240 KB (246,218 bytes) TIME: 01:05 (actual) TIME: 2000ms (reported rendering time) UN-COMMENTED version (5 pages of output): FILENAME - test_NCWC_1.1.XML FILESIZE - 240 KB (246,213 bytes) When I processed the COMMENTED version of the file--which printed out a 1 page document--it took 01:05 (1 minute 5 seconds) watching the system clock, but the DOS output showed ~2000ms: C:\Program Files\Java\fop-0.20.4>fop -d -xml C:\xml_med7\biglinux\test_NCWC_sm_1.1.XML -xsl C:\xml_med7\fo\xml_med7_default.fo -pdf 030123_NCWC_sm_1.1.pdf [DEBUG] Input mode: [DEBUG] xslt transformation [DEBUG] xml input file: C:\xml_med7\biglinux\test_NCWC_sm_1.1.XML [DEBUG] xslt stylesheet: C:\xml_med7\fo\xml_med7_default.fo [DEBUG] Output mode: [DEBUG] pdf [DEBUG] output file: 030123_NCWC_sm_1.1.pdf [DEBUG] OPTIONS [DEBUG] no user configuration file is used [default] [DEBUG] debug mode on [DEBUG] dump configuration [DEBUG] quiet mode on [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [DEBUG] base directory: file:/C:/xml_med7/biglinux/ [INFO] FOP 0.20.4 [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [INFO] building formatting object tree [DEBUG] setting up fonts [INFO] [1] [DEBUG] Last page-sequence produced 1 pages. [INFO] Parsing of document complete, stopping renderer [DEBUG] Initial heap size: 2525Kb [DEBUG] Current heap size: 3641Kb [DEBUG] Total memory used: 1115Kb [DEBUG] Memory use is indicative; no GC was performed [DEBUG] These figures should not be used comparatively [DEBUG] Total time used: 2000ms [DEBUG] Pages rendered: 1 [DEBUG] Avg render time: 2000ms/page C:\Program Files\Java\fop-0.20.4> It took 2 seconds of rendering time compared to 65 seconds of time taken from command line [Enter] to FOP output completion. But when I run the normal un-COMMENTED version, it now took 15 seconds. However, the DOS window showed: C:\Program Files\Java\fop-0.20.4>fop -d -xml C:\xml_med7\biglinux\test_NCWC_1.1.XML -xsl C:\xml_med7\fo\xml_med7_default.fo -pdf 030123_NCWC_1.1.pdf [DEBUG] Input mode: [DEBUG] xslt transformation [DEBUG] xml input file: C:\xml_med7\biglinux\test_NCWC_1.1.XML [DEBUG] xslt stylesheet: C:\xml_med7\fo\xml_med7_default.fo [DEBUG] Output mode: [DEBUG] pdf [DEBUG] output file: 030123_NCWC_1.1.pdf [DEBUG] OPTIONS [DEBUG] no user configuration file is used [default] [DEBUG] debug mode on [DEBUG] dump configuration [DEBUG] quiet mode on [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [DEBUG] base directory: file:/C:/xml_med7/biglinux/ [INFO] FOP 0.20.4 [DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [INFO] building formatting object tree [DEBUG] setting up fonts [INFO] [1] [INFO] [2] [INFO] [3] [INFO] [4] [INFO] [5] [DEBUG] Last page-sequence produced 5 pages. [INFO] Parsing of document complete, stopping renderer [DEBUG] Initial heap size: 2666Kb [DEBUG] Current heap size: 15561Kb [DEBUG] Total memory used: 12894Kb [DEBUG] Memory use is indicative; no GC was performed [DEBUG] These figures should not be used comparatively [DEBUG] Total time used: 9609ms [DEBUG] Pages rendered: 5 [DEBUG] Avg render time: 1921ms/page C:\Program Files\Java\fop-0.20.4> Both times it appeared to "hang" on "[DEBUG] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser". In FOP's defense, I noticed the last line refers to "Avg render time:" BTW, I also tried this with 0.20.5rc and here're my results. COMMENTED version (00:58 seconds): C:\Program Files\Java\fop-0.20.5rc>fop -d -xml C:\xml_med7\biglinux\test_NCWC_sm_1.1.XML -xsl C:\xml_med7\fo\xml_med7_default.fo -pdf 030123_NCWC_sm_1.1.pdf [DEBUG] Input mode: [DEBUG] xslt transformation [DEBUG] xml input file: C:\xml_med7\biglinux\test_NCWC_sm_1.1.XML [DEBUG] xslt stylesheet: C:\xml_med7\fo\xml_med7_default.fo [DEBUG] Output mode: [DEBUG] pdf [DEBUG] output file: 030123_NCWC_sm_1.1.pdf [DEBUG] OPTIONS [DEBUG] no user configuration file is used [default] [DEBUG] debug mode on [DEBUG] dump configuration [DEBUG] quiet mode on [INFO] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [INFO] base directory: file:/C:/xml_med7/biglinux/ [INFO] FOP 0.20.5rc [INFO] Using org.apache.xerces.parsers.SAXParser as SAX2 Parser [INFO] building formatting object tree [INFO] setting up fonts [INFO] [1] [DEBUG] Last page-sequence produced 1 pages. [INFO] Parsin
