Community
Participate
Working Groups
The markup validator currently loads the file's model into memory. This can cause problems for users who have enabled markup validation, since the build validator may come across a large file without the user having ever opened it.
http://git.eclipse.org/c/sourceediting/webtools.sourceediting.git/commit/?id=b9d8adb0866fe6db674b4d569f6889ce9ca640b7 I've pushed the changes to master. The markup validator now tokenizes an XML file instead of using the structured model to perform validation. This has yielded a 95-99% decrease in the time it takes to perform validation on larger files (tested on a ~2MB file). Memory usage has also has also been dramatically improved. With the same ~2MB file: Garbage Collection using the old markup validator: [GC 83210K->67716K(129468K), 0.0038218 secs] [GC 83489K->76683K(129468K), 0.0256200 secs] [GC 93707K->90081K(129468K), 0.0147162 secs] [GC 90081K(129468K), 0.0051320 secs] [GC 107105K->104152K(129468K), 0.0672853 secs] [GC 121176K->117141K(134844K), 0.0448868 secs] [GC 134165K->126963K(144824K), 0.0283667 secs] [GC 143987K->132660K(150776K), 0.0151360 secs] [GC 132944K(150776K), 0.0040487 secs] [GC 149607K->138446K(236592K), 0.0109819 secs] [GC 155470K->145282K(236592K), 0.0171411 secs] [GC 162306K->146878K(236592K), 0.0089203 secs] [GC 163902K->145559K(236592K), 0.0020895 secs] [GC 162583K->145361K(236592K), 0.0021159 secs] [GC 162385K->145311K(236592K), 0.0021374 secs] [GC 162335K->145303K(236592K), 0.0020723 secs] [GC 162327K->145303K(236592K), 0.0020560 secs] [GC 162327K->145305K(236592K), 0.0021291 secs] [GC 162329K->145303K(236592K), 0.0021311 secs] [GC 162327K->145303K(236592K), 0.0020654 secs] [GC 162327K->145303K(236592K), 0.0021938 secs] [GC 162327K->145305K(236592K), 0.0021320 secs] [GC 162329K->145303K(236592K), 0.0021453 secs] [GC 162327K->145301K(236592K), 0.0022555 secs] [GC 162325K->145305K(236592K), 0.0020959 secs] [GC 162329K->145303K(236592K), 0.0020837 secs] [GC 162327K->145303K(236592K), 0.0020251 secs] [GC 162327K->145301K(236592K), 0.0020939 secs] [GC 162325K->145305K(236592K), 0.0020974 secs] [GC 162329K->145303K(236592K), 0.0020893 secs] [GC 162327K->145303K(236592K), 0.0022085 secs] [GC 162327K->145303K(236592K), 0.0021111 secs] [GC 162327K->145305K(236592K), 0.0020748 secs] [GC 162329K->145303K(236592K), 0.0022079 secs] [Full GC 157249K->67775K(236592K), 0.4674900 secs] Garbage Collection with the new validator: [GC 83272K->67968K(129608K), 0.0037404 secs] [GC 84992K->75629K(129608K), 0.0268719 secs] [GC 92653K->76503K(129608K), 0.0063339 secs] [GC 93527K->75898K(129608K), 0.0024181 secs] [GC 92922K->75691K(129608K), 0.0024072 secs] [Full GC 89762K->67665K(129608K), 0.4362108 secs]
Tested with a 16MB XML file. Validation with new validator: 2511ms Validation with old validator: (I don't know. I gave up after five minutes)
