Revision: 5482
http://sourceforge.net/p/jump-pilot/code/5482
Author: edso
Date: 2017-07-30 16:26:14 +0000 (Sun, 30 Jul 2017)
Log Message:
-----------
speeding up JML/GML reader when reading time attributes by parsing them lazy
(during access later)
- adding FlexibleFeature, FlexibleFeatureSchema
- porting GMLReader to use the flexible classes
Modified Paths:
--------------
core/trunk/src/com/vividsolutions/jump/io/GMLInputTemplate.java
core/trunk/src/com/vividsolutions/jump/io/GMLReader.java
Added Paths:
-----------
core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeature.java
core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeatureSchema.java
Added: core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeature.java
===================================================================
--- core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeature.java
(rev 0)
+++ core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeature.java
2017-07-30 16:26:14 UTC (rev 5482)
@@ -0,0 +1,156 @@
+package com.vividsolutions.jump.feature;
+
+import java.security.InvalidParameterException;
+import java.text.ParseException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Date;
+import java.util.List;
+
+import com.vividsolutions.jts.geom.Geometry;
+import com.vividsolutions.jump.util.FlexibleDateParser;
+
+/**
+ * a FlexibleFeature based on {@link BasicFeature} originally used by the
+ * GeoJSON reader. currently adding
+ * - "autoextends" by returning null for undefined attribs
+ * - lazy conversion of attributes (see {@link #getAttribute(int)})
+ * currently types String, Date, Time, Timestamp
+ */
+public class FlexibleFeature extends BasicFeature {
+ private FlexibleFeatureSchema featureSchema;
+
+ public FlexibleFeature(FlexibleFeatureSchema featureSchema) {
+ super(featureSchema);
+ this.featureSchema = featureSchema;
+ }
+
+ @Override
+ public Object getAttribute(int i) {
+ if (i < 0)
+ throw new InvalidParameterException("index must be greater or equal
zero");
+
+ Object attrib = null;
+
+ Object[] attributes = super.getAttributes();
+ // only grab attrib if stack holds it already
+ if (i < attributes.length)
+ attrib = attributes[i];
+
+ // OJ does not allow null geoms!
+ if (i == featureSchema.getGeometryIndex()) {
+ // create and set an empty geom
+ if (attrib == null) {
+ attrib = featureSchema.createEmptyGeometry();
+ setGeometry((Geometry) attrib);
+ }
+ }
+ // enforce String if schema says so
+ else if (featureSchema.getAttributeType(i).equals(AttributeType.STRING) &&
attrib != null
+ && !(attrib instanceof String)) {
+ attrib = String.valueOf(attrib);
+ }
+ // enforce date object if not converted already
+ else if (featureSchema.getAttributeType(i).equals(AttributeType.DATE) &&
attrib != null
+ && !AttributeType.DATE.toJavaClass().isInstance(attrib)) {
+ try {
+ Date d;
+ // the celleditor replaces us w/ a java.util.Date object
+ if (java.util.Date.class.isInstance(attrib))
+ d = (java.util.Date) attrib;
+ else
+ d = FlexibleDateParser.getDefaultInstance().parse(attrib.toString(),
false);
+ attrib = (d==null) ? null : new java.sql.Date(d.getTime());
+ // update the attribute object, so the conversion does not happen on
+ // every getAttrib()
+ setAttribute(i, attrib);
+ } catch (ParseException e) {
+ // TODO: we should find a way to tell the user
+ attrib = null;
+ }
+ }
+ // enforce time object if not converted already
+ else if (featureSchema.getAttributeType(i).equals(AttributeType.TIME) &&
attrib != null
+ && !AttributeType.TIME.toJavaClass().isInstance(attrib)) {
+ try {
+ Date d;
+ // the celleditor replaces us w/ a java.util.Date object
+ if (java.util.Date.class.isInstance(attrib))
+ d = (java.util.Date) attrib;
+ else
+ d = FlexibleDateParser.getDefaultInstance().parse(attrib.toString(),
false);
+ attrib = (d==null) ? null : new java.sql.Time(d.getTime());
+ // update the attribute object, so the conversion does not happen on
+ // every getAttrib()
+ setAttribute(i, attrib);
+ } catch (ParseException e) {
+ // TODO: we should find a way to tell the user
+ attrib = null;
+ }
+ }
+ // enforce timestamp object if not converted already
+ else if (featureSchema.getAttributeType(i).equals(AttributeType.TIMESTAMP)
&& attrib != null
+ && !AttributeType.TIMESTAMP.toJavaClass().isInstance(attrib)) {
+ try {
+ Date d;
+ // the celleditor replaces us w/ a java.util.Date object
+ if (java.util.Date.class.isInstance(attrib))
+ d = (java.util.Date) attrib;
+ else
+ d = FlexibleDateParser.getDefaultInstance().parse(attrib.toString(),
false);
+ attrib = (d==null) ? null : new java.sql.Timestamp(d.getTime());
+ // update the attribute object, so the conversion does not happen on
+ // every getAttrib()
+ setAttribute(i, attrib);
+ } catch (ParseException e) {
+ // TODO: we should find a way to tell the user
+ attrib = null;
+ }
+ }
+
+ return attrib;
+ }
+
+ /**
+ * setting an attribute, fixing the underlying array in case the schema
+ * changed inbetween
+ */
+ public void setAttribute(int attributeIndex, Object newAttribute) {
+ FeatureSchema schema = super.getSchema();
+ Object[] oldAttribs = super.getAttributes();
+ // add fields if schema changed in between
+ int diffCount = schema.getAttributeCount() - oldAttribs.length;
+ if (diffCount > 0) {
+ List attributes = new ArrayList(Arrays.asList(oldAttribs));
+ attributes.addAll(Arrays.asList(new Object[diffCount]));
+ super.setAttributes(attributes.toArray());
+ }
+ super.setAttribute(attributeIndex, newAttribute);
+ }
+
+ /**
+ * setting the geometry by explicitly using the flexible setAttribute()
method
+ * above
+ */
+ public void setGeometry(Geometry geometry) {
+ setAttribute(getSchema().getGeometryIndex(), geometry);
+ }
+
+ /**
+ * getting the geometry by explicitly using the flexible getAttribute()
method
+ * above
+ */
+ public Geometry getGeometry() {
+ return (Geometry) getAttribute(featureSchema.getGeometryIndex());
+ }
+
+ /**
+ * TODO: the method shouldn't be used anyway, still maybe we will have to
+ * implement it later
+ */
+ @Override
+ public Object[] getAttributes() {
+ throw new UnsupportedOperationException("currently not implemented");
+ }
+
+}
Property changes on:
core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeature.java
___________________________________________________________________
Added: svn:mime-type
## -0,0 +1 ##
+text/plain
\ No newline at end of property
Added: core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeatureSchema.java
===================================================================
--- core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeatureSchema.java
(rev 0)
+++ core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeatureSchema.java
2017-07-30 16:26:14 UTC (rev 5482)
@@ -0,0 +1,61 @@
+package com.vividsolutions.jump.feature;
+
+import org.openjump.core.ui.util.GeometryUtils;
+
+import com.vividsolutions.jts.geom.Geometry;
+import com.vividsolutions.jts.geom.GeometryFactory;
+import com.vividsolutions.jump.workbench.Logger;
+
+/**
+ * a FlexibleFeatureSchema originally used by the GeoJSON reader.
+ * extends the basic {@link FeatureSchema} by
+ * - allow changing attrib types on the fly
+ * - creates empty geoms matching a previous set geomType
+ */
+public class FlexibleFeatureSchema extends FeatureSchema {
+ Class geometryClass = null;
+ GeometryFactory geometryFactory = new GeometryFactory();
+
+ public FlexibleFeatureSchema() {
+ }
+
+ public FlexibleFeatureSchema(FeatureSchema featureSchema) {
+ super(featureSchema);
+ }
+
+ public void setAttributeType(int attributeIndex, AttributeType type) {
+ attributeTypes.set(attributeIndex, type);
+ }
+
+ public void setAttributeType(String name, AttributeType type) {
+ setAttributeType(super.getAttributeIndex(name), type);
+ }
+
+ public void setGeometryType(Class clazz) {
+ geometryClass = clazz;
+ }
+
+ public Class getGeometryType() {
+ return geometryClass;
+ }
+
+ /**
+ * creates an empty geometry matching the geom type set already or an empty
+ * geom collection if that fails
+ *
+ * @return geometry
+ */
+ public Geometry createEmptyGeometry() {
+ if (geometryClass != null) {
+ try {
+ return GeometryUtils
+ .createEmptyGeometry(geometryClass, geometryFactory);
+ } catch (Exception e) {
+ Logger.debug(e);
+ }
+ }
+
+ return geometryFactory.createGeometryCollection(null);
+ }
+
+}
Property changes on:
core/trunk/src/com/vividsolutions/jump/feature/FlexibleFeatureSchema.java
___________________________________________________________________
Added: svn:mime-type
## -0,0 +1 ##
+text/plain
\ No newline at end of property
Modified: core/trunk/src/com/vividsolutions/jump/io/GMLInputTemplate.java
===================================================================
--- core/trunk/src/com/vividsolutions/jump/io/GMLInputTemplate.java
2017-07-30 16:11:49 UTC (rev 5481)
+++ core/trunk/src/com/vividsolutions/jump/io/GMLInputTemplate.java
2017-07-30 16:26:14 UTC (rev 5482)
@@ -36,7 +36,10 @@
import java.io.IOException;
import java.io.InputStream;
import java.io.Reader;
+import java.lang.reflect.Constructor;
+import java.lang.reflect.InvocationTargetException;
import java.util.ArrayList;
+import java.util.HashMap;
import org.xml.sax.Attributes;
import org.xml.sax.InputSource;
@@ -46,7 +49,7 @@
import org.xml.sax.helpers.DefaultHandler;
import com.vividsolutions.jump.feature.AttributeType;
-import com.vividsolutions.jump.feature.FeatureSchema;
+import com.vividsolutions.jump.feature.FlexibleFeatureSchema;
import com.vividsolutions.jump.util.FlexibleDateParser;
@@ -87,6 +90,16 @@
private String lastStartTag_qName;
private Attributes lastStartTag_atts;
+ // a list of date attrib types that are treated differently
+ static private ArrayList<AttributeType> dateAttributeTypes = new
ArrayList();
+ static {
+ dateAttributeTypes.add(AttributeType.DATE);
+ dateAttributeTypes.add(AttributeType.TIMESTAMP);
+ dateAttributeTypes.add(AttributeType.TIME);
+ }
+
+ private HashMap<AttributeType, Constructor> constructorCache = new
HashMap();
+
/**
* constructor - makes a new org.apache.xerces.parser and makes this class
be the SAX
* content and error handler.
@@ -114,13 +127,13 @@
/**
* Converts this GMLInputTemplate to a feature schema.
*/
- public FeatureSchema toFeatureSchema() throws ParseException {
+ public FlexibleFeatureSchema toFeatureSchema() throws ParseException {
if (!(loaded)) {
throw new ParseException(
"requested toFeatureSchema w/o loading the template");
}
- FeatureSchema fcmd = new FeatureSchema();
+ FlexibleFeatureSchema fcmd = new FlexibleFeatureSchema();
fcmd.addAttribute("GEOMETRY", AttributeType.GEOMETRY);
@@ -291,86 +304,110 @@
val = xmlAtts.getValue(cd.valueAttribute);
}
- //have the value as a string, make it an object
- if (cd.type == AttributeType.STRING ||
- cd.type == AttributeType.VARCHAR ||
- cd.type == AttributeType.LONGVARCHAR ||
- cd.type == AttributeType.CHAR ||
- cd.type == AttributeType.TEXT) {
- return val;
- }
-
- if (cd.type == AttributeType.INTEGER ||
- cd.type == AttributeType.SMALLINT ||
- cd.type == AttributeType.TINYINT) {
- try {
- //Was Long, but JUMP expects AttributeType.INTEGER to hold
Integers.
- //e.g. open JML file then save as Shapefile => get
ClassCastException.
- //Dave Blasby says there was a reason for changing it to Long,
but
- //can't remember -- suspects there were datasets whose INTEGER
- //values didn't fit in an Integer. [Jon Aquino 1/13/2004]
-
- //Compromise -- try Long if Integer fails. Some other parts of
JUMP
- //won't like it (exceptions), but it's better than null.
Actually I don't like
- //this null business -- future: warn the user. [Jon Aquino
1/13/2004]
- return Integer.parseInt(val);
- } catch (Exception e) {
- return null;
+ //try running the default constructor from string for the class
defined in AttributeType
+ if (!dateAttributeTypes.contains(cd.type))
+ try {
+ Constructor c = null;
+ c = constructorCache.get(cd.type);
+ if (c == null) {
+ c = cd.type.toJavaClass().getConstructor(new Class[] {
String.class });
+ constructorCache.put(cd.type, c);
}
+
+ return c.newInstance(val);
+ } catch (InstantiationException | IllegalAccessException |
IllegalArgumentException | InvocationTargetException
+ | SecurityException e) {
+ // TODO not sure if ignoring is safe, on the other hand we are
currently
+ // processing lots of attributes, we can't just spit out a
plethora of
+ // error
+ } catch (NoSuchMethodException e) {
+ // TODO maybe printing warning for classes without a String
+ // constructor?;
+ }
+
+ // we can return the string safely as the attrib is lazily converted
by FlexibleFeature
+ // REASON: utilizing FlexibleDateParser slowed down reading JML by
magnitudes here
+ if (dateAttributeTypes.contains(cd.type)) {
+ return val;
}
-
- if (cd.type == AttributeType.LONG || cd.type == AttributeType.BIGINT) {
- try {
- return Long.parseLong(val);
- } catch (Exception e) {
- return null;
- }
- }
-
- if (cd.type == AttributeType.DOUBLE ||
- cd.type == AttributeType.REAL ||
- cd.type == AttributeType.FLOAT ||
- cd.type == AttributeType.NUMERIC ||
- cd.type == AttributeType.DECIMAL ||
- cd.type == AttributeType.BIGDECIMAL) {
- try {
- return new Double(val);
- } catch (Exception e) {
- return null;
- }
- }
- //Adding date support. Can we throw an exception if an exception
- //occurs or if the type is unrecognized? [Jon Aquino]
- if (cd.type == AttributeType.DATE ||
- cd.type == AttributeType.TIMESTAMP ||
- cd.type == AttributeType.TIME) {
- try {
- return dateParser.parse(val, false);
- } catch (Exception e) {
- return null;
- }
- }
+ // previous code for attrib parsing, rplaced by the above 07.2017 [ede]
+// //have the value as a string, make it an object
+// if (cd.type == AttributeType.STRING ||
+// cd.type == AttributeType.VARCHAR ||
+// cd.type == AttributeType.LONGVARCHAR ||
+// cd.type == AttributeType.CHAR ||
+// cd.type == AttributeType.TEXT) {
+// return val;
+// }
+//
+// if (cd.type == AttributeType.INTEGER ||
+// cd.type == AttributeType.SMALLINT ||
+// cd.type == AttributeType.TINYINT) {
+// try {
+// //Was Long, but JUMP expects AttributeType.INTEGER to hold
Integers.
+// //e.g. open JML file then save as Shapefile => get
ClassCastException.
+// //Dave Blasby says there was a reason for changing it to
Long, but
+// //can't remember -- suspects there were datasets whose
INTEGER
+// //values didn't fit in an Integer. [Jon Aquino 1/13/2004]
+//
+// //Compromise -- try Long if Integer fails. Some other parts
of JUMP
+// //won't like it (exceptions), but it's better than null.
Actually I don't like
+// //this null business -- future: warn the user. [Jon Aquino
1/13/2004]
+// return Integer.parseInt(val);
+// } catch (Exception e) {
+// return null;
+// }
+// }
+//
+// if (cd.type == AttributeType.LONG || cd.type ==
AttributeType.BIGINT) {
+// try {
+// return Long.parseLong(val);
+// } catch (Exception e) {
+// return null;
+// }
+// }
+//
+// if (cd.type == AttributeType.DOUBLE ||
+// cd.type == AttributeType.REAL ||
+// cd.type == AttributeType.FLOAT ||
+// cd.type == AttributeType.NUMERIC ||
+// cd.type == AttributeType.DECIMAL ||
+// cd.type == AttributeType.BIGDECIMAL) {
+// try {
+// return new Double(val);
+// } catch (Exception e) {
+// return null;
+// }
+// }
+//
+//// //Adding date support. Can we throw an exception if an exception
+//// //occurs or if the type is unrecognized? [Jon Aquino]
+//// if (dateAttributeTypes.contains(cd.type)) {
+//// try {
+//// return dateParser.parse(val, false);
+//// } catch (Exception e) {
+//// return null;
+//// }
+//// }
+//
+// if (cd.type == AttributeType.BOOLEAN || cd.type ==
AttributeType.BIT) {
+// try {
+// return Boolean.parseBoolean(val);
+// } catch (Exception e) {
+// return null;
+// }
+// }
+//
+// if (cd.type == AttributeType.OBJECT)
+// {
+// return val; // the GML file has text in it and we want to
convert it to an "object"
+// // just return a String since we dont know anything
else about it!
+// }
- if (cd.type == AttributeType.BOOLEAN || cd.type == AttributeType.BIT) {
- try {
- return Boolean.parseBoolean(val);
- } catch (Exception e) {
- return null;
- }
- }
-
- if (cd.type == AttributeType.OBJECT)
- {
- return val; // the GML file has text in it and we want to
convert it to an "object"
- // just return a String since we dont know anything
else about it!
- }
-
return null; //unknown type
}
- private FlexibleDateParser dateParser = new FlexibleDateParser();
-
////////////////////////////////////////////////////////////////////
// Error handlers.
////////////////////////////////////////////////////////////////////
Modified: core/trunk/src/com/vividsolutions/jump/io/GMLReader.java
===================================================================
--- core/trunk/src/com/vividsolutions/jump/io/GMLReader.java 2017-07-30
16:11:49 UTC (rev 5481)
+++ core/trunk/src/com/vividsolutions/jump/io/GMLReader.java 2017-07-30
16:26:14 UTC (rev 5482)
@@ -31,9 +31,39 @@
*/
package com.vividsolutions.jump.io;
-import com.vividsolutions.jts.geom.*;
+import java.io.BufferedInputStream;
+import java.io.FileInputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.io.Reader;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+import java.util.StringTokenizer;
+import java.util.regex.Pattern;
+
+import org.xml.sax.Attributes;
+import org.xml.sax.InputSource;
+import org.xml.sax.SAXException;
+import org.xml.sax.SAXParseException;
+import org.xml.sax.XMLReader;
+import org.xml.sax.helpers.DefaultHandler;
+
+import com.vividsolutions.jts.geom.Coordinate;
+import com.vividsolutions.jts.geom.Geometry;
+import com.vividsolutions.jts.geom.GeometryFactory;
+import com.vividsolutions.jts.geom.LineString;
+import com.vividsolutions.jts.geom.LinearRing;
+import com.vividsolutions.jts.geom.Point;
+import com.vividsolutions.jts.geom.Polygon;
+import com.vividsolutions.jts.geom.PrecisionModel;
import com.vividsolutions.jump.I18N;
-import com.vividsolutions.jump.feature.*;
+import com.vividsolutions.jump.feature.AttributeType;
+import com.vividsolutions.jump.feature.Feature;
+import com.vividsolutions.jump.feature.FeatureCollection;
+import com.vividsolutions.jump.feature.FeatureDataset;
+import com.vividsolutions.jump.feature.FlexibleFeature;
+import com.vividsolutions.jump.feature.FlexibleFeatureSchema;
import com.vividsolutions.jump.task.DummyTaskMonitor;
import com.vividsolutions.jump.task.TaskMonitor;
import com.vividsolutions.jump.task.TaskMonitorSupport;
@@ -40,15 +70,6 @@
import com.vividsolutions.jump.task.TaskMonitorUtil;
import com.vividsolutions.jump.util.Timer;
-import org.xml.sax.*;
-import org.xml.sax.helpers.DefaultHandler;
-
-import java.io.*;
-import java.util.Collection;
-import java.util.List;
-import java.util.ArrayList;
-import java.util.StringTokenizer;
-
/**
* GMLReader is a {@link JUMPReader} specialized to read GML files.
*
@@ -264,7 +285,7 @@
private Feature currentFeature;
private int currentGeometryNumb = 1;
private FeatureCollection fc;
- private FeatureSchema fcmd; // list of geometries
+ private FlexibleFeatureSchema fcmd; // list of geometries
private Geometry finalGeometry; // list of geometrycollections - list of
list of
// geometry
private String current_geom_qname = "";
@@ -287,7 +308,7 @@
// (Coordinate)
private Polygon polygon; // polygon
- // higherlevel geomery object
+ // higher level geometry object
private ArrayList<ArrayList> recursivegeometry = new ArrayList<>();
// low-level geometry objects
@@ -305,6 +326,15 @@
*/
public boolean multiItemsAsLists = false;
+ // precompiled patterns for performance reasons
+ private static Pattern regex_geomMultiPoint, regex_geomMultiLineString,
regex_geomMultiPolygon, regex_geomLinearRing;
+ static {
+ regex_geomMultiPoint = Pattern.compile("^(?i)(gml:)?multipoint$");
+ regex_geomMultiLineString =
Pattern.compile("^(?i)(gml:)?multilinestring$");
+ regex_geomMultiPolygon = Pattern.compile("^(?i)(gml:)?multipolygon$");
+ regex_geomLinearRing = Pattern.compile("^(?i)(gml:)?linearring$");
+ }
+
/**
* Constructor - make a SAXParser and have this GMLReader be its
* ContentHandler and ErrorHandler.
@@ -435,13 +465,13 @@
linearRing = null;
}
- if (current_geom_qname.matches("^(?i)(gml:)?multipoint$"))
+ if (regex_geomMultiPoint.matcher(current_geom_qname).matches())
finalGeometry = geometryFactory
.createMultiPoint(geometry.toArray(new Point[0]));
- else if (current_geom_qname.matches("^(?i)(gml:)?multilinestring$"))
+ else if
(regex_geomMultiLineString.matcher(current_geom_qname).matches())
finalGeometry = geometryFactory
.createMultiLineString((geometry.toArray(new LineString[0])));
- else if (current_geom_qname.matches("^(?i)(gml:)?multipolygon$"))
+ else if
(regex_geomMultiPolygon.matcher(current_geom_qname).matches())
finalGeometry = geometryFactory
.createMultiPolygon((geometry.toArray(new Polygon[0])));
// else if (current_geom_qname.matches("^(?i)(gml:)?linearring$"))
@@ -841,7 +871,7 @@
&& (qName.compareToIgnoreCase(GMLinput.featureTag) == 0)) {
// found the feature tag
// System.out.println("found feature");
- currentFeature = new BasicFeature(fcmd);
+ currentFeature = new FlexibleFeature(fcmd);
STATE = STATE_GET_COLUMNS;
SRID = 0;// default SRID (reset for each feature, but should be
constant
// for a featurecollection)
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Jump-pilot-devel mailing list
Jump-pilot-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel
