embulk 0.4.2 → 0.4.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: d63d08592793b691be7f691d842f73d18a202d45
-  data.tar.gz: 5a14e328213b2a97fe6e0abaa044428bb5caa0fd
+  metadata.gz: b24d1b7fd55459a318f9dbce2fff9b976f40738b
+  data.tar.gz: 88573c322b4d3e32f4fb11775122db1c1b28f3d2
 SHA512:
-  metadata.gz: 8e447c2ae251e6ef3309c0862344d10074797389ea2c55bce2e2302d82eea2425345316efdf7dae9adcd9f08de5b0132c04e6cad43edaef29ac3cf635ad640f8
-  data.tar.gz: 0b33b0a4f2dc7911d1432828e303c464c8b25bda00dc9eb3df8c9fb905c85011c30132292c581163fb557d2e2eaeb3a32a8bbd61bd486509361b4464aa2477ed
+  metadata.gz: 3179a8555d542ae38a785ad308a9b44560ae66889f263c932675e4ef5050b719636126ec0a960c6dba098030e269c3b0dd7b34f841d278857464c143ea0a2940
+  data.tar.gz: 20772970a71e622473f119ebc4ee9316fdbfff48e96948345fb6a001342d727a70ecd525e217fac98eb0838f2e20ccebcfd6d48603c945a0d9b096f7070ea92a

data/.gitignore

@@ -1,5 +1,6 @@
 target/
 build/
+pkg/
 *.iml
 *~
 ._*

data/README.md

@@ -24,7 +24,7 @@ You can release plugins to share your efforts of data cleaning, error handling,
 The single-file package is the simplest way to try Embulk. You can download the latest embulk-VERSION.jar from [the releases page](https://bintray.com/embulk/maven/embulk/view#files) and run it with java:
 
 ```
-wget https://bintray.com/artifact/download/embulk/maven/embulk-0.4.2.jar -O embulk.jar
+wget https://bintray.com/artifact/download/embulk/maven/embulk-0.4.3.jar -O embulk.jar
 java -jar embulk.jar --help
 ```
 

data/bin/embulk

@@ -13,7 +13,7 @@ while i = ARGV.find_index {|arg| arg =~ /^\-D/ }
     end
     ARGV.slice!(i, 2)
   else
-    java_args << ARGV[i]
+    java_args << ARGV[i][2..-1]
     ARGV.slice!(i)
   end
 end

data/build.gradle

@@ -4,19 +4,22 @@ plugins {
     id 'com.github.ben-manes.versions' version '0.7'
     id 'com.github.jruby-gradle.base' version '0.1.5'
     id 'com.github.johnrengelman.shadow' version '1.2.0'
-    id 'java'
 }
 
+def java_projects = [project(":embulk-core"), project(":embulk-standards"), project(":embulk-cli")]
+def release_projects = [project(":embulk-core"), project(":embulk-standards")]
+
 allprojects {
     group = 'org.embulk'
-    version = '0.4.2'
+    version = '0.4.3'
 
-    apply plugin: 'maven'  // install jar files to the local repo: $ gradle install
+    apply plugin: 'java'
     apply plugin: 'maven-publish'
     apply plugin: 'com.jfrog.bintray'
-    apply plugin: 'java'
 
-    // upload artifacts to Bintray: $ gradle bintrayUpload
+    //
+    // bintrayUpload task
+    //
     bintray {
         // write at your bintray user name and api key to ~/.gradle/gradle.properties file:
         // bintray_user=frsyuki
@@ -24,7 +27,13 @@ allprojects {
         user = project.hasProperty('bintray_user') ? bintray_user : ''
         key = project.hasProperty('bintray_api_key') ? bintray_api_key : ''
 
-        publications = ['mavenJava']
+        publications = ['bintrayMavenRelease']
+
+        filesSpec {
+            // include embulk-<version>.jar built by 'cli' task
+            from "pkg/embulk-${project.version}.jar"
+            into "embulk-${project.version}.jar"
+        }
 
         dryRun = false
         publish = false  // TODO automate uploading embulk.jar and make this true
@@ -56,54 +65,24 @@ allprojects {
 }
 
 subprojects {
-    apply plugin: 'findbugs'
-    apply plugin: 'jacoco'
-
-    repositories {
-        mavenCentral()
-        jcenter()
-    }
-
-    compileJava.options.encoding = 'UTF-8' // source encoding
-    sourceCompatibility = 1.7
-    targetCompatibility = 1.7
+    if (java_projects.contains(project)) {
+        apply plugin: 'maven'
+        apply plugin: 'findbugs'
+        apply plugin: 'jacoco'
+
+        repositories {
+            mavenCentral()
+            jcenter()
+        }
 
-    configurations {
-        // guice depends on asm and cglib but version of the libraries conflict
-        // with ones bundled in jruby-complete and cause bytecode compatibility error
-        compile.exclude group: 'asm', module: 'asm'
-        compile.exclude group: 'org.sonatype.sisu.inject', module: 'cglib'
-    }
+        compileJava.options.encoding = 'UTF-8' // source encoding
+        sourceCompatibility = 1.7
+        targetCompatibility = 1.7
 
-    // determine which dependencies have updates: $ gradle dependencyUpdates
-    dependencies {
-        compile 'com.google.guava:guava:18.0'
-        compile 'com.google.inject:guice:3.0'
-        compile 'com.google.inject.extensions:guice-multibindings:3.0'
-        compile 'javax.inject:javax.inject:1'
-        compile 'com.fasterxml.jackson.core:jackson-annotations:2.5.0'
-        compile 'com.fasterxml.jackson.core:jackson-core:2.5.0'
-        compile 'com.fasterxml.jackson.core:jackson-databind:2.5.0'
-        compile 'com.fasterxml.jackson.datatype:jackson-datatype-guava:2.5.0'
-        compile 'com.fasterxml.jackson.datatype:jackson-datatype-joda:2.5.0'
-        compile 'com.fasterxml.jackson.module:jackson-module-guice:2.5.0'
-        compile 'log4j:log4j:1.2.17'
-        compile 'org.slf4j:slf4j-api:1.7.10'
-        compile 'org.slf4j:slf4j-log4j12:1.7.10'
-        compile 'org.jruby:jruby-complete:1.7.19'
-        compile 'com.google.code.findbugs:annotations:3.0.0'
-        compile 'org.yaml:snakeyaml:1.14'
-        compile 'javax.validation:validation-api:1.1.0.Final'
-        compile 'org.apache.bval:bval-jsr303:0.5'
-        compile 'io.airlift:slice:0.9'
-        compile 'joda-time:joda-time:2.7'
-        compile 'io.netty:netty-buffer:5.0.0.Alpha1'
-        compile 'com.ibm.icu:icu4j:54.1.1'
-
-        testCompile 'junit:junit:4.12'
-    }
+        dependencies {
+            testCompile 'junit:junit:4.12'
+        }
 
-    gradle.projectsEvaluated {
         tasks.withType(JavaCompile) {
             options.compilerArgs << "-Xlint:unchecked" //<< "-Xlint:deprecation"
         }
@@ -113,39 +92,36 @@ subprojects {
                 html.enabled = true
             }
         }
-    }
 
-    findbugs {
-        ignoreFailures = true
-    }
+        findbugs {
+            ignoreFailures = true
+        }
 
-    javadoc {
-        options {
-            locale = 'en_US'
-            encoding = 'UTF-8'
+        javadoc {
+            options {
+                locale = 'en_US'
+                encoding = 'UTF-8'
+            }
         }
-    }
 
-    // add javadoc/source jar tasks as artifacts to be released
-    task sourcesJar(type: Jar, dependsOn: classes) {
-        classifier = 'sources'
-        from sourceSets.main.allSource
-    }
-    task javadocJar(type: Jar, dependsOn: javadoc) {
-        classifier = 'javadoc'
-        from javadoc.destinationDir
-    }
-    artifacts {
-        archives sourcesJar, javadocJar
+        // add javadoc/source jar tasks as artifacts to be released
+        task sourcesJar(type: Jar, dependsOn: classes) {
+            classifier = 'sources'
+            from sourceSets.main.allSource
+        }
+        task javadocJar(type: Jar, dependsOn: javadoc) {
+            classifier = 'javadoc'
+            from javadoc.destinationDir
+        }
+        artifacts {
+            archives sourcesJar, javadocJar
+        }
     }
 
     publishing {
         publications {
-            if (!project.name.equals("embulk-docs")) {
-                mavenJava(MavenPublication) {
-                    groupId project.group
-                    artifactId project.name
-                    version project.version
+            if (release_projects.contains(project)) {
+                bintrayMavenRelease(MavenPublication) {
                     from components.java
                     artifact sourcesJar
                     artifact javadocJar
@@ -155,6 +131,28 @@ subprojects {
     }
 }
 
+//
+// classpath task
+//
+task classpath(dependsOn: ['build', ':embulk-cli:classpath']) << { }
+clean { delete 'classpath' }
+
+//
+// cli task
+//
+task cli(dependsOn: ':embulk-cli:shadowJar') << {
+    file('pkg').mkdirs()
+    File f = file("pkg/embulk-${project.version}.jar")
+    f.write('''\
+#!/bin/sh
+exec java -jar "$0" "$@"
+exit 127
+''')
+    f.append(file("embulk-cli/build/libs/embulk-cli-${project.version}-all.jar").readBytes())
+    f.setExecutable(true)
+}
+bintrayUpload.dependsOn(['cli'])
+
 project(':embulk-cli') {
     apply plugin: 'com.github.johnrengelman.shadow'
 
@@ -167,31 +165,19 @@ project(':embulk-cli') {
                        'Specification-Version': project.version,
                        'Main-Class': 'org.embulk.cli.Main'
         }
-        append("${parent.projectDir}/COPYING")
+        append("${rootProject.projectDir}/COPYING")
     }
 
     task classpath(type: Copy) {
-        doFirst { file("${parent.projectDir}/classpath").mkdirs() }
+        doFirst { file("${rootProject.projectDir}/classpath").mkdirs() }
         from configurations.runtime
-        into "${parent.projectDir}/classpath"
+        into "${rootProject.projectDir}/classpath"
     }
 }
 
-task classpath(dependsOn: ['build', ':embulk-cli:classpath']) << { }
-clean { delete 'classpath' }
-
-task cli(dependsOn: ':embulk-cli:shadowJar') << {
-    file('pkg').mkdirs()
-    File f = file("pkg/embulk-${project.version}.jar")
-    f.write('''\
-#!/bin/sh
-exec java -jar "$0" "$@"
-exit 127
-''')
-    f.append(file("embulk-cli/build/libs/embulk-cli-${project.version}-all.jar").readBytes())
-    f.setExecutable(true)
-}
-
+//
+// gem task
+//
 import com.github.jrubygradle.JRubyExec
 task gem(type: JRubyExec) {
     jrubyArgs '-rrubygems/gem_runner', '-eGem::GemRunner.new.run(ARGV)', 'build'
@@ -201,6 +187,18 @@ task gem(type: JRubyExec) {
 gem.dependsOn('gemspec')
 gem.dependsOn('classpath')
 
+//
+// rubyGemsUpload task
+//
+task rubyGemsUpload(type: JRubyExec, dependsOn: ["gem"]) {
+    jrubyArgs '-rrubygems/gem_runner', '-eGem::GemRunner.new.run(ARGV)', 'push'
+    script "pkg/embulk-${project.version}.gem"
+}
+gem.dependsOn('gemspec')
+
+//
+// releaseCheck and release tasks
+//
 task releaseCheck << {
     if (!file("lib/embulk/version.rb").getText().contains("${project.version}")) {
         throw new GradleException("lib/embulk/version.rb doesn't include ${project.version}")
@@ -215,21 +213,21 @@ task releaseCheck << {
     if (!file("embulk-docs/src/release/release-${project.version}.rst").getText().contains(date)) {
         throw new GradleException("embulk-docs/src/release/release-${project.version}.rst doesn't include today's release date")
     }
-    println "Ready. Run ./gradlew release"
+    // TODO check git-ls-files includes release-<version>.rst file
+    println "Ready. Run 'release' task."
 }
 
-task release(dependsOn: ["cli", "gem"]) << {
+task release(dependsOn: ["releaseCheck", "bintrayUpload", "rubyGemsUpload"]) << {
     println """
-manual operations:
+Manual operations:
 
-  git commit -a -m v${project.version}
+  git commit -am v${project.version}
   git tag v${project.version}
-  gem push pkg/embulk-${project.version}.gem"
-  ./gradlew bintrayUpload
-  open "https://bintray.com/embulk/maven/embulk/${project.version}/view"  # and upload pkg/embulk-${project.version}.jar
 
 """
 }
+bintrayUpload.mustRunAfter('releaseCheck')
+rubyGemsUpload.mustRunAfter('releaseCheck')
 
 task gemspec << {
     file('build').mkdirs()

data/embulk-core/build.gradle

@@ -1,6 +1,41 @@
+// include ruby scripts
 sourceSets {
     main.resources {
-        srcDirs 'src/main/resources'
-        srcDirs "${parent.projectDir}/lib"
+        srcDirs "${rootProject.projectDir}/lib"
     }
 }
+
+configurations {
+    // com.google.inject:guice depends on asm and cglib but version of the libraries conflict
+    // with ones bundled in jruby-complete and cause bytecode compatibility error
+    compile.exclude group: 'asm', module: 'asm'
+    compile.exclude group: 'org.sonatype.sisu.inject', module: 'cglib'
+}
+
+// determine which dependencies have updates: $ gradle dependencyUpdates
+dependencies {
+    compile 'com.google.guava:guava:18.0'
+    compile 'com.google.inject:guice:3.0'
+    compile 'com.google.inject.extensions:guice-multibindings:3.0'
+    compile 'javax.inject:javax.inject:1'
+    compile 'com.fasterxml.jackson.core:jackson-annotations:2.5.0'
+    compile 'com.fasterxml.jackson.core:jackson-core:2.5.0'
+    compile 'com.fasterxml.jackson.core:jackson-databind:2.5.0'
+    compile 'com.fasterxml.jackson.datatype:jackson-datatype-guava:2.5.0'
+    compile 'com.fasterxml.jackson.datatype:jackson-datatype-joda:2.5.0'
+    compile 'com.fasterxml.jackson.module:jackson-module-guice:2.5.0'
+    compile 'log4j:log4j:1.2.17'
+    compile 'org.slf4j:slf4j-api:1.7.10'
+    compile 'org.slf4j:slf4j-log4j12:1.7.10'
+    compile 'org.jruby:jruby-complete:1.7.19'
+    compile 'com.google.code.findbugs:annotations:3.0.0'
+    compile 'org.yaml:snakeyaml:1.14'
+    compile 'javax.validation:validation-api:1.1.0.Final'
+    compile 'org.apache.bval:bval-jsr303:0.5'
+    compile 'io.airlift:slice:0.9'
+    compile 'joda-time:joda-time:2.7'
+    compile 'io.netty:netty-buffer:5.0.0.Alpha1'
+
+    // for embulk/guess/charset.rb
+    compile 'com.ibm.icu:icu4j:54.1.1'
+}

data/embulk-core/src/main/java/org/embulk/command/Runner.java

@@ -169,7 +169,11 @@ public class Runner
 
         String yml = writeNextConfig(options.getNextConfigOutputPath(), config, configDiff);
         System.err.println(yml);
-        System.out.println("Created "+options.getNextConfigOutputPath());
+        if (options.getNextConfigOutputPath() == null) {
+            System.out.println("Use -o PATH option to write the guessed config file to a file.");
+        } else {
+            System.out.println("Created '"+options.getNextConfigOutputPath()+"' file.");
+        }
     }
 
     private void checkFileWritable(String path)

data/embulk-docs/src/release.rst

@@ -13,4 +13,5 @@ Release Notes
     release/release-0.4.0
     release/release-0.4.1
     release/release-0.4.2
+    release/release-0.4.3
 

data/embulk-docs/src/release/release-0.4.3.rst

@@ -0,0 +1,34 @@
+Release 0.4.3
+==================================
+
+CLI
+------------------
+
+* All subcommands show current time with timezone and embulk's version number at the beginning.
+
+Plugin API Changes
+------------------
+
+* ``Thread.currentThread().getContextClassLoader()`` no longer returns JRuby's classloader. It returns null so that dependent libraries fallback to appropriate ``this.getClass().getContextClassLoader()`` call.
+
+Built-in plugins
+------------------
+
+* ``guess/csv`` guesses ``escape`` and ``null_string`` options.
+* Fixed ``guess/csv`` fails if the csv file includes a timestamp value with timezone (@kinyuka++).
+* Fixed memory leak at ``output/file`` (@akirakw++).
+* Fixed ``input/file`` loads unnecessary files when it lists files from ``.``.
+
+General Changes
+------------------
+
+* embulk-cli artifact is no longer released.
+* embulk-standards artifact doesn't directly depend on dependencies of embulk-core.
+* Updated the build script.
+
+  * ``bintrayUpload`` task uploads embulk-<version>.jar.
+  * ``release`` task actually releases gem to RubyGems and publishes jar files to Bintray.
+
+Release Date
+------------------
+2015-02-17

data/embulk-standards/src/main/java/org/embulk/standards/LocalFileInputPlugin.java

@@ -51,6 +51,8 @@ public class LocalFileInputPlugin
 
     private final Logger log = Exec.getLogger(getClass());
 
+    private final static Path CURRENT_DIR = Paths.get(".").normalize();
+
     @Override
     public ConfigDiff transaction(ConfigSource config, FileInputPlugin.Control control)
     {
@@ -92,27 +94,58 @@ public class LocalFileInputPlugin
         } else {
             fileNamePrefix = pathPrefix.getFileName().toString();
             Path d = pathPrefix.getParent();
-            directory = (d == null ? Paths.get(".") : d);
+            directory = (d == null ? CURRENT_DIR : d);
         }
 
         final ImmutableList.Builder<String> builder = ImmutableList.builder();
         final String lastPath = task.getLastPath().orNull();
         try {
-            log.info("Listing local files at directory '{}' filtering filename by prefix '{}'", directory, fileNamePrefix);
+            log.info("Listing local files at directory '{}' filtering filename by prefix '{}'", directory.equals(CURRENT_DIR) ? "." : directory.toString(), fileNamePrefix);
             Files.walkFileTree(directory, new SimpleFileVisitor<Path>() {
                 @Override
-                public FileVisitResult visitFile(Path path, BasicFileAttributes aAttrs)
+                public FileVisitResult preVisitDirectory(Path path, BasicFileAttributes attrs)
+                {
+                    if (path.equals(directory)) {
+                        return FileVisitResult.CONTINUE;
+                    } else if (lastPath != null && path.toString().compareTo(lastPath) <= 0) {
+                        return FileVisitResult.SKIP_SUBTREE;
+                    } else {
+                        Path parent = path.getParent();
+                        if (parent == null) {
+                            parent = CURRENT_DIR;
+                        }
+                        if (parent.equals(directory)) {
+                            if (path.getFileName().toString().startsWith(fileNamePrefix)) {
+                                return FileVisitResult.CONTINUE;
+                            } else {
+                                return FileVisitResult.SKIP_SUBTREE;
+                            }
+                        } else {
+                            return FileVisitResult.CONTINUE;
+                        }
+                    }
+                }
+
+                @Override
+                public FileVisitResult visitFile(Path path, BasicFileAttributes attrs)
                 {
-                    if (lastPath == null || path.toString().compareTo(lastPath) > 0) {
-                        if (path.getParent().equals(directory)) {
+                    if (lastPath != null && path.toString().compareTo(lastPath) <= 0) {
+                        return FileVisitResult.CONTINUE;
+                    } else {
+                        Path parent = path.getParent();
+                        if (parent == null) {
+                            parent = CURRENT_DIR;
+                        }
+                        if (parent.equals(directory)) {
                             if (path.getFileName().toString().startsWith(fileNamePrefix)) {
                                 builder.add(path.toString());
+                                return FileVisitResult.CONTINUE;
                             }
                         } else {
                             builder.add(path.toString());
                         }
+                        return FileVisitResult.CONTINUE;
                     }
-                    return FileVisitResult.CONTINUE;
                 }
             });
         } catch (IOException ex) {

data/embulk-standards/src/main/java/org/embulk/standards/LocalFileOutputPlugin.java

@@ -112,6 +112,8 @@ public class LocalFileOutputPlugin
                     output.write(buffer.array(), buffer.offset(), buffer.limit());
                 } catch (IOException ex) {
                     throw new RuntimeException(ex);
+                } finally {
+                    buffer.release();
                 }
             }
 

data/lib/embulk/command/embulk.rb

@@ -1,3 +1,7 @@
+# reset context class loader set by org.jruby.Main.main to nil. embulk manages
+# multiple classloaders. default classloader should be Plugin.class.getClassloader().
+java.lang.Thread.current_thread.set_context_class_loader(nil)
+
 bundle_path = ENV['EMBULK_BUNDLE_PATH'].to_s
 bundle_path = nil if bundle_path.empty?
 

data/lib/embulk/command/embulk_example.rb

@@ -13,8 +13,8 @@ module Embulk
 id,account,time,purchase,comment
 1,32864,2015-01-27 19:23:49,20150127,embulk
 2,14824,2015-01-27 19:01:23,20150127,embulk jruby
-3,27559,2015-01-28 02:20:02,20150128,embulk core
-4,11270,2015-01-29 11:54:36,20150129,"Embulk ""csv"" parser plugin"
+3,27559,2015-01-28 02:20:02,20150128,"Embulk ""csv"" parser plugin"
+4,11270,2015-01-29 11:54:36,20150129,NULL
 EOF
     end
 

data/lib/embulk/command/embulk_run.rb

@@ -17,8 +17,9 @@ module Embulk
     # to make sure org.embulk.jruby.JRubyScriptingModule can require 'embulk/java/bootstrap'
     $LOAD_PATH << Embulk.home('lib')
 
+    require 'embulk/version'
+
     if argv.include?('--version')
-      require 'embulk/version'
       puts "embulk #{Embulk::VERSION}"
       exit 0
     end
@@ -31,6 +32,8 @@ module Embulk
     require 'optparse'
     op = OptionParser.new
 
+    puts "#{Time.now.strftime("%Y-%m-%d %H:%M:%S %Z")}: Embulk v#{Embulk::VERSION}"
+
     load_paths = []
     classpaths = []
     classpath_separator = java.io.File.pathSeparator

data/lib/embulk/data/new/README.md.erb

@@ -4,7 +4,7 @@ TODO: Write short description here
 
 ## Overview
 
-* **Plugin type**: <%= embulk_category %>
+* **Plugin type**: <%= display_category %>
 * **Load all or nothing**: yes
 * **Resume supported**: no
 

data/lib/embulk/guess/csv.rb

@@ -13,6 +13,17 @@ module Embulk
         "\"", "'"
       ]
 
+      ESCAPE_CANDIDATES = [
+        "\\"
+      ]
+
+      NULL_STRING_CANDIDATES = [
+        "null",
+        "NULL",
+        "#N/A",
+        "\\N",  # MySQL LOAD, Hive STORED AS TEXTFILE
+      ]
+
       # CsvParserPlugin.TRUE_STRINGS
       TRUE_STRINGS = Hash[*%w[
         true True TRUE
@@ -35,6 +46,13 @@ module Embulk
         quote = guess_quote(sample_lines, delim)
         parser_guessed["quote"] = quote ? quote : ''
 
+        escape = guess_escape(sample_lines, delim, quote)
+        parser_guessed["escape"] = escape ? escape : ''
+
+        null_string = guess_null_string(sample_lines, delim)
+        parser_guessed["null_string"] = null_string if null_string
+        # don't even set null_string to avoid confusion of null and 'null' in YAML format
+
         sample_records = sample_lines.map {|line| line.split(delim) }  # TODO use CsvTokenizer
         first_types = guess_field_types(sample_records[0, 1])
         other_types = guess_field_types(sample_records[1..-1])
@@ -119,6 +137,32 @@ module Embulk
         end
       end
 
+      def guess_escape(sample_lines, delim, optional_quote)
+        guessed = ESCAPE_CANDIDATES.map do |str|
+          if optional_quote
+            regexp = /#{Regexp.quote(str)}(?:#{Regexp.quote(delim)}|#{Regexp.quote(optional_quote)})/
+          else
+            regexp = /#{Regexp.quote(str)}#{Regexp.quote(delim)}/
+          end
+          counts = sample_lines.map {|line| line.scan(regexp).count }
+          count = counts.inject(0) {|r,c| r + c }
+          [str, count]
+        end.select {|str,count| count > 0 }.sort_by {|str,count| -count }
+        found = guessed.first
+        return found ? found[0] : nil
+      end
+
+      def guess_null_string(sample_lines, delim)
+        guessed = NULL_STRING_CANDIDATES.map do |str|
+          regexp = /(?:^|#{Regexp.quote(delim)})#{Regexp.quote(str)}(?:$|#{Regexp.quote(delim)})/
+          counts = sample_lines.map {|line| line.scan(regexp).count }
+          count = counts.inject(0) {|r,c| r + c }
+          [str, count]
+        end.select {|str,count| count > 0 }.sort_by {|str,count| -count }
+        found = guessed.first
+        return found ? found[0] : nil
+      end
+
       def guess_field_types(field_lines)
         column_lines = []
         field_lines.each do |fields|

data/lib/embulk/guess/time_format_guess.rb

@@ -231,7 +231,7 @@ module Embulk::Guess
         end
 
         if zm = /^#{TZ}$/.match(rest)
-          delimiters << zm["zone_space"] || ''
+          delimiters << (zm["zone_space"] || '')
           if zm["z"]
             # TODO ISO 8601
             parts << :zone_off

data/lib/embulk/version.rb

@@ -1,3 +1,3 @@
 module Embulk
-  VERSION = "0.4.2"
+  VERSION = "0.4.3"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: embulk
 version: !ruby/object:Gem::Version
-  version: 0.4.2
+  version: 0.4.3
 platform: ruby
 authors:
 - Sadayuki Furuhashi
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2015-02-16 00:00:00.000000000 Z
+date: 2015-02-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -161,7 +161,6 @@ files:
 - embulk-core/src/main/java/org/embulk/plugin/PluginSource.java
 - embulk-core/src/main/java/org/embulk/plugin/PluginSourceNotMatchException.java
 - embulk-core/src/main/java/org/embulk/plugin/PluginType.java
-- embulk-core/src/main/java/org/embulk/plugin/SetThreadContextClassLoader.java
 - embulk-core/src/main/java/org/embulk/spi/Buffer.java
 - embulk-core/src/main/java/org/embulk/spi/BufferAllocator.java
 - embulk-core/src/main/java/org/embulk/spi/Column.java
@@ -268,6 +267,7 @@ files:
 - embulk-docs/src/release/release-0.4.0.rst
 - embulk-docs/src/release/release-0.4.1.rst
 - embulk-docs/src/release/release-0.4.2.rst
+- embulk-docs/src/release/release-0.4.3.rst
 - embulk-standards/build.gradle
 - embulk-standards/src/main/java/org/embulk/standards/CsvFormatterPlugin.java
 - embulk-standards/src/main/java/org/embulk/standards/CsvParserPlugin.java
@@ -362,8 +362,8 @@ files:
 - classpath/bval-jsr303-0.5.jar
 - classpath/commons-beanutils-core-1.8.3.jar
 - classpath/commons-lang3-3.1.jar
-- classpath/embulk-core-0.4.2.jar
-- classpath/embulk-standards-0.4.2.jar
+- classpath/embulk-core-0.4.3.jar
+- classpath/embulk-standards-0.4.3.jar
 - classpath/guava-18.0.jar
 - classpath/guice-3.0.jar
 - classpath/guice-multibindings-3.0.jar

data/embulk-core/src/main/java/org/embulk/plugin/SetThreadContextClassLoader.java

@@ -1,19 +0,0 @@
-package org.embulk.plugin;
-
-public class SetThreadContextClassLoader
-        implements AutoCloseable
-{
-    private final ClassLoader original;
-
-    public SetThreadContextClassLoader(ClassLoader classLoader)
-    {
-        this.original = Thread.currentThread().getContextClassLoader();
-        Thread.currentThread().setContextClassLoader(classLoader);
-    }
-
-    @Override
-    public void close()
-    {
-        Thread.currentThread().setContextClassLoader(original);
-    }
-}