Merge branch 'develop'

Norconex · Feb 29, 2016 · cd280a1 · cd280a1
2 parents c4291a1 + 69529ae
commit cd280a1
Show file tree

Hide file tree

Showing 9 changed files with 57 additions and 8 deletions.
diff --git a/norconex-collector-filesystem/pom.xml b/norconex-collector-filesystem/pom.xml
@@ -19,7 +19,7 @@
   <modelVersion>4.0.0</modelVersion>
   <groupId>com.norconex.collectors</groupId>
   <artifactId>norconex-collector-filesystem</artifactId>
-  <version>2.3.0</version>
+  <version>2.4.0</version>
   <name>Norconex Filesystem Collector</name>
 
   <properties>
@@ -60,12 +60,12 @@
     <dependency>
       <groupId>com.norconex.collectors</groupId>
       <artifactId>norconex-collector-core</artifactId>
-      <version>1.3.0</version>
+      <version>1.4.0</version>
     </dependency>
     <dependency>
         <groupId>joda-time</groupId>
         <artifactId>joda-time</artifactId>
-        <version>2.7</version>
+        <version>2.9.2</version>
     </dependency>    
     <dependency>
 	    <groupId>org.apache.commons</groupId>

diff --git a/norconex-collector-filesystem/src/changes/changes.xml b/norconex-collector-filesystem/src/changes/changes.xml
@@ -7,6 +7,21 @@
   </properties>
   <body>
 
+    <release version="2.4.0" date="2016-02-28" description="Minor release">
+      <action dev="essiembre" type="add">
+        Now supports specifying relative paths in startPaths (for local
+        file systems only).
+      </action>
+      <action dev="essiembre" type="update">
+        The "log4j.properties" file has been moved from classes to the 
+        installation root directory.
+      </action>
+      <action dev="essiembre" type="update">
+        Maven dependency updates: Norconex Collector Core 1.4.0, 
+        Joda Time 2.9.2.
+      </action>
+    </release>
+
     <release version="2.3.0" date="2015-11-06" description="Feature release">
       <action dev="essiembre" type="update">
         Maven dependency updates: Norconex Collector Core 1.3.0 and Norconex

diff --git a/norconex-collector-filesystem/src/main/assembly/CLASSES_README.txt b/norconex-collector-filesystem/src/main/assembly/CLASSES_README.txt
@@ -0,0 +1,12 @@
+What is the "classes" folder?
+-----------------------------
+
+This folder is for Java developers wishing to create their custom features.
+You can put your custom Java classes in this folder and they will be 
+automatically picked up on the next execution (provided you are using
+the supplied .bat or .sh launch script).
+
+If you package your classes into a Jar file, put that file in the "lib" folder
+instead.
+
+You can safely delete this file.
diff --git a/norconex-collector-filesystem/src/main/assembly/dist.xml b/norconex-collector-filesystem/src/main/assembly/dist.xml
@@ -1,5 +1,5 @@
 <!-- 
-   Copyright 2010-2014 Norconex Inc.
+   Copyright 2010-2016 Norconex Inc.
 
    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.
@@ -36,6 +36,18 @@
       </includes>
       <outputDirectory>/examples</outputDirectory>
     </fileSet>
+    <fileSet>
+      <directory>${basedir}/src/main/assembly/</directory>
+      <excludes>
+        <exclude>dist.xml</exclude>
+        <exclude>CLASSES_README.txt</exclude>        
+      </excludes>
+      <includes>
+        <include>third-party/**</include>
+        <include>*.txt</include>
+      </includes>
+      <outputDirectory>/</outputDirectory>
+    </fileSet>
     <fileSet>
       <directory>${project.build.directory}/site/apidocs</directory>
       <outputDirectory>apidocs</outputDirectory>
@@ -55,6 +67,10 @@
   <files>
     <file>
       <source>src/main/assembly/log4j.properties</source>
+      <outputDirectory>/</outputDirectory>
+    </file>
+    <file>
+      <source>src/main/assembly/CLASSES_README.txt</source>
       <outputDirectory>/classes</outputDirectory>
     </file>
   </files>

diff --git a/...llector-filesystem/src/main/java/com/norconex/collector/fs/crawler/FilesystemCrawler.java b/...llector-filesystem/src/main/java/com/norconex/collector/fs/crawler/FilesystemCrawler.java
@@ -14,6 +14,7 @@
  */
 package com.norconex.collector.fs.crawler;
 
+import java.io.File;
 import java.io.FileInputStream;
 import java.io.IOException;
 
@@ -95,6 +96,11 @@ private void queueStartPaths(ICrawlDataStore crawlDataStore) {
         if (startPaths != null) {
             for (int i = 0; i < startPaths.length; i++) {
                 String startPath = startPaths[i];
+                // No protocol specified: we assume local file, and we get 
+                // the absolute version.
+                if (!startPath.contains("://")) {
+                    startPath = new File(startPath).getAbsolutePath();
+                }
                 executeQueuePipeline(
                         new BaseCrawlData(startPath), crawlDataStore);
             }

diff --git a/norconex-collector-filesystem/src/main/scripts/collector-fs.bat b/norconex-collector-filesystem/src/main/scripts/collector-fs.bat
@@ -2,5 +2,5 @@
 cd %~dp0
 set ROOT_DIR=%~dp0
 
-java -Dlog4j.configuration="file:///%ROOT_DIR%classes/log4j.properties" -Dfile.encoding=UTF8 -cp "./lib/*;./classes" com.norconex.collector.fs.FilesystemCollector %*
+java -Dlog4j.configuration="file:///%ROOT_DIR%log4j.properties" -Dfile.encoding=UTF8 -cp "./lib/*;./classes" com.norconex.collector.fs.FilesystemCollector %*
 
diff --git a/norconex-collector-filesystem/src/main/scripts/collector-fs.sh b/norconex-collector-filesystem/src/main/scripts/collector-fs.sh
@@ -3,4 +3,4 @@
 cd $(dirname $0)
 export ROOT_DIR=$(dirname $0)
 
-java -Dlog4j.configuration="file:${ROOT_DIR}/classes/log4j.properties" -Dfile.encoding=UTF8 -cp "./lib/*:./classes" com.norconex.collector.fs.FilesystemCollector "$@"
+java -Dlog4j.configuration="file:${ROOT_DIR}/log4j.properties" -Dfile.encoding=UTF8 -cp "./lib/*:./classes" com.norconex.collector.fs.FilesystemCollector "$@"
diff --git a/...ollector-filesystem/src/site/resources/examples/collector-filesystem-config-reference.xml b/...ollector-filesystem/src/site/resources/examples/collector-filesystem-config-reference.xml
@@ -207,7 +207,7 @@
          options.  Please refer to committer for complete documentation.
          Below is an example using the FileSystemCommitter.
          -->
-    <committer class="$committer.impl.FileSystemCommitter">
+    <committer class="$committer.core.impl.FileSystemCommitter">
       <directory>$workdir\crawledFiles</directory>
     </committer>
 

diff --git a/norconex-collector-filesystem/src/site/resources/examples/sample-config.xml b/norconex-collector-filesystem/src/site/resources/examples/sample-config.xml
@@ -56,7 +56,7 @@
         </postParseHandlers>
       </importer>
 
-      <committer class="com.norconex.committer.impl.FileSystemCommitter">
+      <committer class="com.norconex.committer.core.impl.FileSystemCommitter">
         <directory>${workdir}/crawledFiles</directory>
       </committer>
     </crawler>