Deep-blue

Scala, Geotoolkit and UTM Zone 10N gridded data (Part 1)

2012-01-26T14:00:00.000-08:00

Overview

I am currently working on some detailed GIS analysis. It's very cool stuff; I'm analyzing about 4 million undersea video annotations collected using VARS (which I also wrote). The problem that I keep encountering again and again is coordinate transforms and data formats.

Why are coordinate transforms a problem? Well, I like to analyze data using UTM grids. UTM has an advantage for analysis in that pixels represent a uniform area, for example, a single pixel might represent a 25mx25m area. Geographic grids do not usually have uniform pixel sizes, which makes inter-pixel comparisons difficult (e.g. I found 20 critters in pixel A and 20 in pixel B, but in geographic coordinate systems pixel A and B represent different surface areas, so I can't say that they have the same density of organisms). The problem is that all the positions of our ROVs and annotations are in geographic coordinates (i.e. latitude and longitude) so a transform from geographic coordinates to projected coordinates (UTM) is required. Since I'm writing automated software to generate thousands (actually hundreds of thousands) of data products, I need software libraries that I can use for the transforms. And since I'm doing a huge number of analysis the software has to be fast. The speed issue alone rules out ESRI products; they supply python as a scripting language. As great as python is, it just isn't fast enough for what I need. Instead, I'm working on the JVM (which is pretty darn fast). So in the examples below I'll be using Scala to illustrate how to do the analysis.

The second issue, data formats, is one of those insidious annoyances that just pervades the GIS world. For one, there's a lot of different formats. For another, one format does not rule them all, so it's tough to find a grid format that's broadly usable in different software applications. For this analysis, I'll be using GMT's 'GRD' format. Why? For one, it's NetCDF, so there's a zillion libraries to read it in any number of a zillion programming languages. I do most of my analysis on the JVM, so I can use NetCDF-Java there. In addition, I often do ad-hoc analysis in Matlab, so I can use nctoolbox (which I also wrote) to work with NetCDF there. The other reason that GRD is nice is that the CDL is very simple to understand. Here's an example CDL:

netcdf ETOPO1_Bed_g_gmt4 {
dimensions:
 x = 21601 ;
 y = 10801 ;
variables:
 double x(x) ;
  x:long_name = "Longitude" ;
  x:actual_range = -180., 180. ;
  x:units = "degrees" ;
 double y(y) ;
  y:long_name = "Latitude" ;
  y:actual_range = -90., 90. ;
  y:units = "degrees" ;
 int z(y, x) ;
  z:long_name = "z" ;
  z:_FillValue = -2147483648 ;
  z:actual_range = -10898., 8271. ;

// global attributes:
  :Conventions = "COARDS/CF-1.0" ;
  :title = "ETOPO1_Bed_g_gmt4.grd" ;
  :history = "grdreformat ETOPO1_Bed_g_gdal.grd ETOPO1_Bed_g_gmt4.grd=ni" ;
  :GMT_version = "4.4.0" ;
  :node_offset = 0 ;
}

The Setup

For this analysis, I'll need to include the NetCDF-Java libraries as well as Geotoolkit libraries. Geotoolkit is an open source Java library for developing geospatial applications. I use it because: I can actually get it to work, it's NetCDF support is decent, and the lead developer, Martin Desruisseaux, is excellent about responding to questions. There are several Geotoolkit modules that need to be included, they're listed in the pom.xml example below. At the time of this writing, the most recent released version of Geotoolkit, 3.19, has a bug that prevents it from working with my grd files. However, the 3.x-SNAPSHOT release seems to work just fine.

I'm using Maven for this project, so I thought it would be handy to post the pom.xml for reference. Yes, it's long. Yes, it's XML. Yes, I know you hate Maven and prefer Ant/Sbt/Gradle/Buildr/Make. Get over it. I'm posting the pom because it shows the dependencies needed to run the code samples below.

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
    <modelVersion>4.0.0</modelVersion>

    <groupId>org.mbari</groupId>
    <artifactId>geotools</artifactId>
    <version>1.0-SNAPSHOT</version>
    <packaging>jar</packaging>

    <name>geotools</name>
    <url>http://maven.apache.org</url>

    <properties>
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
        <geotoolkit.version>3.x-SNAPSHOT</geotoolkit.version>
        <scala.version>2.9.1</scala.version>
    </properties>

    <dependencies>


        <!-- GIS Referencing libraries -->
        <dependency>
            <groupId>org.geotoolkit</groupId>
            <artifactId>geotk-referencing</artifactId>
            <version>${geotoolkit.version}</version>
        </dependency>
        <dependency>
            <groupId>org.geotoolkit</groupId>
            <artifactId>geotk-epsg</artifactId>
            <version>${geotoolkit.version}</version>
        </dependency>

        <!-- JDBC Driver required for GIS referencing libraries -->
        <dependency>
         <groupId>org.apache.derby</groupId>
         <artifactId>derbyclient</artifactId>
         <version>10.8.2.2</version>
        </dependency>
        <dependency>
            <groupId>org.apache.derby</groupId>
            <artifactId>derbynet</artifactId>
            <version>10.8.2.2</version>
        </dependency>


        <!-- GIS Coverage -->
        <dependency>
            <groupId>org.geotoolkit</groupId>
            <artifactId>geotk-coverage</artifactId>
            <version>${geotoolkit.version}</version>
        </dependency>
        <dependency>
            <groupId>org.geotoolkit</groupId>
            <artifactId>geotk-coverageio</artifactId>
            <version>${geotoolkit.version}</version>
        </dependency>
        <dependency>
            <groupId>org.geotoolkit</groupId>
            <artifactId>geotk-coverageio-netcdf</artifactId>
            <version>${geotoolkit.version}</version>
        </dependency>

        <!-- Scala dependencies -->
        <dependency>
            <groupId>org.scala-lang</groupId>
            <artifactId>scala-library</artifactId>
            <version>${scala.version}</version>
        </dependency>
        <dependency>
            <groupId>org.scala-lang</groupId>
            <artifactId>jline</artifactId>
            <version>${scala.version}</version>
            <scope>runtime</scope>
        </dependency>

        <!-- Logger via SLF4J -->
        <dependency>
            <groupId>org.slf4j</groupId>
            <artifactId>slf4j-api</artifactId>
            <version>1.6.4</version>
            <scope>compile</scope>
        </dependency>
        <dependency>
            <groupId>ch.qos.logback</groupId>
            <artifactId>logback-classic</artifactId>
            <version>1.0.0</version>
            <scope>test</scope>
        </dependency>

        <!-- Ubiquitous JUnit dependency -->
        <dependency>
            <groupId>junit</groupId>
            <artifactId>junit</artifactId>
            <version>4.10</version>
            <scope>test</scope>
        </dependency>
    </dependencies>

    <repositories>
        <repository>
            <id>maven.geotoolkit.org/</id>
            <name>Geotoolkit repository</name>
            <url>http://maven.geotoolkit.org/</url>
        </repository>
        <repository>
            <id>scala-tools.org</id>
            <name>Scala-Tools Maven2 Repository</name>
            <url>http://scala-tools.org/repo-releases</url>
            <releases>
                <enabled>true</enabled>
            </releases>
            <snapshots>
                <enabled>false</enabled>
            </snapshots>
        </repository>
    </repositories>

    <pluginRepositories>
        <pluginRepository>
            <id>scala-tools.org</id>
            <name>Scala-Tools Maven2 Repository</name>
            <url>http://scala-tools.org/repo-releases</url>
        </pluginRepository>
    </pluginRepositories>

    <build>
        <sourceDirectory>src/main/scala</sourceDirectory>
        <testSourceDirectory>src/test/scala</testSourceDirectory>
        <plugins>
            <plugin>
                <groupId>org.apache.maven.plugins</groupId>
                <artifactId>maven-compiler-plugin</artifactId>
                <version>2.3.2</version>
                <configuration>
                    <encoding>${project.build.sourceEncoding}</encoding>
                    <source>1.6</source>
                    <target>1.6</target>
                    <optimize>true</optimize>
                </configuration>
            </plugin>
            <plugin>
                <groupId>org.scala-tools</groupId>
                <artifactId>maven-scala-plugin</artifactId>
                <version>2.15.2</version>
                <executions>
                    <execution>
                        <goals>
                            <goal>compile</goal>
                            <goal>testCompile</goal>
                        </goals>
                        <configuration>
                            <scalaVersion>${scala.version}</scalaVersion>
                            <encoding>${project.build.sourceEncoding}</encoding>
                            <charset>${project.build.sourceEncoding}</charset>
                            <args>
                                <arg>-optimise</arg>
                                <arg>-unchecked</arg>
                                <arg>-deprecation</arg>
                                <arg>-dependencyfile</arg>
                                <arg>${project.build.directory}/.scala_dependencies</arg>
                            </args>
                        </configuration>
                    </execution>
                </executions>
            </plugin>
            <plugin>
                <groupId>org.apache.maven.plugins</groupId>
                <artifactId>maven-surefire-plugin</artifactId>
                <version>2.9</version>
                <configuration>
                    <parallel>classes</parallel>
                    <argLine>-Xmx1024m</argLine>
                    <encoding>${project.build.sourceEncoding}</encoding>
                    <forkMode>once</forkMode>
                    <useFile>false</useFile>
                    <disableXmlReport>true</disableXmlReport>
                    <!-- If you have classpath issue like NoDefClassError,... -->
                    <!-- useManifestOnlyJar>false</useManifestOnlyJar -->
                    <includes>
                        <include>**/*Test.*</include>
                        <include>**/*Suite.*</include>
                    </includes>
                </configuration>
            </plugin>
        </plugins>
    </build>
</project>

Reading a Grid

First, I need to read the gridded data so I can work with it. The CDL for my dataset is:

netcdf BIGGRID_UTM50 {
dimensions:
 x = 5083 ;
 y = 4067 ;
variables:
 double x(x) ;
  x:long_name = "Easting (meters)" ;
  x:actual_range = 422985.277598382, 677085.277598383 ;
 double y(y) ;
  y:long_name = "Northing (meters)" ;
  y:actual_range = 3915514.90677777, 4118814.90677777 ;
 float z(y, x) ;
  z:long_name = "Topography (m)" ;
  z:actual_range = -4032.438f, 964.f ;

// global attributes:
  :Conventions = "COARDS/CF-1.0" ;
  :description = "\tProjection: UTM10N\n",
   "\tGrid created by: GMTGridReader.scala" ;
  :CreationDate = "20110214T193537Z" ;
}

The Scala code to read the file is:

import java.io.File
import org.geotoolkit.coverage.io.CoverageIO

val file = new File("BIGGRID_UTM50.grd")
// read returns GridCoverage, but really it's a GridCoverage2D
val coverage = CoverageIO.read(file).asInstanceOf[GridCoverage2D]

Unfortunately, the coordinate reference system, or CRS, isn't correct. How do I know that? You can examine the CRS like so:

println(coverage.getCoordinateReferenceSystem)
// NetCDF:"x y"

We need to force it to be UTM Zone 10N. If you're using a different zone, you can look up the CRS code at http://www.epsg-registry.org/. Here's how I'm forcing the zone:

import org.geotoolkit.coverage.CoverageFactoryFinder
import org.geotoolkit.referencing.CRS


// Be warned. This fires up an Apache Derby database in the background
val utmZone10N = CRS.decode("EPSG:32610")
val gridCoverageFactory = CoverageFactoryFinder.getGridCoverageFactory(null)
val sourceGeometrey = coverage.getGridGeometry
val utmCoverage = gridCoverageFactory.create(
            coverage.getName,               // The grid coverage name
            coverage.getRenderedImage,      // The pixel values
            utmZone10N,                     // The coordinate reference system
            sourceGeometrey.getGridToCRS(), // The transform from pixel to geodetic coordinates
            coverage.getSampleDimensions,   // Information about each band
            null,                           // The sources (Not relevant here)
            null)                           // Optional properties


// View CRS
println(utmCoverage.getCoordinateReferenceSystem)

/*
PROJCS["WGS 84 / UTM zone 10N", 
  GEOGCS["WGS 84", 
    DATUM["WGS84", 
      SPHEROID["WGS 84", 6378137.0, 298.257223563, AUTHORITY["EPSG","7030"]], 
      AUTHORITY["EPSG","6326"]], 
    PRIMEM["Greenwich", 0.0, AUTHORITY["EPSG","8901"]], 
    UNIT["degree", 0.017453292519943295], 
    AXIS["Geodetic latitude", NORTH], 
    AXIS["Geodetic longitude", EAST], 
    AUTHORITY["EPSG","4326"]], 
  PROJECTION["Transverse Mercator", AUTHORITY["EPSG","9807"]], 
  PARAMETER["central_meridian", -123.0], 
  PARAMETER["latitude_of_origin", 0.0], 
  PARAMETER["scale_factor", 0.9996], 
  PARAMETER["false_easting", 500000.0], 
  PARAMETER["false_northing", 0.0], 
  UNIT["metre", 1.0], 
  AXIS["Easting", EAST], 
  AXIS["Northing", NORTH], 
  AUTHORITY["EPSG","32610"]]
*/

Coming in Part 2

In part 2, I'll walk through accessing data values from the grid.

PartialFunction in Scala

2012-01-05T16:51:00.000-08:00

From: Weblogs Forum - Scala, Patterns and The Perl Effect

I came across this comment thread on Bruce Eckel's blog. Trond Olsen's comments provided an 'aha' moment that helped me understand PartialFunctions:

Partial functions are defined for specific value ranges. The example just showed how you can compose two partial functions, one on {"1","2","3"} and another on {1,2,3}, into one of the union of those. Partial functions gives a runtime error when you hit outside their defined value ranges.

And here's the code:

val f1: PartialFunction[Any,String] = {
  case "1" => "One"
  case "2" => "Two"
  case "3" => "Three"
}

val f2: PartialFunction[Any,String] = {
  case 1 => "One"
  case 2 => "Two"
  case 3 => "Three"
}

val f3 = f1 orElse f2

f3("1")
// res0: String = One

f3(1)
// res1: String = One

f3("3")
// res2: String = Three

f3("a")
// scala.MatchError: a (of class java.lang.String)

And here's further extension of the example:

val f4: PartialFunction[Any,String] = {
  case _ => "undefined"
}

val f5 = f3 orElse f4

f5("a")
// res3: String = undefined

Settings in Sublime Text 2 get Clobbered on Updates

2012-01-05T16:03:00.000-08:00

I've been playing around with different text editors lately. Although TextMate 2 Alpha is finally out and is pretty nice, I've been playing around with Sublime Text 2. I've kept my eye on it for a while now and it just wasn't quite there yet. However, in the last few weeks it's reached a state where it's gotten very usable. However, when it self-updates some of my settings get lost. So I thought, I'd write them down for future reference:

Base File.sublime-settings

{
 "color_scheme": "Packages/Theme - Hohonuuli/Hohonuuli.tmTheme",
 "caret_style": "phase",
 "find_selected_text": true,
 "font_face": "Ubuntu Mono",
 "font_size": 14.0,
 "indent_guide_options": ["draw_normal", "draw_active"],
 "translate_tabs_to_spaces": false,
 "wide_caret": true, /* hidden option */
 "word_wrap": false
}

Default (OSX).sublime-keymap

[
 { "keys": ["ctrl+w"], "command": "expand_selection", "args": {"to": "word"} },
 { "keys": ["ctrl+l"], "command": "expand_selection", "args": {"to": "line"} },
 { "keys": ["ctrl+b"], "command": "expand_selection", "args": {"to": "scope"} },
 { "keys": ["super+d"], "command": "run_macro_file", "args": {"file": "Packages/Default/Delete Line.sublime-macro"} },
 { "keys": ["super+shift+o"], "command": "prompt_add_folder"}
]

Global.sublime-settings

{
 "highlight_modified_tabs": true,
 "hot_exit": false,
 "ignored_packages":
 [
  "Vintage"
 ],
 "remember_open_files": true,
 "theme": "Soda Dark.sublime-theme",
 "word_wrap": false
}

Word Highlight.sublime-settings

{
 "color_scope_name" : "wordhighlight",
 "draw_outlined": true,
 "selection_delay": 0.04,
 "highlight_when_selection_is_empty": true,
 "highlight_word_under_cursor_when_selection_is_empty": false,
 "file_size_limit": 4194304
}

BracketHighlighter.sublime-settings

{
 //Enable bracket matching on type
 "quote_enable" : true,
 "curly_enable" : true,
 "round_enable" : true,
 "square_enable": true,
 "angle_enable" : true,
 //Higlight tags if tag is selected. angle_enable must be true for tags to work
 //if you would like to ignore non tags then set ignore_non_tags to true
 "tag_enable"   : true,

 //don't show non tag angle matches
 "ignore_non_tags" : false,

 //match brackets only when the cursor is adjacent to one
 "match_adjacent_only" : false,

 //exclude or include bracket by language type (example: C++ for C++.tmLanguage)
 //Note: this is particularly useful with Angle. Some languages do not attribute any special scope to angle brackets making it hard
 //to tell when the bracket is beign used in a valid scenario.  C++ is a perfect example.  It is better to disable completely than
 //to ignore angle.  Ignoring will not show angle matches, but will also interfere with other bracket matches.
 "angle_language_list" : ["HTML","HTML 5","XML","PHP"],
 "tag_language_list"   : ["HTML","HTML 5","XML","PHP"],
 "quote_language_list" : ["Plain text"],
 "curly_language_list" : ["Plain text"],
 "round_language_list" : ["Plain text"],
 "square_language_list": ["Plain text"],

 //filter type (whitelist|blacklist)
 "angle_language_filter" : "whitelist",
 "tag_language_filter"   : "whitelist",
 "quote_language_filter" : "blacklist",
 "curly_language_filter" : "blacklist",
 "round_language_filter" : "blacklist",
 "square_language_filter": "blacklist",

 //Scope? (Defined in theme files.) ->
 //Examples: (keyword/string/number)
 "quote_scope" : "entity.name.class",
 "curly_scope" : "curlybrackets",
 "round_scope" : "curlybrackets",
 "square_scope": "curlybrackets",
 "angle_scope" : "curlybrackets",
 "tag_scope"   : "entity.name.class",

 //Outline? (solid/outline/underline/none) ->
 "quote_style" : "solid",
 "curly_style" : "solid",
 "round_style" : "solid",
 "square_style": "solid",
 "angle_style" : "solid",
 "tag_style"   : "solid",

 //Icon? (dot/circle/bookmark/cross)
 "quote_icon" : "dot",
 "curly_icon" : "dot",
 "round_icon" : "dot",
 "square_icon": "dot",
 "angle_icon" : "dot",
 "tag_icon"   : "dot",

 //Higlight brackets of tag only instead of entire tag
 "tag_brackets_only" : false,

 //Type of tag selection? (html/xhtml)
 "tag_type" : "html",

 //threshold distance to search
 "search_threshold"     : 2000,
 "tag_search_threshold" : 2000,
 "use_search_threshold"     : true,
 "tag_use_search_threshold" : true,

 //Debounce delay in miliseconds
 "debounce_delay" : 100
}

Screenshot

Functor — incompatible types

2011-12-06T09:05:00.000-08:00

Here's a handy tidbit for handling types in Functors. From: Functor — incompatible types - Google Groups:

trait Functor[F[_],A] { 
  def map[B](f: A => B): F[B] 
}
 
case class Box[T](val content: T) extends Functor[Box, T] {
  def map[B](f: T => B): Box[B] = Box(f(content))
  override def toString = "Box(" + content + ")"
}

Command line Twitter search in Groovy and Scala

2011-10-24T14:18:00.001-07:00

I found a nice post about doing command line searches of Twitter using Groovy. In a nutshell the code was this: twitsearch.groovy

#!/usr/bin/env groovy
/*
  From http://www.ibm.com/developerworks/java/library/j-groovy09299/
 */
 
if (args) {
    def queryterm = args[0]
    def q = "http://search.twitter.com/search.atom?q=${queryterm}"
    def feed = new XmlSlurper().parse(q)
    feed.entry.each { entry ->
        println  """\
${entry.author.name} - ${entry.published}
\"${entry.title}\"
${'-' * 72}"""
    }
}
else {
    println  "USAGE: twitsearch.groovy "
}

I tried the same thing in Scala: twitsearch.scala

#!/bin/sh
exec scala -savecompiled "$0" "$@"
!#

if (args.size > 0) {
    import java.net.{URLConnection, URL}
    import scala.xml._
    val queryterm = args(0)
    val q = "http://search.twitter.com/search.atom?q=" + queryterm
    val feed = XML.load(new URL(q))
    feed \ "entry" foreach { entry =>
        println((entry \\ "author" \\ "name" text) + " - " + (entry \\ "published" text))
        println("\"" + (entry \\ "title" text) + "\"")
        println("-" * 72 )
    }
}
else {
    println("USAGE: twitsearch.scala ")
}

Groovy Script for Dumping NCML

2011-10-24T14:17:00.000-07:00

If you work with NetCDF and want to see the NcML you can create a groovy script like:

#!/usr/bin/env groovy

@Grab(group = 'edu.ucar.netcdf', module = 'netcdf-all', version = '4.0')
def nc = ucar.nc2.dataset.NetcdfDataset.openDataset(args[0])
nc.writeNcML(System.out, '')

Extracting Image metadata using Java/Groovy

2010-10-07T12:57:00.000-07:00

Extracting Image metadata using Java/Groovy

I'm currently working on a project that wants to use lossless PNG images for an image archive format. The format has to be able to store metadata about the image. We've found that the EXIFTool will put the metadata in PNG's using XMP XML. The trick is trying to read that metadata back out. Using Sanselan, it's possible to access this metadata. The script below will dump out all the metadata I'm able to extract from the image: Note: The Grab annotation that's fetching Sanselan requires you to configure Grapes to pull from a repository that actually contains Sanselan. There's one such repo at http://code.google.com/p/mbari-maven-repository/

#!/usr/bin/env groovy 

@Grab(group='org.apache.commons', module='sanselan', version='0.97')
import javax.imageio.ImageIO
import javax.imageio.ImageReader
import javax.imageio.metadata.IIOMetadata
import javax.imageio.stream.ImageInputStream
import org.apache.sanselan.ImageReadException
import org.apache.sanselan.Sanselan
import org.apache.sanselan.common.IImageMetadata
import org.apache.sanselan.common.RationalNumber
import org.apache.sanselan.formats.jpeg.JpegImageMetadata
import org.apache.sanselan.formats.tiff.TiffField
import org.apache.sanselan.formats.tiff.TiffImageMetadata
import org.apache.sanselan.formats.tiff.constants.GPSTagConstants
import org.apache.sanselan.formats.tiff.constants.TagInfo
import org.apache.sanselan.formats.tiff.constants.TiffConstants
import org.w3c.dom.NamedNodeMap
import org.w3c.dom.Node

def file = new File(args[0])
println("file: ${file.getPath()}")
println("---- Extracting metadata using Sanselan")
printSanselan(file)
println("---- Extracting XMP XML using Sanselan")
printSanselanXMP(file)
println("---- Extracting metadata using Java ImageIO")
printImageIO(file)



void printImageIO(File file) {
    
    ImageInputStream input = ImageIO.createImageInputStream(file)
    Iterator readers = ImageIO.getImageReaders(input)
    while (readers.hasNext()) {
        ImageReader reader = readers.next()
        reader.setInput(input, true)
        IIOMetadata metadata = reader.getImageMetadata(0)
        String[] names = metadata.metadataFormatNames
        for (name in names) {
            println("")
            println("-> Format name: ${name}" )
            printImageIOMetadata(metadata.getAsTree(name))
        }
    }
    input.close()
}

void printImageIOMetadata(Node node, int level = 0) {
    print("    " * level)
    print("<${node.nodeName}")
    NamedNodeMap map = node.attributes
    if (map) {
        for (i in 0..")
        while(child) {
             printImageIOMetadata(child, level + 1)
             child = child.nextSibling
        }
        print("    " * level)
        println("")
    }
    else {
        println(" />")
    }
    
}



// ---- Using Sanselan ------------------------------------------------
void printSanselan(File file) {
    //        get all metadata stored in EXIF format (ie. from JPEG or TIFF).
    //            org.w3c.dom.Node node = Sanselan.getMetadataObsolete(imageBytes);
    IImageMetadata metadata = Sanselan.getMetadata(file)
    if (!metadata) {
        println("\tNo image metadata was found")
        return
    }
    
    if (metadata instanceof JpegImageMetadata) {
        printSanselanJpegMetadata(metadata)
    }
    else {
        printSanselanOtherMetadata(metadata)
    }
}

void printSanselanJpegMetadata(JpegImageMetadata metadata) {
    JpegImageMetadata jpegMetadata = metadata

    // Jpeg EXIF metadata is stored in a TIFF-based directory structure
    // and is identified with TIFF tags.
    // Here we look for the "x resolution" tag, but
    // we could just as easily search for any other tag.
    //
    // see the TiffConstants file for a list of TIFF tags.

    // print out various interesting EXIF tags.
    println("  -- Standard EXIF Tags")
    printTagValue(jpegMetadata, TiffConstants.TIFF_TAG_XRESOLUTION)
    printTagValue(jpegMetadata, TiffConstants.TIFF_TAG_DATE_TIME)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_DATE_TIME_ORIGINAL)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_CREATE_DATE)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_ISO)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_SHUTTER_SPEED_VALUE)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_APERTURE_VALUE)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_BRIGHTNESS_VALUE)

    // simple interface to GPS data
    TiffImageMetadata exifMetadata = jpegMetadata.exif
    println("  -- GPS Info (using ${exifMetadata.getClass().getName()})")
    if (exifMetadata) {
        TiffImageMetadata.GPSInfo gpsInfo = exifMetadata.getGPS()
        if (gpsInfo) {
            double longitude = gpsInfo.longitudeAsDegreesEast
            double latitude = gpsInfo.latitudeAsDegreesNorth

            println("        GPS Description: ${gpsInfo}")
            println("        GPS Longitude (Degrees East): ${longitude}")
            println("        GPS Latitude (Degrees North): ${latitude}")
            
        }
        def tags = GPSTagConstants.ALL_GPS_TAGS
        tags.each { tag ->
            def field = exifMetadata.findField(tag)
            if (field) {
                def item = new TiffImageMetadata.Item(field)
                println("        ${tag.description}${item.text}")
            }
        }
        /*println("  -- All Tiff Info (using ${exifMetadata.getClass().getName()})")
        def directories = exifMetadata.directories.sort { it.toString() }
        directories.each { dir ->
            def fields = dir.allFields.sort { it.tagInfo.description }
            fields.each { field ->
                def item = new TiffImageMetadata.Item(field)
                println("        ${dir} -> ${field.tagInfo.description}${item.text}") 
            }
        } */
    }

    println("  -- All EXIF info (using ${metadata.getClass().getName()})")
    def items = jpegMetadata.items.sort { it.keyword } // List
    items.each { item ->
       println("        ${item}")   
    }
    println("")
}

void printSanselanOtherMetadata(IImageMetadata metadata) {
    println("  -- All image metadata info (using ${metadata.getClass().getName()})")
    def items = metadata.items.sort { it.keyword } // List
    items.each { item ->
       println("        ${item}")   
    }
    println("")
}

void printSanselanXMP(File file) {
    String xml = Sanselan.getXmpXml(file)
    if (xml) {
        println(xml)
    }
    println("")
}

void printTagValue(JpegImageMetadata jpegMetadata, TagInfo tagInfo) {
    TiffField field = jpegMetadata.findEXIFValue(tagInfo);
    if (field == null) {
        println("        (${tagInfo.name} not found.)")
    }
    else {
        println("        ${tagInfo.name}: ${field.valueDescription}")
    }
}

Maven Versions Plugin

2010-08-20T09:03:00.000-07:00

Maven Versions Plugin

I came across this article (at http://www.wakaleo.com/blog/281-managing-version-numbers-in-maven-with-the-maven-versions-plugin that discusses a Maven plugin that can assist you with updating the dependencies in a Maven project to more recent versions. The short guide is:

mvn versions:display-dependency-updates: Lists the dependencies you are using and which ones have an update available
mvn versions:display-plugin-updates: Same as above but looks at the plugins that your project is using
mvn versions:set -DnewVersion=1.0.2-SNAPSHOT: Changes your projects version number. Here we're setting it to 1.0.2-SNAPSHOT
mvn versions:use-latest-versions: Updates all dependencies to the latest version
- mvn versions:use-latest-releases: Replaces SNAPSHOT dependencies with latest release
- mvn versions:use-releases: Replaces SNAPSHOT dependencies with the corresponding release version
- versions:use-next-snapshots: Use the next available SNAPSHOT release
- versions:use-latest-snapshots: Use the bleeding edge (i.e. the latest SNAPSHOT release
mvn versions:commit: Commits any changes you made with mvn versions to your pom and removes the backup files that mvn versions created.
mvn versions:revert: Rollback any changes you made with the mvn versions to it's previous state.

Safe-dereference operator in Scala

2010-07-15T11:48:00.000-07:00

I've always loved Groovy's null safe operators (i.e ?.) and wanted something similar in Scala. In keeping with Scala's preference of using Option for managing null's, this little object seems to do the trick:

object SafeOption { 
    def apply[A](x: => A):Option[A] = {
        try { 
            Option(x) 
        } 
        catch { 
            case _ => None 
        }
    }
}

This allows me to try pretty much anything and have it safely handled. For example:

// Straight Scala
scala> val a = Option(Nil.head)
java.util.NoSuchElementException: head of empty list
 at scala.collection.immutable.Nil$.head(List.scala:386)
 at .(:5)
 at .()
 at RequestResult$.(:9)
 at RequestResult$.()
 at RequestResult$scala_repl_result()
 at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
 at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
 at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
 at java.lang.reflect.Method.invoke(Method.java:597)
 at scala.tools.nsc.Interpreter$Request$$anonfun$loadAndRun$1$$anonfun$apply$18.apply(Interpreter.scala:981)
 at scala.tools.nsc.Interpreter$Request$$anonfun$loadAndRun$1$$anonfun$apply$18.apply(Interpreter.scala:981)
 at s...


// With 'safe-dereferencing'
scala> val a = SafeOption(Nil.head)
a: Option[Nothing] = None

Calling Clojure from Java

2010-06-16T19:11:00.000-07:00

Calling Clojure from Java

From rzezeski on the clojure mailing list:

------ CallClojure.java ------
import clojure.lang.RT;
import clojure.lang.Var;
import clojure.lang.PersistentVector;

public class CallClojure {

   static PersistentVector toVec(int[][] arr) {
       PersistentVector pv = PersistentVector.EMPTY;
       for (int[] a : arr) {
           PersistentVector temp = PersistentVector.EMPTY;
           for (int n : a) {
               temp = temp.cons(n);
           }
           pv = pv.cons(temp);
       }
       return pv;
   }

   public static void main(String[] args) throws Exception {
       RT.loadResourceScript("foo.clj");
       Var gimmie_vec = RT.var("foo", "gimmie-vec");
       int[][] ar = {{1, 2}, {3, 4}, {5, 6}};
       Object result = gimmie_vec.invoke(toVec(ar));
       System.out.println(result);
   }
}

------ foo.clj ------
(ns foo)

(defn gimmie-vec [v]
 (println "class:" (class v))
 (println "v:" v))

Dynamically Setting the Name of the output of a Maven build

2010-04-05T16:32:00.000-07:00

Dynamically Setting the Name of the output of a Maven build

I came across this post at DZone where essential the author was complaining that he couldn't set the name of the build product in Maven without modifying the pom.xml file. In his case he wanted to pull the version number from Git. The solution was actually pretty straightforward with a little scripting help from GMaven. Here's an example


 <plugins>
   <plugin>
     <groupId>org.codehaus.groovy.maven</groupId>
     <artifactId>gmaven-plugin</artifactId>
     <executions>
       <execution>
         <id>groovy-magic</id>
         <phase>prepare-package</phase>
         <goals>
           <goal>execute</goal>
         </goals>
         <configuration>
           <source>
             def scmversion = 'git describe'.execute().text.trim()
             project.build.finalName = "${project.artifactId}-${scmversion}"
           </source>
         </configuration>
       </execution>
     </executions>
   </plugin>
 </plugins>
</build>

Working as an Independant programmer

2010-02-25T12:49:00.001-08:00

Working as an Independant programmer

Found the very informative response (below) to this post on Slashdot:

Step 1. Form an LLC. It's not hard, you can do it yourself for under $100 in most cases

Step 2. Get an EIN number from the feds. Free and easy

Step 3. Open a checking account for your new LLC. might require a credit check.

Step 4. Get a decent accounting package.

Step 5. Keep track of EVERY business expense. Keep milage logs in your car. Keep receipts. What percentage of your utilities, etc are business related? Track it.

Step 6. If you think you need the additional coverage get E&O Insurance. It can be pricey, true. On the other hand if you LLC doesn't have a lot of hard assets, why worry?

Step 7. Get health coverage. We found insurance through a local trade group for $600 a month for my wife and I. Pay it out of the company, it's a write off.

Step 8. Work your ass off and enjoy the benefits of being able to write-off things you probably would have purchased anyway.

This should have been step 6 - get a good tax guy (or girl) to help you figure shit out.

Now get creative. Like to go to theme parks? Set up another LLC and create a website dedicated to reviewing them, talking about which ones have what etc. Now you get to write off trips to Six Flags and Cedar point as legitimate business research.

Life is far more enjoyable when you do what you want, when you want, for whom you want. All the accounting is a pain in the ass, yes, but not as big of a pain in the ass as working for Bill Lumberg the rest of your life.

Show HTTP Headers

2010-02-19T11:42:00.000-08:00

Show HTTP Headers

Here's a little Groovy script to show the HTTP headers from a URL.

#!/usr/bin/env groovy

if (args.size() < 1) {
    println """
USAGE: 
  urlheaders.groovy  
  
ARGS:
  url = The URL to fetch headers from
  proxy = The proxy to go through. In the form of a column name:port. 
          For example "amoeba.ucsd.edu:9001"
    """
    System.exit(-1)
}

def url =  new URL(args[0])

def proxy = Proxy.NO_PROXY
if (args.size() > 1) {
    def proxyparts = args[1].split(":")
    proxy = new Proxy(Proxy.Type.HTTP, new InetSocketAddress(proxyparts[0], proxyparts[1] as int))
}

url.openConnection(proxy).headerFields.each {key, value ->
    if (key) {
        print("${key}: ")
    }
    println("${value.join(",\n ")}")
}

and a Scala version (for Scala 2.8+):

#!/bin/sh
exec scala "$0" "$@"
!#

import java.net.{InetSocketAddress, Proxy, URL}
import scala.collection.JavaConversions._

def printer(key: String, value: Iterable[String]): Unit = {
    if (key != null) {
        print(key + ": ")
    }
    println(headers(key).reduceLeft(_ + ",\n " + _))
}

if (args.length < 1) {
    println("""
USAGE: 
  urlheaders.groovy  
  
ARGS:
  url = The URL to fetch headers from
  proxy = The proxy to go through. In the form of a column name:port. 
          For example "amoeba.ucsd.edu:9001"
    """)
    System.exit(-1)
}

val url = new URL(args(0))

var proxy = Proxy.NO_PROXY
if (args.length > 1) {
    val proxyparts = args(1).split(":")
    proxy = new Proxy(Proxy.Type.HTTP, new InetSocketAddress(proxyparts(0), proxyparts(1).toInt))
}
val connection = url.openConnection()
val headers = connection.getHeaderFields()

headers.keySet.foreach(key => printer(key, headers(key)))

Getting started with JEE 6, Glassfish and Maven

2010-02-05T16:41:00.000-08:00

Getting started with JEE 6, Glassfish and Maven

I'm ramping back up on JEE development, so I'm catching up with the latest stuff in JEE 6. To get started I've created a basic Maven POM file for getting me going. The nice thing about this is that the POM will start the glassfish server to execute the tests and the shut it down when they are completed. It can also start up a glassfish instance on the command line with

mvn embedded-glassfish:run

pom.xml

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
    <modelVersion>4.0.0</modelVersion>
    <groupId>bms</groupId>
    <artifactId>jee-testing</artifactId>
    <packaging>war</packaging>
    <version>1.0-SNAPSHOT</version>
    <name>jee-testing Maven Webapp</name>
    <url>http://maven.apache.org</url>
    <properties>
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
    </properties>
    <dependencies>
        <dependency>
            <groupId>junit</groupId>
            <artifactId>junit</artifactId>
            <version>4.7</version>
            <scope>test</scope>
        </dependency>
        <dependency>
            <groupId>javax</groupId>
            <artifactId>javaee-web-api</artifactId>
            <scope>provided</scope>
            <version>6.0</version>
        </dependency>
    </dependencies>
    <build>
        <finalName>${project.artifactId}</finalName>
        <plugins>
            <plugin>
                <groupId>org.apache.maven.plugins</groupId>
                <artifactId>maven-compiler-plugin</artifactId>
                <configuration>
                    <source>1.6</source>
                    <target>1.6</target>
                </configuration>
            </plugin>
            <plugin>
                <groupId>org.glassfish</groupId>
                <artifactId>maven-embedded-glassfish-plugin</artifactId>
                <version>3.0</version>
                <configuration>
                    <goalPrefix>embedded-glassfish</goalPrefix>
                    <app>${basedir}/target/${project.artifactId}</app>
                    <port>8080</port>
                    <contextRoot>${project.artifactId}</contextRoot>
                    <autoDelete>true</autoDelete>
                </configuration>
                <!-- Configure glassfish for executing tests -->
                <executions>
                    <execution>
                        <id>start-glassfish</id>
                        <phase>pre-integration-test</phase>
                        <goals>
                            <goal>start</goal>
                        </goals>
                    </execution>
                    <execution>
                        <id>glassfish-deploy</id>
                        <phase>pre-integration-test</phase>
                        <goals>
                            <goal>deploy</goal>
                        </goals>
                    </execution>
                    <execution>
                        <id>glassfish-undeploy</id>
                        <phase>post-integration-test</phase>
                        <goals>
                            <goal>undeploy</goal>
                        </goals>
                    </execution>
                    <execution>
                        <id>stop-glassfish</id>
                        <phase>post-integration-test</phase>
                        <goals>
                            <goal>stop</goal>
                        </goals>
                    </execution>
                </executions>
            </plugin>
        </plugins>
    </build>
    <pluginRepositories>
        <pluginRepository>
            <id>maven2.java.net</id>
            <name>Java.net Repository for Maven 2</name>
            <url>http://download.java.net/maven/glassfish</url>
        </pluginRepository>
    </pluginRepositories>
</project>

Maven and Clojure

2010-01-17T22:49:00.000-08:00

Maven and Clojure

I'm trying to setup a Maven build for a Clojure project. I'm playing around with Incanter and need to a) package up the project and b) have easy REPL interaction. The following POM seems to be working so far:

<project 
  xmlns="http://maven.apache.org/POM/4.0.0"
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
  xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 
    http://maven.apache.org/maven-v4_0_0.xsd">
    <modelVersion>4.0.0</modelVersion>
    <groupId>org.mbari</groupId>
    <artifactId>clojuretest</artifactId>
    <packaging>jar</packaging>
    <version>1.0-SNAPSHOT</version>
    <name>clojuretest [library]</name>
    <url>http://maven.apache.org</url>
    
    <!-- 
        Combining info from:
            Packaging: http://pupeno.com/blog/how-to-create-a-clojure-application/
            Command line integration: http://github.com/talios/clojure-maven-plugin
    -->
            
    <build>
        <resources>
            <resource>
                <directory>src/main/clojure</directory>
            </resource>
        </resources>

        <plugins>
            <!-- Filter resources using UTF-8 encoding -->
            <plugin>
                <artifactId>maven-resources-plugin</artifactId>
                <configuration>
                    <encoding>UTF-8</encoding>
                </configuration>
            </plugin>
            <!-- Creates executable jar -->
            <plugin>
                <groupId>org.codehaus.mojo</groupId>
                <artifactId>appassembler-maven-plugin</artifactId>
                <configuration>
                    <programs>
                        <program>
                            <mainclass>org.mbari.clojuretest.App</mainclass>
                            <name>App</name>
                        </program>
                    </programs>
                </configuration>
            </plugin>
            <!--  Integrate Clojure command line tools -->
            <plugin>
                <groupId>com.theoryinpractise</groupId>
                <artifactId>clojure-maven-plugin</artifactId>
                <version>1.3.1</version>
                <configuration>
                    <clojureOptions>-Xmx512m</clojureOptions>
                    <warnOnReflection>true</warnOnReflection>
                </configuration>
            </plugin>
        </plugins>
    </build>

    <dependencies>
        <dependency>
            <groupId>junit</groupId>
            <artifactId>junit</artifactId>
            <version>4.7</version>
            <scope>test</scope>
        </dependency>


        <dependency>
            <groupId>incanter</groupId>
            <artifactId>incanter</artifactId>
            <version>1.0-master-SNAPSHOT</version>
        </dependency>

    </dependencies>

    <repositories>
        <repository>
            <id>clojars.org</id>
            <url>http://clojars.org/repo</url>
        </repository>
    </repositories>
</project>

Viewing Groovy's Classpath

2010-01-11T08:51:00.000-08:00

Viewing Groovy's Classpath

I came across this nice tip found here. The tip is:

this.class.classLoader.rootLoader.URLs.each{ println it }

EXIF Metadata Extraction in Clojure

2010-01-08T12:28:00.000-08:00

EXIF Metadata Extraction in Clojure

Since I've been playing around with Clojure lately, I thought I'd port the Groovy and Scala versions of the EXIF metadata extraction code to Clojure. Here's some things I ran into with Clojure:

Have to admit, although I'm very new at the Lisp syntax, Clojure is kinda fun to work with. As with both Groovy and Scala it's nice to have a REPL to test out new ideas out. The Clojure REPL's built in help (e.g. (doc instance?) and (find-doc "instance") is a very nice touch.
I'm not a fan of the numerous parenthesis, I suppose I'm just used to the curly brackets. I find it's hard to visualize the scope of variables. A text editor with nice bracket matching helps though (Sorry TextMate, I had to use jEdit for Clojure work)
Writing a script still left me jonesing for an equivalent to Grape
The clj script used in the shebang line, posted on Wikibooks, was wrong. Took me a little bit to figure out what was going on. I fixed the script on Wikibooks.
I couldn't figure out how to neatly embed the classpath info, so I had to launch it with clj -cp /path/to/sanselan-0.97.jar readexif.clj /path/to/image.jpg
I have absolutely no idea how to indent Clojure/Lisp code

And the code:

#! /usr/bin/env clj

(import
  '(java.io File)
  '(org.apache.sanselan ImageReadException Sanselan)
  '(org.apache.sanselan.common IImageMetadata, RationalNumber)
  '(org.apache.sanselan.formats.jpeg JpegImageMetadata)
  '(org.apache.sanselan.formats.tiff TiffField TiffImageMetadata)
  '(org.apache.sanselan.formats.tiff.constants ExifTagConstants GPSTagConstants TagInfo TiffConstants TiffTagConstants))
  
(defn print-tag-value [metadata tagInfo]
  (let [field  (. metadata findEXIFValue tagInfo)]
    (if (nil? field) 
      (println (str "        (" (. tagInfo name) " not found.)"))
      (println (str "        (" (. tagInfo name) ": " (. field getValueDescription))))))
    
(defn readexif [file]
  (let [metadata (Sanselan/getMetadata file)]
    (if (nil? metadata) (println "\tNo EXIF metdata was found"))
    (if (instance? JpegImageMetadata metadata)
      (do 
        (println "  -- Standard EXIF Tags")
        (print-tag-value metadata (TiffTagConstants/TIFF_TAG_XRESOLUTION))
        (print-tag-value metadata (TiffTagConstants/TIFF_TAG_DATE_TIME))
        (print-tag-value metadata (ExifTagConstants/EXIF_TAG_DATE_TIME_ORIGINAL))
        (print-tag-value metadata (ExifTagConstants/EXIF_TAG_CREATE_DATE))
        (print-tag-value metadata (ExifTagConstants/EXIF_TAG_ISO))
        (print-tag-value metadata (ExifTagConstants/EXIF_TAG_SHUTTER_SPEED_VALUE))
        (print-tag-value metadata (ExifTagConstants/EXIF_TAG_APERTURE_VALUE))
        (print-tag-value metadata (ExifTagConstants/EXIF_TAG_BRIGHTNESS_VALUE))
        (print-tag-value metadata (GPSTagConstants/GPS_TAG_GPS_LATITUDE_REF))
        (print-tag-value metadata (GPSTagConstants/GPS_TAG_GPS_LATITUDE))
        (print-tag-value metadata (GPSTagConstants/GPS_TAG_GPS_LONGITUDE_REF))
        (print-tag-value metadata (GPSTagConstants/GPS_TAG_GPS_LONGITUDE_REF))
        
        ; simple interface to GPS data
        (println "  -- GPS Info")
        (let [exifMetadata (. metadata getExif)]
          (if (false? (nil? exifMetadata)) 
            (do 
              (let [gpsInfo (. exifMetadata getGPS)]
                (if (false? (nil? gpsInfo))
                  (let [longitude (. gpsInfo getLongitudeAsDegreesEast)
                        latitude (. gpsInfo getLatitudeAsDegreesNorth)]
                    (println (str "        GPS Description: " gpsInfo))
                    (println (str "        GPS Longitude (Degrees East):" longitude))
                    (println (str "        GPS Latitude (Degrees North):" latitude))))))))
                  
         ; Print all EXIF data
         (println "  -- All EXIF info")
         (doseq [item (. metadata getItems)]
           (println (str "        " item)))
         (println "")))))

;(readexif (new File (first *command-line-args*)))
(do
  (let [file (new File (first *command-line-args*))]
    (println (str "file: " (. file getPath)))
    (readexif file)))

EXIF Metadata Extraction in Scala

2010-01-07T09:34:00.000-08:00

EXIF Metadata Extraction in Scala

Just for grins I ported the EXIF metadata extraction script in yesterdays post from Groovy to Scala. I ran into a few interesting gotchas while working with Scala.

The class TiffConstants inherits a ton of constants (e.g. public static final ints) from other interfaces. However, in scala I had to explicitly use each interface where the constant was defined, using TiffConstants didn't work.
There's no Grape for package management. Which is too bad, Scala would be far more useful if there was some equivalent; defining the classpath for a small script is annoying. My work around was just to embed the classpath in the shebang preamble.
Got to remember to import implicit List conversions (i.e. import scala.collection.jcl.Conversions._) in order to work with Java Lists as Scala Lists

Here's the code:

#!/bin/sh
exec scala -cp $HOME/.groovy/grapes/org.apache.commons/sanselan/jars/sanselan-0.97.jar $0 $@
!#

import java.io.File
import org.apache.sanselan.{ImageReadException, Sanselan}
import org.apache.sanselan.common.{IImageMetadata, RationalNumber}
import org.apache.sanselan.formats.jpeg.JpegImageMetadata
import org.apache.sanselan.formats.tiff.{TiffField, TiffImageMetadata}
import org.apache.sanselan.formats.tiff.constants.{ExifTagConstants, GPSTagConstants, TagInfo, TiffConstants, TiffTagConstants}
import scala.collection.jcl.Conversions._

val file = new File(args(0))
val metadata = Sanselan.getMetadata(file)
def printTagValue(metadata: JpegImageMetadata, tagInfo: TagInfo): Unit = {
    val field = metadata.findEXIFValue(tagInfo)
    field match {
        case null => println("        (" + tagInfo.name + " not found.)")
        case _ => println("        " + tagInfo.name + ": " + field.getValueDescription())
    }
}

println("file:" + file.getPath())

if (metadata == null) {
    println("\tNo EXIF metadata was found")
}

if (metadata.isInstanceOf[JpegImageMetadata]) {
    val jpegMetadata = metadata.asInstanceOf[JpegImageMetadata]
    
    println("  -- Standard EXIF Tags")
    printTagValue(jpegMetadata,  TiffTagConstants.TIFF_TAG_XRESOLUTION)
    printTagValue(jpegMetadata,  TiffTagConstants.TIFF_TAG_DATE_TIME)
    printTagValue(jpegMetadata,  ExifTagConstants.EXIF_TAG_DATE_TIME_ORIGINAL)
    printTagValue(jpegMetadata,  ExifTagConstants.EXIF_TAG_CREATE_DATE)
    printTagValue(jpegMetadata,  ExifTagConstants.EXIF_TAG_ISO)
    printTagValue(jpegMetadata,  ExifTagConstants.EXIF_TAG_SHUTTER_SPEED_VALUE)
    printTagValue(jpegMetadata,  ExifTagConstants.EXIF_TAG_APERTURE_VALUE)
    printTagValue(jpegMetadata,  ExifTagConstants.EXIF_TAG_BRIGHTNESS_VALUE)
    printTagValue(jpegMetadata,  GPSTagConstants.GPS_TAG_GPS_LATITUDE_REF)
    printTagValue(jpegMetadata,  GPSTagConstants.GPS_TAG_GPS_LATITUDE)
    printTagValue(jpegMetadata,  GPSTagConstants.GPS_TAG_GPS_LONGITUDE_REF)
    printTagValue(jpegMetadata,  GPSTagConstants.GPS_TAG_GPS_LONGITUDE) 
    
    // simple interface to GPS data
    println("  -- GPS Info")
    val exifMetadata = jpegMetadata.getExif()
    if (exifMetadata != null) {
        val gpsInfo = exifMetadata.getGPS()   
        if (gpsInfo != null) {
            val longitude = gpsInfo.getLongitudeAsDegreesEast()
            val latitude = gpsInfo.getLatitudeAsDegreesNorth()
            
            println("        GPS Description: " + gpsInfo)
            println("        GPS Longitude (Degrees East): " + longitude)
            println("        GPS Latitude (Degrees North): " + latitude)
        }
    }
    
    println("  -- All EXIF info")
    jpegMetadata.getItems().foreach(item => println("        " + item))
    println("")
}

EXIF Metadata Extraction in Groovy

2010-01-06T16:47:00.000-08:00

EXIF Metadata Extraction in Groovy

For a work project, I need to extract EXIF metadata from JPG's in Java code. The Apache Sanselan project supports reading and writing of EXIF metadata. I wrote a Groovy script as a test for extracting metadata. The script uses Grape to fetch the Sanselan dependency, however, Sanselan in not yet in one of the main Maven repos so I had to install it into our internal repository. (Just be aware that the script won't run 'out-of-the-box'). The code was just a quick port of this example:

#!/usr/bin/env groovy 

@Grab(group='org.apache.commons', module='sanselan', version='0.97')
import org.apache.sanselan.ImageReadException
import org.apache.sanselan.Sanselan
import org.apache.sanselan.common.IImageMetadata
import org.apache.sanselan.common.RationalNumber
import org.apache.sanselan.formats.jpeg.JpegImageMetadata
import org.apache.sanselan.formats.tiff.TiffField
import org.apache.sanselan.formats.tiff.TiffImageMetadata
import org.apache.sanselan.formats.tiff.constants.TagInfo
import org.apache.sanselan.formats.tiff.constants.TiffConstants

def file = new File(args[0])

//        get all metadata stored in EXIF format (ie. from JPEG or TIFF).
//            org.w3c.dom.Node node = Sanselan.getMetadataObsolete(imageBytes);
IImageMetadata metadata = Sanselan.getMetadata(file)
println("file: ${file.getPath()}")

if (!metadata) {
    println("\tNo EXIF metadata was found")
}

if (metadata instanceof JpegImageMetadata) {
    JpegImageMetadata jpegMetadata = metadata

    // Jpeg EXIF metadata is stored in a TIFF-based directory structure
    // and is identified with TIFF tags.
    // Here we look for the "x resolution" tag, but
    // we could just as easily search for any other tag.
    //
    // see the TiffConstants file for a list of TIFF tags.

    

    // print out various interesting EXIF tags.
    println("  -- Standard EXIF Tags")
    printTagValue(jpegMetadata, TiffConstants.TIFF_TAG_XRESOLUTION)
    printTagValue(jpegMetadata, TiffConstants.TIFF_TAG_DATE_TIME)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_DATE_TIME_ORIGINAL)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_CREATE_DATE)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_ISO)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_SHUTTER_SPEED_VALUE)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_APERTURE_VALUE)
    printTagValue(jpegMetadata, TiffConstants.EXIF_TAG_BRIGHTNESS_VALUE)
    printTagValue(jpegMetadata, TiffConstants.GPS_TAG_GPS_LATITUDE_REF)
    printTagValue(jpegMetadata, TiffConstants.GPS_TAG_GPS_LATITUDE)
    printTagValue(jpegMetadata, TiffConstants.GPS_TAG_GPS_LONGITUDE_REF)
    printTagValue(jpegMetadata, TiffConstants.GPS_TAG_GPS_LONGITUDE)
    


    // simple interface to GPS data
    println("  -- GPS Info")
    TiffImageMetadata exifMetadata = jpegMetadata.exif
    if (exifMetadata) {
        TiffImageMetadata.GPSInfo gpsInfo = exifMetadata.getGPS()
        if (gpsInfo) {
            double longitude = gpsInfo.longitudeAsDegreesEast
            double latitude = gpsInfo.latitudeAsDegreesNorth

            println("        GPS Description: ${gpsInfo}")
            println("        GPS Longitude (Degrees East): ${longitude}")
            println("        GPS Latitude (Degrees North): ${latitude}")
        }
    }

    println("  -- All EXIF info")
    ArrayList items = jpegMetadata.items.each { item ->
       println("        ${item}")   
    }
    println("")
}

void printTagValue(JpegImageMetadata jpegMetadata, TagInfo tagInfo) {
    TiffField field = jpegMetadata.findEXIFValue(tagInfo);
    if (field == null) {
        println("        (" + tagInfo.name + " not found.)")
    }
    else {
        println("        " + tagInfo.name + ": " + field.valueDescription)
    }
}

Examining architectures in a Fat-binary on Mac OS X

2010-01-06T12:48:00.000-08:00

Examining architectures in a Fat-binary on Mac OS X

Apple's Mac OS X supports fat binaries (i.e. multi-architecture). I've been trying to debug a problem with bad kernel access using RXTX and found this reference for the very useful tool lipo

lipo
/usr/bin/lipo creates or operates on multi-architecture ("fat") files. It can list the architecture types in a fat file, create a single fat file from one or more input files, thin out a single fat file to a specified architecture type, and extract, replace and/or remove architecture types from the input file.

To see the architectures in a binary it's used like:

lipo -info librxtxSerial.jnilib

Which returns output like:

Architectures in the fat file: librxtxSerial.jnilib are: x86_64 i386 ppc7400

Playing with JXLayer

2010-01-01T19:42:00.000-08:00

Playing with JXLayer

I started using JXLayer in a project. Here's a sample UI class that decorates a Swing componenet with an XOR'd cross that tracks the mouse:

/*
 * @(#)CrossHairLayerUI.java   2009.12.29 at 03:21:13 PST
 *
 * Copyright 2009 MBARI
 *
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */


package vars.annotation.ui.imagepanel;

import java.awt.Color;
import java.awt.Graphics2D;
import java.awt.Point;
import java.awt.event.MouseEvent;
import java.awt.geom.GeneralPath;
import javax.swing.JComponent;
import javax.swing.SwingUtilities;
import org.jdesktop.jxlayer.JXLayer;
import org.jdesktop.jxlayer.plaf.AbstractLayerUI;

/**
 * Draws a crosshair over a JComponent. Use as:
 * {@code
 * JXLayer layer = new JXLayer(component);
 * layer.setUI(new CrossHairLayerUI());
 * }
 * @author brian
 */
public class CrossHairLayerUI extends AbstractLayerUI {

    private GeneralPath crosshair = new GeneralPath();

    @Override
    protected void paintLayer(Graphics2D gd, JXLayer jxl) {
        super.paintLayer(gd, jxl);
        gd.setXORMode(Color.WHITE);
        gd.draw(crosshair);
        gd.setPaintMode();
    }

    @Override
    protected void processMouseMotionEvent(MouseEvent me, JXLayer jxl) {
        super.processMouseMotionEvent(me, jxl);

        if (me.getID() == MouseEvent.MOUSE_MOVED || me.getID() == MouseEvent.MOUSE_DRAGGED) {
            Point point = SwingUtilities.convertPoint(me.getComponent(), me.getPoint(), jxl);
            int w = jxl.getWidth();
            int h = jxl.getHeight();

            /*
             * Create crosshair
             */
            crosshair.reset();

            if (point.y <= h) {
                crosshair.moveTo(0, point.y);
                crosshair.lineTo(w, point.y);
            }

            if (point.x <= w) {
                crosshair.moveTo(point.x, 0);
                crosshair.lineTo(point.x, h);
            }

            // mark the ui as dirty and needed to be repainted
            setDirty(true);
        }
    }
}

My First Lisp Program

2009-12-13T22:11:00.000-08:00

My First Lisp Program

I wrote a simple timecode generator for testing of the Video Annotation and Reference System in Groovy. The program listens for requests on a UDP port and responds with a String formated as a timecode. For grins, I converted the program to Clojure

Groovy

#!/usr/bin/env groovy

/**
 * Script for simulation of a udp VCR
 */

def port = (args.size() > 0) ? Integer.valueOf(args[0]) : 9000

println("Localtime timecode generator attached to port ${port}")

 
def socket = new DatagramSocket(port)
def buffer = (' ' * 4096) as byte[]
while(true) {
    def incoming = new DatagramPacket(buffer, buffer.length)
    socket.receive(incoming)
    def c = new GregorianCalendar()
    def f = c.get(Calendar.MILLISECOND) / 1000.0 * 29.97;
    String reply = String.format('%tT:%02d', c, Math.round(f))
    outgoing = new DatagramPacket(reply.bytes, reply.size(),
            incoming.address, incoming.port);
    socket.send(outgoing)
}

Clojure

(import 
  '(java.net DatagramPacket DatagramSocket)
  '(java.util GregorianCalendar Calendar))
  
(defn fmt-time 
  "Format a java Calendar as a timecode"
  [cal] 
  (let [frames (* (/ (. cal get (Calendar/MILLISECOND)) 1000) 29.97)]
    (format "%tT:%02d" cal (Math/round frames))))
  
(defn run-generator 
  "Function for simulation of a udp VCR"
  [port] 
  (println (str "Localtime timecode generator attached to port " port))
  (let [socket (new DatagramSocket port)
        buffer (make-array (Byte/TYPE) 4096)]
    (while true 
      (let [incoming (new DatagramPacket buffer 4096) 
            reply (fmt-time (new GregorianCalendar))]
         (. socket receive incoming)
         (let [outgoing (new DatagramPacket (. reply getBytes) (. reply length) (. incoming getAddress) (. incoming getPort))]
           (. socket send outgoing))))))
  
 (run-generator (Integer/parseInt (first *command-line-args*)))

Eclipse template for quickly inserting a SLF4J logging statement

2009-12-03T14:27:00.001-08:00

Eclipse template for quickly inserting a SLF4J logging statement

${imp:import(org.slf4j.LoggerFactory)}
private final ${type:newType(org.slf4j.Logger)} ${log:newName(org.slf4j.Logger)} = LoggerFactory.getLogger(getClass());${cursor}

Deleteing all the files Memeo scatters across my hard drive using Groovy

2009-12-03T09:51:00.001-08:00

Deleteing all the files Memeo scatters across my hard drive using Groovy

(new File('.')).eachFileRecurse { file -> if (file.isFile() && file.name.startsWith('.Memeo ')) {file.delete()} }

Maven and Google Code

2009-06-01T16:21:00.000-07:00

Maven and Google Code

I have a few projects I keep on Google Code. Since I do quite a bit of work using Maven and Java, it's nice to be able to use the subversion repo on Google Code to host my maven repo and my javadocs. Notes for hosting a Maven repo can be found at http://www.jroller.com/mrdon/entry/find_of_the_day_wagon and Javadoc details are at http://stuffthathappens.com/blog/2007/11/09/howto-publish-javadoc-on-google-code/

2009-03-05T09:17:00.000-08:00

Adding Log4j DOCTYPE to log4j.xml

If you're using an XML aware editor, it's nice to take advantage of built in XML validation when editing log4j.xml configuration files. Here's a DOCTYPE that works with these editors:

<!DOCTYPE log4j:configuration PUBLIC 
        "-//LOG4J//DTD LOG4J 1.2//EN" 
        "http://svn.apache.org/viewvc/logging/log4j/trunk/src/main/resources/org/apache/log4j/xml/log4j.dtd?view=co">

Generating a task list from JIRA issues using XML-RPC and Groovy

2008-12-02T14:07:00.000-08:00

Generating a task list from JIRA issues using XML-RPC and Groovy

This is just a quick note demonstrating how to generate a task list from issues entered into JIRA. I use TaskPaper for creating GTD-style task lists for work and thought it might be handy to tack issues assigned to me in our JIRA issue tracker in the same way. To do so do the following:

Install Groovy
Download the Groovy XML-RPC module and copy it to $HOME/.groovy/lib
Create a filter in JIRA to return tasks that you're interested in. Make note of the filterIdfor the filter you created.

Create a shell script, here I'll call it jiratasks.groovy, and copy the following code into it:

#!/usr/bin/env groovy
import groovy.net.xmlrpc.XMLRPCServerProxy
import java.text.SimpleDateFormat

// Included to make pretty dates
def dateFormat1 = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss")
def dateFormat2 = new SimpleDateFormat("yyyy-MM-dd")

// Connect to your server
//Your URL may vary
def server = new XMLRPCServerProxy("https://your.jira.server.net/rpc/xmlrpc") 
def jira = server.jira1
// Subsitute your username and password here
def token = jira.login("your_username", "your_password")
// Replace "10021" with your filterId  
def issues = jira.getIssuesFromFilter(token, "10021")  

// Sort by priority
issues.sort { it.priority }

// Get a list of projects
def projects = issues.collect { it["project"]}
projects.unique().sort()

println "JIRA Issues:"
projects.each { project ->
  println("\t${project}:")
  issues.each { issue ->
    if (issue["project"] == project) {
      def creationDate = dateFormat2.format(dateFormat1.parse(issue.created))
      println("\t- ${issue.key}: ${issue.summary} @created(${creationDate})")
    }
  }
}

jira.logout(token)

Run it like: jiratasks.groovy > JIRATasks.taskpaper
Voila. A GTD style tasks list that works with TaskPaper

Spinning up on Google Guice

2008-06-04T09:27:00.000-07:00

Spinning up on Google Guice

I'm spinning up on using dependency-injection for a new desktop app. Here's a few references to get me going: more to come...

2008-02-21T09:06:00.001-08:00

Peformance tip for SQL: If Exists Update Else Insert

See http://blogs.msdn.com/miah/archive/2008/02/17/sql-if-exists-update-else-insert.aspx

2008-02-20T14:52:00.000-08:00

Learing Cocoa Notes

I'm working through the tutorials on Cocoa at MacResearch

Notes from 'Living Objects'

Cocoa is a framework built upon conventions. If you stick to the conventions, you will be fine; if you flout them, it will hurt.

Typical init block:

-(id)initWithLeftValue:(double)lv andRightValue:(double)rv {
    if ( self = [super init] ) {
        leftValue = lv;
        rightValue = rv;
    }
    return self;
}

RULE: You should check the return value to see if it is nil after you call an initializer.
RULE: Cocoa has the concept of a designated initializer, which is the initializer you should generally chain to from a subclass. There is no language support for the designated initializer; it is purely a convention, and, as such, typically involves no more than a statement to the effect in the comments or documentation.
In Cocoa, when an object is ready to disappear, the method dealloc gets called.

Typical pattern:

-(id)init... {
    // Chain to designated initializer of superclass
    if ( self = [super init] ) {
        ...
    }
    return self;
}

-(void)dealloc {
    ...
    [super dealloc];
}

Notes from 'Good References'

Cocoa uses a system of memory management called reference counting
Whenever an object is allocated, it automatically has a retain count of one. If you do nothing more with that object, it will remain in existence until the program exits. Other methods for creating objects, such as copy and new, also leave the object with a retain count of one.
Often, some part of your program will want to make use of an object that it did not create itself. In such cases, you want to ensure that the object in question does not disappear when it is still needed. To achieve this, the object's retain method can be called, which increases the retain count by one.
When an object is no longer needed, you call either the release method, or the autorelease method. Both methods decrease the retain count by one, but they differ in the timing of the change. release decreases the count immediatly.
When you use autorelease, you are saying you are finished with the object, but maybe some other part of the program would like to make use of it for a bit longer. An example of this is when you need to return an object from a method, and you don't need the object anymore. If you call release, it will disappear before you can return it; if you call autorelease, you have time to return the object, and the calling code has a chance to retain it if it wants to keep it around.
RULE: Whenever you call one of the retain count increasing methods, like alloc, retain, or copy, you should balance that with a call to a retain count decreasing method like release or autorelease.

Notes from 'All in the Genes'

You can actually apply constraints on the accessibility of instance variables in Objective-C, as in most object-oriented programming languages. The default, which we have been using to this point, is @protected, which means that an instance variable can be accessed from within the class in which it is defined, and from any descendent (eg subclass). The other two options are @private, which restricts access to the defining class, excluding its descendent classes, and @public, which allows free-for-all access from anywhere in the program.
In Objective-C, all methods are effectively public
You can fake a private method by declaring it not in the public header file, but in the .m implementation file.
Objective-C does not support multiple inheritance.
Protocols are blocks that declare one or methods that a class must implement. They are basically the same as interfaces in the Java language.

Protcol syntax:

// Declaration
@protocol Add
-(id)add:(id)toAdd;
@end

// Inclusion
@interface AdditionOperator : Operator <Add> {
}
@end

// Implementation
@implementation AdditionOperator
-(id)add:(id)toAdd {
    // Implementation here
    ...
}
@end

Objective-C allows overriding methods by using a method of the same name in a subclass

Creating simple 'toString()' methods for Plain Old Groovy Beans (POGO's)

2007-12-07T14:26:00.000-08:00

Creating simple 'toString()' methods for Plain Old Groovy Beans (POGO's)

Hate to write toString() methods? Me too. I created a simple Category for generating formatted strings for my beans. This is nothing fancy but it illustrates how simple it is. First, I created a Category for dynamically adding methods to my classes.

package hohonuuli

class PropertyCategory {
    
    /**
     * List the property names of an object excluding class and metaClass
     */
    static List listKeys(Object self) {
        List keys = []
        self.properties.each {
            if (it.key != "metaClass" && it.key != "class") {
                keys << it.key
            }
        }
        return keys
    }  
    
    static String toXML(Object self) {
        def stringBuilder = new StringBuilder()
        stringBuilder.append("<${self.getClass().name}>")
        use (PropertyCategory) {
            List keys = listKeys(self)
            keys.sort().each { key ->
                stringBuilder.append("<${key}>").append(self."${key}".toString()).append("</${key}>")
            }
        }
        stringBuilder.append("</${self.getClass().name}>")
    }
    
}

Next, I create a bean and use the Cateogry in the toString method. In this case, I'm just dumping everything out as an XML fragment.

import hohonuuli.PropertyCategory

class AncillaryDatum 

    Date date
    def block
    def mode
    def type
    def depthMeters
    def temperatureCelsius
    def salinity

 
    String toString() {
        def s = ""
        // Here's the magic
        use (PropertyCategory) {
            s = this.toXML()
        }
        return s
    }
}

The output looks like:

<AncillaryDatum><block>2</block><date>Tue Jun 29 21:40:23 UTC</date><depthMeters>0.0050</depthMeters><mode>1</mode><salinity>0.0097</salinity><temperatureCelsius>19.1766</temperatureCelsius><type>18</type></AncillaryDatum>

You can format the output however you like and control what properties are written out. For example, you might want to ignore transient and static fields in you output. Enjoy

AppleScript for "Open Terminal Here" command.

2007-11-17T14:06:00.000-08:00

AppleScript for "Open Terminal Here" command.

Variants of this are all over the place. I modified one to work with the new tabbed feature n Leopards Terminal.app. This applescript opens a new tab in Terminal in the directory of the current Finder window. Open it in Script Editor, compile, save as an application and then drag the application to the toolbar of a Finder window.

(*

 Open Terminal Here
 
 A toolbar script for Mac OS X 10.3
 
 Written by Marc Liyanage

 
 See http://www.apple.com/applescript/macosx/toolbar_scripts/ for
 more information about toolbar scripts.
 
 See http://www.entropy.ch/software/applescript/ for the latest
 version of this script.
 
 
 History:
 17-NOV-2007: Modified by Brian Schlining to open in a new Tab in a terminal window
 18-AUG-2004: Version 2.0 by Allan Marcus. uses posix path 
 30-OCT-2001: Version 1.0, adapted from one of the example toolbar scripts
 30-OCT-2001: Now handles embedded single quote characters in file names
 30-OCT-2001: Now handles folders on volumes other than the startup volume
 30-OCT-2001: Now handles click on icon in top-level (machine) window
 31-OCT-2001: Now displays a nicer terminal window title, courtesy of Alain Content
 11-NOV-2001: Now folders within application packages (.app directories) and has a new icon
 12-NOV-2001: New properties to set terminal columns and rows as the Terminal does not use default settings
 14-NOV-2001: Major change, now handles 8-bit characters in all shells, and quotes and spaces in tcsh
 18-NOV-2001: Version 1.1: Rewrite, now uses a temporary file  ~/.OpenTerminalHere to communicate
 the directory name between AppleScript and the shell because this is much more reliable for 8-bit characters
 
*)

property terminal_rows : 24
property terminal_columns : 90
property debug : true

-- when the toolbar script icon is clicked
--
on run
 tell application "Finder"
  activate
  try
   set this_folder to (the target of the front window) as alias
  on error
   set this_folder to startup disk
  end try
  my process_item(this_folder)
 end tell
end run

-- This handler processes folders dropped onto the toolbar script icon
--
on open these_items
 repeat with i from 1 to the count of these_items
  set this_item to item i of these_items
  my process_item(this_item)
 end repeat
end open

-- this subroutine processes does the actual work
--
on process_item(this_item)
 tell application "Terminal"
  activate
  tell application "System Events" to tell process "Terminal" to keystroke "t" using command down
  do script with command "cd " & quoted form of POSIX path of this_item & ";clear" in selected tab of the front window
 end tell
end process_item

2007-11-04T15:10:00.000-08:00

Hey Apple, bring Java 6 to Leopard!

13949712720901ForOSX

Icons hosted on Google for Google Earth, Google Maps and KML

2007-09-18T21:22:00.000-07:00

Icons hosted on Google for Google Earth, Google Maps and KML

I did a little sleuthing to find out what paddles are available on Google for marking points in Google Earth. Here's a list of the icons that Google hosts. I don't know if this is complete or not (I suspect not) but it's a pretty handy reference for those working with KML. There's a few more listed here

Large Letters and Numbers

Small Letters and Numbers

Icons

Shadow Icons

Large Paddles

Small Paddles

Miscellaneous

SLF4J Code Template for Netbeans

2007-04-24T09:32:00.000-07:00

SLF4J Code Template for Netbeans

I use Netbeans alot for Java development. I also used Loggers a great deal and Netbeans Code Template feature is a quick an easy way to insert this boiler plate code. Here's an example code template for inserting the code for creating an SLF4J logger [Note I've wrapped it in this blog, but it should all be on one line]:

private static final 
   ${LOG_TYPE type="org.slf4j.Logger" default="Logger" editable=false} 
   log = ${LOG_FACT type="org.slf4j.LoggerFactory" 
   default="LoggerFactory" editable=false}.getLogger(${CLASS});

I've added this tip to Netbeans Community Docs at http://wiki.netbeans.org/wiki/view/SLF4JCodeTemplate.

2007-04-12T13:43:00.000-07:00

Delete Line Macro in TextMate

TextMate is a great text editor. One key command that I really miss in it is a 'Delete Line' command. It's pretty simple to implement though. Here's how:

Under the Bundles menu select Macros->Start recording
Select a line in an editor window([Option]+[Shift]+L)
Delete the line ([Backspace])
Under the Bundles menu select Macros->Stop recording
Under the Bundles menu select Macros->Save last recording
Give it a name; 'Delete Line' seems most appropriate
Assign it a keystroke. I use [Option]+D.

Scripting a Java Project with Groovy

2007-03-30T09:04:00.001-07:00

Scripting a Java Project with Groovy

For java projects, after you've finished writing the the API, business logic, O/R mapping layer, etc, you have to get down to the nitty, gritty details of making the application actually run. I often find that I need to script together different API's to perform some task. For example, I might want to call some of my libraries that add a time-code track to a QuickTime movie. I could write a Java class to do this but generally this is cumbersome and creates a fairly heavy-weight development process (write-compile-test-deploy); and honestly, isn't really much fun. And even after I've written the class to do the particular task, I still need to write a shell script to execute the classes main method. For desktop applications this means creating a launcher, such as a JNLP file or shell script.
Recently, I've started poking around with Groovy . I've been very impressed with it's scripting ability. Here, I will present what I've done to script together different API's using Groovy. Below, I present a technique that allows you to script a Java applications without requiring you to install anything outside of the project. (i.e. you don't need to install groovy on the system that's running your scripts.) In order to build this project you will need to have the following installed: Java, Ant
First, I create an Ant build.xml file. This particular project doens't actually compile any code, the build file is just going to use Maven to fetch the dependencies I need for my project and store them in a lib directory. This technique is described here.
My project is set up like so:

PRJ_HOME
|~~ bin         [Shell scripts]
|~~ conf        [Configuration - e.g. log4j.xml]
|~~ lib         [Java jars]
`~~ scripts    
    `~~ groovy  [Groovy scripts]

My build.xml looks like this:

<?xml version="1.0"?>
<!-- build file for eits-db program -->
<!-- Created by:  Brian Schlining -->

<project name="eits-db" default="init" basedir="." xmlns:artifact="antlib:org.apache.maven.artifact.ant">
  <property name="lib" value="lib" />
  <property name="groovy" value="scripts/groovy" />
  
<!-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
artifact.install
   this target is not necessary if you put ivy.jar in your ant lib directory
   if you already have ivy 1.4 in your ant lib, you can simply remove this
   target and the dependency the 'go' target has on it
-->

  <property name="artifact.jar.dir" value="${basedir}/.ant" />
  <property name="artifact.jar.file" value="${artifact.jar.dir}/maven.artifact-ant-2.0.2-dep.jar" />
  
  <target name="artifact.install" description="Install artifact lib">
      <mkdir dir="${artifact.jar.dir}"/>
      <!--
         download maven antlib from web site so that it can be used even without any
         special installation
       -->
      <available file="${artifact.jar.file}" property="artifact.exists"/>
      <echo message="Installing Maven Antlib..."/>
      <get src="http://www.apache.org/dist/maven/binaries/maven-artifact-ant-2.0.2-dep.jar"
           dest="${artifact.jar.file}" usetimestamp="true" ignoreerrors="${artifact.exists}"/>
      
      <typedef resource="org/apache/maven/artifact/ant/antlib.xml" uri="antlib:org.apache.maven.artifact.ant">
          <classpath>
              <pathelement location="${artifact.jar.file}" />
          </classpath>
      </typedef>
      
  </target>
  
  <target name="init" description="Setup project and fetch dependencies from internet" depends="artifact.install">
      <mkdir dir="${lib}" />
      <artifact:pom id="pom" file="pom.xml" />
      <echo>Fetching dependencies for ${pom.artifactId}-${pom.version}</echo>
      <artifact:dependencies filesetId="dependency.fileset" useScope="runtime">
          <pom refid="pom"/>
      </artifact:dependencies>
      <copy todir="${lib}" verbose="true">
          <fileset refid="dependency.fileset" />
          <mapper type="flatten" />
      </copy>
  </target>
  
  <target name="clean" description="Cleanup the project" >
      <delete dir="${lib}" />
  </target>

</project>

Next, I'll declare my dependencies in a Maven pom.xml file. For this example the 2 dependencies we really care about are groovy-all-1.0.jar and commons-cli-1.0.jar. Both of these are required in order to script from the command line with groovy.

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
    <modelVersion>4.0.0</modelVersion>
    <groupId>eits</groupId>
    <artifactId>eits-db</artifactId>
    <packaging>jar</packaging>
    <version>1.0-SNAPSHOT</version>
    <name>EITS Database Tools</name>
    <url>http://foo.bar.org/projects/${project.name}</url>
    <dependencies>
 <!-- Other dependencies not included for brevity -->
        <dependency>
            <groupId>junit</groupId>
            <artifactId>junit</artifactId>
            <version>4.2</version>
            <scope>test</scope>
        </dependency>
        <dependency>
            <groupId>groovy</groupId>
            <artifactId>groovy-all</artifactId>
            <version>1.0</version>
            <scope>runtime</scope>
        </dependency>
        <dependency>
            <groupId>commons-cli</groupId>
            <artifactId>commons-cli</artifactId>
            <version>1.0</version>
            <scope>runtime</scope>
        </dependency>
    </dependencies>
</project>

Next, I'll need to create a launcher script to set up the environment and execute and of my Groovy scripts. I'm developing on Mac OS X, so I'll create a bash script. The script is pretty simple, it sets the application classpath to include all jars in PRJ_HOME/lib. if your working on windows, setting up the classpath at runtime is a bit more of a pain; the details for it can be found here.

#!/bin/sh
# GSH: Execute groovy script using the projects library directory.

PRJ_HOME=`dirname "$0"`/..

CP=${PRJ_HOME}/scripts/groovy:$PRJ_HOME/conf:$PRJ_HOME/lib
PRJ_LIB=$PRJ_HOME/lib
for jar in $(ls $PRJ_LIB/*.jar)
do
    CP=$CP:$jar
done

java -classpath "$CP" groovy.ui.GroovyMain "$@"

With this in place, all I need to do script any of my java apps is create a Groovy script in PRJ_HOME/scripts/groovy then run 'gsh myscript' on the command line. Groovy does allow for command line switches using a groovy builder called 'CliBuilder'. Discussion about CliBuilder is way beyond what I want to cover here, but there's and example of it's use in the script below. And of course, the groovy script:

/**
 * addtimecode.groovy 
 *
 * This script is used takes a QuickTime movie as input, adds a timecode track 
 * and saves it out to a different 'flattened' QuickTime movie file
 */
 
import org.mbari.movie.Timecode
import org.mbari.qt.FileExistsException
import org.mbari.qt.QT
import org.mbari.qt.TimeUtil
import org.mbari.qt.VideoStandard

/*
 * Setup command line interface
 */
def cli = new CliBuilder(usage: "run_groovy ../scripts/groovy/addtimecode -i [input move] -o [output movie] -t [hh:mm:ss:ff]")
cli.h(longOpt: "help", "usage information")
cli.i(argName: "input", longOpt: "input", args:1, required: true, "movie file to read")
cli.o(argName: "output", longOpt: "output", args:1, required: true, "movie file to write")
cli.t(argName: "timecode", longOpt: "timecode", args:1, required: true, "starting timecode (hh:mm:ss:ff) - assuming NTSC (29.97 fps)")

/*
 * Parse the command line options
 */
def options = cli.parse(args)
if (!options) {
    return
}
if (options.h) {
    println "Script for adding a time-code track to QuickTime movies"
    cli.usage()
}

// Create a file reference
def inputFile = new File(options.i)
URL inputUrl = inputFile.toURL()

// Add a timecode track
def timecode = new Timecode(options.t)

// Export to the output file
def exitCode = 0
try {
    def movie = QT.openMovieFromUrl(inputUrl)
    TimeUtil.addTimeCodeTrack(movie, timecode, VideoStandard.NTSC)
    QT.flattenMovie(movie, new File(options.o), false)
}
catch (FileExistsException e) {
    println "The file ${inputFile} already exists. You must delete it first or save to a different file."
    exitCode = 1
}
catch (Exception e) {
    println "An error occurred during processing. \n  ERROR: ${e.class} \n  MESSAGE:${e.message}"
    exitCode = 2
}

System.exit(exitCode)

To execute this script I would run the following on the command line: gsh addtimecode -i /Some/movie.mov -o /Some/newmovie.mov -t 00:12:34:01.

Boostrapping Maven's Antlib

2007-03-20T08:24:00.000-07:00

Boostrapping Maven's Antlib

Maven repositories can be used from Ant build scripts using Maven's antlib. To do this you must have
both Maven and
Ant installed. When using non-standard Ant extensions it's much easier on your fellow developers if you have your build script bootstrap the use of the libraries so that they don't have to manually install the Ant extensions. Below is an example of how to bootstrap Maven's antlib and fetch all the dependencies, including transitive ones, declared in a Project Object Model (i.e. pom.xml) and copy them to the projects lib directory.
For this example, in addition to your build.xml file, you will need to create a simple pom.xml that defines your dependencies. Here's an example pom.xml:

<?xml version="1.0" encoding="UTF-8"?>
<project 
xmlns="http://maven.apache.org/POM/4.0.0" 
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
    <modelVersion>4.0.0</modelVersion>
    <groupId>eits</groupId>
    <artifactId>eits-db</artifactId>
    <packaging>jar</packaging>
    <version>1.0-SNAPSHOT</version>
    <name>EITS Database Tools</name>
    <url>http://oceana.shore.mbari.org/projects/${project.name}</url>
    <dependencies>
        <dependency>
            <groupId>junit</groupId>
            <artifactId>junit</artifactId>
            <version>4.2</version>
            <scope>test</scope>
        </dependency>
        <dependency>
            <groupId>mysql</groupId>
            <artifactId>mysql-connector-java</artifactId>
            <version>5.0.4</version>
            <scope>runtime</scope>
        </dependency>
        <dependency>
            <groupId>org.slf4j</groupId>
            <artifactId>slf4j-log4j12</artifactId>
            <version>1.3.0</version>
            <scope>runtime</scope>
        </dependency>
    </dependencies>
</project>

To bootstrap antlib into your projects (so users can build without having to install antlib) modify your build.xml file like so:

<?xml version="1.0" ?>
<!-- $Id: build.xml,v 1.21 2007/01/05 01:04:43 brian Exp $ -->
<project 
    name="YourProjectName" 
    default="init" 
    basedir="." 
    xmlns:artifact="antlib:org.apache.maven.artifact.ant">
    
    ... 

    <!-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
         Install the Maven artifact and configure project properties
    -->

    <property name="artifact.jar.dir" value="${basedir}/.ant" />
    <property name="artifact.jar.file" value="${artifact.jar.dir}/maven.artifact-ant-2.0.2-dep.jar" />
    <mkdir dir="${artifact.jar.dir}"/>
    <!--
       download maven antlib from web site so that it can be used even without any
       special installation
     -->
    <available file="${artifact.jar.file}" property="artifact.exists"/>
    <echo message="Installing Maven Antlib..."/>
    <get src="http://www.apache.org/dist/maven/binaries/maven-artifact-ant-2.0.2-dep.jar"
         dest="${artifact.jar.file}" usetimestamp="true" ignoreerrors="${artifact.exists}"/>

    <typedef resource="org/apache/maven/artifact/ant/antlib.xml" uri="antlib:org.apache.maven.artifact.ant">
        <classpath>
            <pathelement location="${artifact.jar.file}" />
        </classpath>
    </typedef>
    
    <!-- 
        Configure properties based on pom.xml and typical Maven build structure.
        Modify this for your own project structure
    -->
    <artifact:pom id="pom" file="pom.xml" />
    <property name="version" value="${pom.version}" />
    <property name="project" value="${pom.artifactId}" />
    <property name="build" value="${pom.build.directory}" />
    <property name="lib" value="${build}/lib" />

    <!-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    init
        Sets up the project for build

    -->
    <target name="init" description="Create the build directory structure">
        <!-- Create the timestamp -->
        <tstamp/>
        
        <!-- Create the directory build structure used by compile -->
        <mkdir dir="${build}"/>
        <mkdir dir="${lib}"/>
        
        <!-- Copy dependencies to target/lib -->
        <echo>Finding dependencies of ${pom.artifactId}-${version}</echo>
        <artifact:dependencies filesetId="dependency.fileset" useScope="runtime">
            <pom refid="pom"/>
        </artifact:dependencies>
        <copy todir="${lib}" verbose="true">
            <fileset refid="dependency.fileset" />
            <mapper type="flatten" />
        </copy>
            
        ...
    </target>

2007-02-01T14:46:00.000-08:00

Creating a Java CLASSPATH dynamically from jars in a directory

When creating DOS batch file to launch a Java application, many people build the classpath by hand. For example:

APP_CLASSPATH=C:\APP\junit.jar;C:\APP\myapp.jar;C:\APP\commons-logging.jar

This is kind of a pain. Especially, when working on an evolving project where the jars change relatively frequently. It would be much easier if the batch file built the classpath dynamically based on the jars found in a directory. Here's how to do this:

First, here's the directory layout for our imaginary application. I'm just explaing this so that you aware of how this example is structured.

APP_HOME
|~~ bin (Batch files live here)
`~~ lib (Jars live here)

Next, create a batch file to launch your Java application or to setup your environment. It's contents will look something like:

@echo off
REM $Id:  $

REM Setup the APP environment
REM ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
REM APP_HOME assumes that your batch file is in APP_HOME\bin.
REM If your batch file is in APP_HOME change the next line
REM to 'SET APP_HOME=%~dp0'
SET APP_HOME=%~dp0..
SET APP_LIB=%APP_HOME%\lib
SET APP_CLASSPATH=
FOR %%i IN ("%APP_LIB%\*.jar") DO CALL cpappend.bat %%i

Notice the call to cpappend.bat. You need to create that batch file; it's contents are:

REM ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
REM Append to APP_CLASSPATH
REM
REM $Id: $
REM ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

SET _TEMP=

REM Process the first argument
IF ""%1"" == """" GOTO end
SET _TEMP=%1
SHIFT

REM Process the remaining arguments.
REM Paths with spaces are handled here
:setArgs
IF ""%1"" == """" GOTO doneSetArgs
SET _TEMP=%_TEMP% %1
SHIFT
GOTO setArgs

REM Build the classpath
:doneSetArgs
SET APP_CLASSPATH=%APP_CLASSPATH%;"%_TEMP%"
GOTO end

:end

If you're working on Mac OS X, Linux, or UNIX setting the path dynamically is much simpler. Here's the equivalent shell scripting code:

APP_HOME=`dirname "$0"`/..
APP_LIB= $APP_HOME/lib
APP_CLASSPATH=
for jar in $(ls $APP_LIB/*.jar)
do
    APP_CLASSPATH=$APP_CLASSPATH:$jar
done

Open an iTerm tab at the current Finder selection

2007-01-04T15:13:00.000-08:00

Open an iTerm tab at the current Finder selection

For you do-it-yourselfers, Here's the Applescript for opening an iTerm tab for the current finder window. To use, open Apple's Script Editor, paste it in and press the compile button. Then 'Save As...' an application; I usually save it into ~/Library/Scripts as 'iTermHere'. Drag the saved application onto your finder toolbar and your ready to roll.

(* 

    Open Terminal Here 
     
    A toolbar script for Mac OS X 10.3/10.4
     
    Written by Marc Liyanage 

     
    See http://www.apple.com/applescript/macosx/toolbar_scripts/ for 
    more information about toolbar scripts. 
     
    See http://www.entropy.ch/software/applescript/ for the latest 
    version of this script. 
     
     
    History: 
    04-Jan-2007: Version 2.2.1 by Brian Schlining. Modified to work with iterm
    11-AUG-2005: Version 2.1.1 by magnamous. minor changes to process_item(this_item)
    21-MAR-2005: Version 2.1 by Will Norris. code cleanup and minor additions 
    18-AUG-2004: Version 2.0 by Allan Marcus. uses posix path    
    30-OCT-2001: Version 1.0, adapted from one of the example toolbar scripts 
    30-OCT-2001: Now handles embedded single quote characters in file names 
    30-OCT-2001: Now handles folders on volumes other than the startup volume 
    30-OCT-2001: Now handles click on icon in top-level (machine) window 
    31-OCT-2001: Now displays a nicer terminal window title, courtesy of Alain Content 
    11-NOV-2001: Now folders within application packages (.app directories) and has a new icon 
    12-NOV-2001: New properties to set terminal columns and rows as the Terminal does not use default settings 
    14-NOV-2001: Major change, now handles 8-bit characters in all shells, and quotes and spaces in tcsh 
    18-NOV-2001: Version 1.1: Rewrite, now uses a temporary file  ~/.OpenTerminalHere to communicate 
    the directory name between AppleScript and the shell because this is much more reliable for 8-bit characters 
     
 *)


property debug : false

-- when the toolbar script icon is clicked 
-- 
on run
 tell application "Finder"
  try
   set this_folder to (the target of the front window) as alias
  on error
   set this_folder to startup disk
  end try
  
  my process_item(this_folder)
  
 end tell
end run


-- This handler processes folders dropped onto the toolbar script icon 
-- 
on open these_items
 repeat with i from 1 to the count of these_items
  set this_item to item i of these_items
  my process_item(this_item)
 end repeat
end open


-- this subroutine processes does the actual work 
-- this version can handle this weirdo case: a folder named "te'st"ö te%s`t"

on process_item(this_item)
 
 set thePath to quoted form of POSIX path of this_item
 
 tell application "iTerm"
  activate
  -- just open a terminal and cd to thePath
  make new terminal
  tell the first terminal
   activate current session
   launch session "Default Session"
   tell the last session
    write text "cd '" & thePath & "'; clear"
   end tell
  end tell
 end tell
 
 
end process_item

This script is based on one that's been circulated around for a while. I found the original script here

2006-12-27T13:58:00.000-08:00

How to use Google to search for a movie

Here's an example that finds Battlestar Galactica AVI's

-inurl:htm -inurl:html intitle:”index of” “Last modified” avi “battlestar”

This valid Google search returns music files of certain types with the word Nirvana in the link from certain pages all over the web:

-inurl:(htm|html|php) intitle:"index of" +"last modified" +"parent directory" 
+description +size +(wma|mp3) "Nirvana"

2006-11-27T08:45:00.000-08:00

Blame Canada

2006-11-20T08:43:00.000-08:00

Los Gringos Salvajes

This is my brother-in-laws band. They're hard to describe in words; but they sound great. You can take a listen at http://www.myspace.com/gringossalvajes

2006-11-20T08:41:00.000-08:00

The McDonalds Order Song...

2006-11-16T23:44:00.000-08:00

Now this is the right way to sing Hey-ya!

macosxhints.com - Use Quicksilver for quick timed reminders

2006-10-05T09:03:00.000-07:00

This is a great hint for using Quicksilver (for Mac OS X) to do timed reminders or actions.

QuickTime for Java: Threading issue

2006-08-07T09:21:00.000-07:00

QTJ's capture manager, SequenceGrabber, periodically needs some CPU time or it will crash on Intel Macs. A solution was posted on java.net.

Mark Occurrences for IntelliJ IDEA

2006-06-27T11:37:00.000-07:00

I've been using IntelliJ IDEA for some development lately. It's a great IDE but one thing that I've missed is the 'Mark Occurences' feature available in Eclipse and as a plugin in Netbeans. Fortunatly, it is available in IntelliJ, it's just a bit hidden. To activate it, use [CTRL]+[SHIFT]+[F7] (for Mac folks that's [META]+[SHIFT]+[F7]). Pressing [ESC] disables it.

Using Runtime.getRuntime().exec() in java

2006-03-13T10:36:00.000-08:00

Using Runtime.getRuntime().exec() in java

This is not well documented by Sun but if you make a call using Runtime.getRuntime().exec() you should always get a reference to the Process object that's returned. When you're done with the Process you should either:

Explicitly get and close the input, output and error streams for the Process, whether you've used them or not
Call Process.destroy()

If you don't make these calls you will leak resources. Once you start leaking resources you may start to see "IOException: Broken Pipe" errors

2006-02-22T10:26:00.000-08:00

Enabling readline support with Jython 2.2a on Mac OS X

If you do much work using Java, you're eventually going to try a scripting language that runs on the JVM. My personal favorite is Jython. The only problem is that Jython doesn't have readline support right out of the box. This makes command line usage very frustrating since all the shortcuts that command line users love are missing. Fortunatly, it's relatively easy to enable readline support. Here's the steps for setting jython up with it on Mac OS X:

Install Fink
Use fink to install 'readline-java'
Edit the Jython registry file. Normally, this is located in the same directory that the jython.jar file is in. In case you're wondering, it's named 'registry'. You'll want to add the following lines:
```
python.console=org.python.util.ReadlineConsole
python.console.readlinelib=GnuReadline
```
The last step is to add 'libJavaReadline.jnilib' to the java.library.path. An easy way to do this is to edit your script that launches Jython and add the line:
```
DYLD_LIBRARY_PATH=$DYLD_LIBRARY_PATH:/sw/lib
```
or you can just copy libJavaReadline.jnilib and java-readline.jar to /Library/Java/Extensions

SVN Client for the Mac

2006-02-17T14:03:00.000-08:00

Here's the details on how to install SCPlugin which adds SVN context to Mac OS X's Finder application

LearningCommons/Documentation/Subversion - Wiki.ucalgary.ca

2006-02-15T10:17:00.000-08:00

Handy tips for using Subversion...

Cross Fade images using javaScript

http://slayeroffice.com/code/imageCrossFade/

2006-01-21T19:53:00.000-08:00

Marina hosts first debate on general plan: "Saying it wants to demystify the Monterey County general plan initiative, a Marina-based organization has scheduled a debate so leaders on both sides of the issue can hash out the ballot measure's finer points."

2006-01-19T14:41:00.000-08:00

Invert the Screen on Mac OS X

From http://www.patriceschneider.com/apple-osx/blog/?p=8

Another pretty cool one is to hit the keyboard combination “Control ALT Apple 8″ which will invert the colours on your screen. (Hitting the same combination will change it back to the standard setting.) This can at times also give you an extra 30 minutes or more. It can under certain bright lights also make it easier to read text, btw. Give it a try!

Make a Remote Mac Sing

2006-01-12T08:48:00.000-08:00

Make a Remote Mac Sing

I found this at http://lowendmac.com/crews/06/0110.html
This one requires you to have a administrative login account on a remote Mac that has Remote Login (SSH) enabled. Assuming your friend or co-worker's Mac meets these requirements, you can play a heck of a April fool's joke.
In the Terminal, first remotely login to the system by using ssh <username>@<remote Mac's IP address or hostname>.
Once logged in, enter this command:

sudo osascript -e "set Volume 20"

And then press return. This sends an AppleScript to set the volume to an audible level. Then enter the following command:
sudo osascript -e 'say "Dum dum dum dum dum dum dum he he he ho ho ho fa lah lah lah lah lah lah fa lah full hoo hoo hoo" using "Cellos"'
This sends an AppleScript to say the phrase using the Cellos voice. The results are quite humorous. You can try it out on your own system to hear what it sounds like.

2005-12-15T14:59:00.000-08:00

Handy commands for interacting between Finder and Terminal on Mac OS X

http://pdb.finkproject.org/pdb/package.php/pos

2005-11-23T08:35:00.000-08:00

Runing Java SE 5 on Panther

http://www.macosxhints.com/article.php?story=20051117173104762&lsrc=osxh

2005-09-01T08:35:00.000-07:00

Setting system properties in Java Web Start

This is a .jnlp snippet for setting a system property;here we set the timezone:

  <resources>
    <j2se version="1.5+"/>
    <jar href="dist/hooves-prey.jar"/>
    <property name="user.timezone" value="UTC"/>
  </resources>

The property *does* get set, but j2see 1.4.2 ignores it. It works fine with Java 5 though.

2005-08-18T16:47:00.000-07:00

Efficient String Concatenation in Python

If you've ever created long strings in Python, you may have noticed that string concatenation is slow. There's a great site that tries out the different methods for concatenating strings in python at http://www.skymind.com/~ocrow/python_string/. In a nutshell, I've found this one works best for my operations:

def method5():
  # Example that concatenates numbers
  from cStringIO import StringIO
  file_str = StringIO()
  for num in xrange(loop_count):
    file_str.write(`num`)

  return file_str.getvalue()

2005-07-26T10:16:00.000-07:00

See complete output of 'ps' command on Mac OS X

This is a tip from macosxhints.com The output of 'ps' is only as wide as there are columns in the current terminal. Needless to say this gets very annoying since one rarely can read the entire output. According to the man file for ps, using the w option forces the output to use 132 columns, which helps little. But if you specify it twice, it forces ps to use as many columns as necessary. Thus long lines are displayed wrapping to the the next line. So if you are looking for lines that you think may be longer than the Terminal's width, use:

$ ps -auxww | grep stringToSearchFor

2005-05-25T09:26:00.000-07:00

Music of the week

Anna Ternheim - Better Be
Venus Hum - Soul Sloshing
Vinylshakerz - One Night in Bangkok
Jem - Just a Ride
Filter - Take a Picture
Jimmy Buffett - Mademoiselle (Voulez-vous Danser)

2005-05-12T12:49:00.000-07:00

Generating epoch seconds in python

I often need to convert a specific time to epoch seconds in various languages. Here's a python way:

# A date string to epoch seconds
epoch = time.mktime(time.strptime('1982-01-01 00:00:00', '%Y-%m-%d %H:%M:%S'))

# Epoch seconds to a date string
time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(epoch))

2005-05-09T10:11:00.000-07:00

A simple Linux bash script for reading from a serial port

#/bin/bash
while true
do
  read LINE < /dev/tty0S
  echo $LINE
done

2005-05-06T10:53:00.000-07:00

Here's what I'm listening to today

Artist	Song
Relient K	Be My Escape
Olu Dara	Zora
Massive Attack	Teardrop
Eminem	Lose Yourself
Ani DiFranco	The Arrivals Gate
Outkast	Hey Ya!
Beth Orton	Stolen Car
Darude vs. Robert Miles	Children of the Sandstorm
Eddi Reader	Simple Soul
Israel Kamakawiwo'ole	Somewhere Over The Rainbow

2005-03-11T08:16:00.000-08:00

Integrating JProfiler 3.3 into Eclipse on Mac OS X

Start Jprofiler
From the file menu select 'IDE Integration'. Follow the steps in the wizard. This will install the plugin. However, it won't run because eclipse tries to run it using a 2.x compatiblility layer.
Edit the file $ECLIPSE_HOME/plugins/com.jprofiler.integrations.eclipse/plugin.xml. Change the line <?eclipse version="3.0"?> to <?eclipse version="3.0.1"?> . This will take care of the compatibility layer
Go to Windows->Customize Perspective->Commands and activate "Profile". The "Run" menu will then have "Profile" entries and the toolbar will have an additional button.

2005-02-11T09:19:00.000-08:00

Enable readline support for Python on Mac OS X

To enable readline for python on Mac OS X:
sudo python `python -c "import pimp; print pimp.__file__"` -i readline

This hint was found here.

A couple of other useful python links are:

IPython - An enhancement to the python interactive shell
Python Tutorial - Written by the man himself
Installing TKinter on MacOS X

2005-02-02T15:19:00.000-08:00

New Matlab featues

There are 2 handy new features in Matlab 7.0. The first is mlint, an m-file similar to C's lint that checks the code for style, language usage and portability. Mathworks has an article about it at http://www.mathworks.com/company/newsletters/news_notes/dec04/cleanup.html. The second handy feature is the new textscan function which really simplifies access to data in arbirarily formatted text files. Here's some info from an article on http://www.mathworks.com/company/newsletters/digest/nov04/newfeatures.html?mld_na_tech2.

Text File Reading

The new textscan function enables you to access very large text files that have arbitrary format. This function is similar to textread but adds the ability to specify a file identifier so that a file pointer can be tracked and traversed through the file. The file can therefore be read a block at a time, changing the format on each occasion. For example, suppose we have a text file, test12_80211b.txt, which contains multiple different-sized blocks of data, each with the following format:

Two headerlines of description
A parameter m
A p x m table of data

Here is how test12_80211b.txt looks:

*       Mobile1
*       SNR Vs test No
Num tests=19
,-5.00E+00,-4.00E+00,-3.00E+00,-2.00E+00,...
1.00E+00,6.19E-07,8.63E-07,6.43E-07,1.84E-07,...
2.00E+00,2.88E-07,4.71E-07,6.92E-07,1.43E-07,...
3.00E+00,2.52E-07,8.11E-07,4.74E-07,8.48E-07,...
4.00E+00,...
...

*       Mobile2
*       SNR Vs test No
Num tests=20
,-5.00E+00,-4.00E+00,-3.00E+00,-2.00E+00,-1.00E+00,0.00E+00,...
1.00E+00,6.19E-07,8.63E-07,6.43E-07,1.84E-07,6.86E-07,3.73E-,...
2.00E+00,...

You could use the following MATLAB commands to read it in:

fid = fopen('test12_80211b.txt', 'r'); % Open text file
InputText = textscan(fid, '%s', 2, 'delimiter', '\n'); % Read header lines
HeaderLines = InputText{1} HeaderLines =
'* Mobile1'
'* SNR Vs test No'

InputText = textscan(fid, 'Num tests=%f'); % Read parameter value
NumCols=InputText{1} NumCols =
19

InputText=textscan(fid, '%f', 'delimiter', ','); % Read data block
Data=reshape(InputText{1},[],NumCols)';
format short g
Section=Data(1:5,1:5) Section =
NaN 	-5 	-4 	-3 	-2
1 	6.19e-007 	8.63e-007 	6.43e-007 	1.84e-007
2 	2.88e-007 	4.71e-007 	6.92e-007 	1.43e-007
3 	2.52e-007 	8.11e-007 	4.74e-007 	8.48e-007
4 	1.97e-007 	1.64e-007 	1.38e-007 	6.17e-007

For improved data access speed, in this release of MATLAB the reading of comma-separated-value (CSV) files is an order of magnitude faster.

2005-02-02T15:15:00.000-08:00

Burning ISO images to CD on Mac OS X

To burn ISOs use fink to install cdrecord. Then use the command (here it's burning Ubuntu linux to CD):

sudo cdrecord -v speed=24 dev=IODVDServices warty-release-install-i386.iso

2004-04-16T14:38:00.000-07:00

Creating an image from SGT in java

JPlotLayout jpl; 
// initialize and add plots to jpl ... ommitted here.

// Get graphic
int width = jpl.getWidth();
int height = jpl.getHeight();
BufferedImage img = new BufferedImage(width, height, BufferedImage.TYPE_INT_RGB);
Graphics g = img.getGraphics();
jpl.paint(g);

// Save to a file
try {
    BufferedOutputStream out = new BufferedOutputStream(new FileOutputStream("some.jpg"));
    JPEGImageEncoder encoder = JPEGCodec.createJPEGEncoder(out);
    JPEGEncoderParam param = encoder.getDefaultJPEGEncoderParam(img);
    int quality = 100;
    quality = Math.max(0, Math.min(quality, 100));
    param.setQuality((float) quality/100F, false);
    encoder.setJPEGEncoderParam(param);
    encoder.encode(img);
    out.close();
}
catch (IOException e) {
    // Log or deai with exception
}