Refactor reverse dependencies in org.opensha.commons

[Peter Powers]

Building javadoc for just org.opensha.commons revealed numerous reverse dependencies on org.opensha.sha.*

[Peter Powers]

org.opensha.commons.data.region.SitesInGriddedRegion? -->

• org.opensha.sha.util.SiteTranslator?

org.opensha.commons.data.siteData.gui.beans.OrderedSiteDataGUIBean -->

• org.opensha.sha.imr.ScalarIMR
• org.opensha.sha.util.SiteTranslator?

org.opensha.commons.data.siteData.util.SiteDataTypeParameterNameMap? -->

• org.opensha.sha.imr.ScalarIMR;
• org.opensha.sha.imr.attenRelImpl.Campbell_1997_AttenRel;
• org.opensha.sha.imr.param.SiteParams?.Vs30_Param;
• org.opensha.sha.util.SiteTranslator?;

org.opensha.commons.mapping.gmt.gui.GMT_MapGuiBean -->

• org.opensha.sha.gui.infoTools.ImageViewerWindow?

among others

[Kevin Milner]

This is still theoretically though not immediately important and has some nasty consequences. Delaying for the future.

[Kevin Milner]

I just checked out another version of the project where I deleted everything but commons to find all of the reverse dependencies (results below). Most are simple, but a few are tricky. If we decide to enforce this, then we should add a test (I guess using reflection...ugly) which validates this requirement to the svn post-commit hook.

• org.opensha.commons.metadata.MetadataLoader?
  • various ERF/IMR classes, only in main method which can just be removed
• org.opensha.commons.gui.plot.jfreechart.GriddedSurfaceXYDataSet
  • surface classes, this class should just be moved to SHA (no commons classes use it)
• org.opensha.commons.geo.LocationUtils?
  • distanceToSurf and distanceToSurfFast both take RuptureSurface? instances. This one is a little tougher, should I guess be moved but I don't have a good other location in mind. Thoughts?
• org.opensha.commons.data.function.HistogramFunction?
  • main class references BPT_DistCalc, should be removed
• org.opensha.commons.data.siteData.util.SiteDataTypeParameterNameMap?
  • ScalarIMR, should be moved to SHA (with OrderedSiteDataGUIBean
• org.opensha.commons.hpc.mpj.taskDispatch.MPJTaskCalculator
  • DeadlockDetectionThread?, which can/should be moved to commons
• org.opensha.commons.param.editor.impl.ConstrainedEstimateParameterEditor?
  • a couple ref fault param db classes, and EstimateViewer? in sha. might be a little tricky
• org.opensha.commons.util.FaultUtils?
  • relies on FaultTrace? as would be assumed. Could be moved to sha, but MagScalingRelationship? uses assertValidRake
• org.opensha.commons.gui.plot.jfreechart.tornado.TornadoDiagram?
  • relies on HeadlessGraphPanel?, which definitely should be in commons instead of sha
• org.opensha.commons.mapping.gmt.gui.GMT_MapGeneratorApplet
  • relies on IconFetcher?, which might be moveable to commons
• org.opensha.commons.mapping.gmt.gui.GMT_MapGuiBean
  • relies on ImageViewerWindow?, which can likely be moved to commons
• org.opensha.commons.calc.WeightedSampler?<T>
  • relies on IntegerPDF_FunctionSampler in scratch, which should probably be moved to commons
• org.opensha.commons.gui.AppVersionDialog?
  • relies on "Resources" class in ui package. Not sure why the "ui" package exists or what belongs there. Thoughts?
• org.opensha.commons.gui.DisclaimerDialog?
  • same as above, Resources class in ui
• org.opensha.commons.data.siteData.gui.beans.OrderedSiteDataGUIBean
  • many IMR/site data specific things, should be moved to SHA
• org.opensha.commons.param.editor.demo.ParameterApplet?
  • uses SimpleFaultParameter?, can be deleted or moved. probably not useful anymore
• org.opensha.commons.data.siteData.impl.WillsMap2000 & WillsMap2000TranslatedVs30
  • uses SiteTranslator? for Vs30/Wills conversions, needs some thought
• org.opensha.commons.data.siteData.impl.CVM4BasinDepth & CVM4i26BasinDepth
  • unused import can be removed
• org.opensha.commons.hpc.mpj.taskDispatch.DispatcherThread?
  • unused import can be removed

[Kevin Milner]

I just fixed all of the trial ones in [10933]. I also added a class which can generate the list of reverse references (and found a few not in the previous comments): scratch.kevin.CommonsImportCheck?

Here's the current list (with notes):

org.opensha.commons.util.FaultUtils?

• org.opensha.sha.faultSurface.FaultTrace?
  • Could be moved to sha, but MagScalingRelationship?? uses assertValidRake

org.opensha.commons.geo.LocationUtils?

• org.opensha.sha.faultSurface.AbstractEvenlyGriddedSurface?
• org.opensha.sha.faultSurface.RuptureSurface?
  • This one is a little tougher, should I guess be moved but I don't have a good other location in mind. Thoughts?

org.opensha.commons.param.editor.impl.ConstrainedEstimateParameterEditor?

• org.opensha.refFaultParamDb.gui.infotools.GUI_Utils
• org.opensha.refFaultParamDb.gui.infotools.InfoLabel?
• org.opensha.sha.gui.infoTools.EstimateViewer?

org.opensha.commons.geo.RegionUtils?

• org.opensha.sha.faultSurface.AbstractEvenlyGriddedSurface?

org.opensha.commons.data.region.SitesInGriddedRegion?

• org.opensha.sha.util.SiteTranslator?

org.opensha.commons.mapping.gmt.gui.GMT_MapGeneratorApplet

• org.opensha.sha.gui.util.IconFetcher?
  • tricky because IconFetcher? relies on JNLPGen, which references all of the apps. I guess JNLPGen could be turned into a general utility in commons, with code in sha which calls it for the apps (e.g. SHA_JNLPGen)

org.opensha.commons.gui.DisclaimerDialog?

• org.opensha.ui.components.Resources

org.opensha.commons.data.siteData.impl.WillsMap2000TranslatedVs30

• org.opensha.sha.util.SiteTranslator?
  • Wills Class -> Vs30 done in SiteTranslator?

org.opensha.commons.data.siteData.gui.beans.OrderedSiteDataGUIBean

• org.opensha.sha.imr.ScalarIMR
• org.opensha.sha.util.SiteDataTypeParameterNameMap?
• org.opensha.sha.util.SiteTranslator?

org.opensha.commons.data.siteData.impl.WillsMap2000

• org.opensha.sha.gui.servlets.siteEffect.WillsSiteClass?
• org.opensha.sha.util.SiteTranslator?
  • Wills Class -> Vs30 done in SiteTranslator?

org.opensha.commons.gui.AppVersionDialog?

• org.opensha.ui.components.Resources

11 offenders