jenkins/core/src/main/java/hudson/FilePath.java

1710 lines
63 KiB
Java
Raw Normal View History

/*
* The MIT License
*
* Copyright (c) 2004-2009, Sun Microsystems, Inc., Kohsuke Kawaguchi, Eric Lefevre-Ardant, Erik Ramfelt, Michael B. Donohue
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
package hudson;
import hudson.Launcher.LocalLauncher;
import hudson.Launcher.RemoteLauncher;
import hudson.model.Hudson;
import hudson.model.TaskListener;
import hudson.model.AbstractProject;
import hudson.model.Item;
import hudson.remoting.Callable;
import hudson.remoting.Channel;
import hudson.remoting.DelegatingCallable;
import hudson.remoting.Future;
import hudson.remoting.Pipe;
import hudson.remoting.RemoteOutputStream;
import hudson.remoting.VirtualChannel;
import hudson.remoting.RemoteInputStream;
import hudson.util.IOException2;
import hudson.util.HeadBufferingStream;
import hudson.util.FormValidation;
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
import static hudson.util.jna.GNUCLibrary.LIBC;
import static hudson.Util.fixEmpty;
import static hudson.FilePath.TarCompression.GZIP;
import org.apache.tools.ant.BuildException;
import org.apache.tools.ant.DirectoryScanner;
import org.apache.tools.ant.Project;
import org.apache.tools.ant.taskdefs.Copy;
import org.apache.tools.ant.types.FileSet;
import org.apache.tools.tar.TarEntry;
import org.apache.tools.tar.TarOutputStream;
import org.apache.tools.tar.TarInputStream;
import org.apache.tools.zip.ZipOutputStream;
import org.apache.tools.zip.ZipEntry;
import org.apache.commons.io.IOUtils;
import org.apache.commons.fileupload.FileItem;
import org.kohsuke.stapler.Stapler;
import java.io.BufferedInputStream;
import java.io.BufferedOutputStream;
import java.io.File;
import java.io.FileFilter;
import java.io.FileInputStream;
import java.io.FileOutputStream;
import java.io.FileWriter;
import java.io.IOException;
import java.io.InputStream;
import java.io.ObjectInputStream;
import java.io.ObjectOutputStream;
import java.io.OutputStream;
import java.io.Serializable;
import java.io.Writer;
import java.io.OutputStreamWriter;
import java.net.URI;
import java.net.URL;
import java.net.URLConnection;
import java.util.ArrayList;
import java.util.List;
import java.util.StringTokenizer;
import java.util.Arrays;
import java.util.Comparator;
import java.util.regex.Pattern;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;
import java.util.zip.GZIPOutputStream;
import java.util.zip.GZIPInputStream;
import java.util.zip.ZipInputStream;
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
import com.sun.jna.Native;
/**
* {@link File} like object with remoting support.
*
* <p>
* Unlike {@link File}, which always implies a file path on the current computer,
* {@link FilePath} represents a file path on a specific slave or the master.
*
* Despite that, {@link FilePath} can be used much like {@link File}. It exposes
* a bunch of operations (and we should add more operations as long as they are
* generally useful), and when invoked against a file on a remote node, {@link FilePath}
* executes the necessary code remotely, thereby providing semi-transparent file
* operations.
*
* <h2>Using {@link FilePath} smartly</h2>
* <p>
* The transparency makes it easy to write plugins without worrying too much about
* remoting, by making it works like NFS, where remoting happens at the file-system
* later.
*
* <p>
* But one should note that such use of remoting may not be optional. Sometimes,
* it makes more sense to move some computation closer to the data, as opposed to
* move the data to the computation. For example, if you are just computing a MD5
* digest of a file, then it would make sense to do the digest on the host where
* the file is located, as opposed to send the whole data to the master and do MD5
* digesting there.
*
* <p>
* {@link FilePath} supports this "code migration" by in the
* {@link #act(FileCallable)} method. One can pass in a custom implementation
* of {@link FileCallable}, to be executed on the node where the data is located.
* The following code shows the example:
*
* <pre>
* FilePath file = ...;
*
* // make 'file' a fresh empty directory.
* file.act(new FileCallable&lt;Void>() {
* // if 'file' is on a different node, this FileCallable will
* // be transfered to that node and executed there.
* public Void invoke(File f,VirtualChannel channel) {
* // f and file represents the same thing
* f.deleteContents();
* f.mkdirs();
* }
* });
* </pre>
*
* <p>
* When {@link FileCallable} is transfered to a remote node, it will be done so
* by using the same Java serialization scheme that the remoting module uses.
* See {@link Channel} for more about this.
*
* <p>
* {@link FilePath} itself can be sent over to a remote node as a part of {@link Callable}
* serialization. For example, sending a {@link FilePath} of a remote node to that
* node causes {@link FilePath} to become "local". Similarly, sending a
* {@link FilePath} that represents the local computer causes it to become "remote."
*
* @author Kohsuke Kawaguchi
*/
public final class FilePath implements Serializable {
/**
* When this {@link FilePath} represents the remote path,
* this field is always non-null on master (the field represents
* the channel to the remote slave.) When transferred to a slave via remoting,
* this field reverts back to null, since it's transient.
*
* When this {@link FilePath} represents a path on the master,
* this field is null on master. When transferred to a slave via remoting,
* this field becomes non-null, representing the {@link Channel}
* back to the master.
*
* This is used to determine whether we are running on the master or the slave.
*/
private transient VirtualChannel channel;
// since the platform of the slave might be different, can't use java.io.File
private final String remote;
/**
* Creates a {@link FilePath} that represents a path on the given node.
*
* @param channel
* To create a path that represents a remote path, pass in a {@link Channel}
* that's connected to that machine. If null, that means the local file path.
*/
public FilePath(VirtualChannel channel, String remote) {
this.channel = channel;
this.remote = remote;
}
/**
* To create {@link FilePath} that represents a "local" path.
*
* <p>
* A "local" path means a file path on the computer where the
* constructor invocation happened.
*/
public FilePath(File localPath) {
this.channel = null;
this.remote = localPath.getPath();
}
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
/**
* Construct a path starting with a base location.
* @param base starting point for resolution, and defines channel
* @param rel a path which if relative will be resolved against base
*/
public FilePath(FilePath base, String rel) {
this.channel = base.channel;
if(isAbsolute(rel)) {
// absolute
this.remote = rel;
} else
if(base.isUnix()) {
this.remote = base.remote+'/'+rel;
} else {
this.remote = base.remote+'\\'+rel;
}
}
private static boolean isAbsolute(String rel) {
return rel.startsWith("/") || DRIVE_PATTERN.matcher(rel).matches();
}
private static final Pattern DRIVE_PATTERN = Pattern.compile("[A-Za-z]:\\\\.+");
/**
* Checks if the remote path is Unix.
*/
private boolean isUnix() {
// if the path represents a local path, there' no need to guess.
if(!isRemote())
return File.pathSeparatorChar!=';';
// note that we can't use the usual File.pathSeparator and etc., as the OS of
// the machine where this code runs and the OS that this FilePath refers to may be different.
// Windows absolute path is 'X:\...', so this is usually a good indication of Windows path
if(remote.length()>3 && remote.charAt(1)==':' && remote.charAt(2)=='\\')
return false;
// Windows can handle '/' as a path separator but Unix can't,
// so err on Unix side
return remote.indexOf("\\")==-1;
}
public String getRemote() {
return remote;
}
/**
* Creates a zip file from this directory or a file and sends that to the given output stream.
*/
public void createZipArchive(OutputStream os) throws IOException, InterruptedException {
final OutputStream out = (channel!=null)?new RemoteOutputStream(os):os;
act(new FileCallable<Void>() {
private transient byte[] buf;
public Void invoke(File f, VirtualChannel channel) throws IOException {
buf = new byte[8192];
ZipOutputStream zip = new ZipOutputStream(out);
zip.setEncoding(System.getProperty("file.encoding"));
scan(f,zip,"");
zip.close();
return null;
}
private void scan(File f, ZipOutputStream zip, String path) throws IOException {
// Bitmask indicating directories in 'external attributes' of a ZIP archive entry.
final long BITMASK_IS_DIRECTORY = 1<<4;
if (f.canRead()) {
if(f.isDirectory()) {
ZipEntry dirZipEntry = new ZipEntry(path+f.getName()+'/');
// Setting this bit explicitly is needed by some unzipping applications (see HUDSON-3294).
dirZipEntry.setExternalAttributes(BITMASK_IS_DIRECTORY);
zip.putNextEntry(dirZipEntry);
zip.closeEntry();
for( File child : f.listFiles() )
scan(child,zip,path+f.getName()+'/');
} else {
zip.putNextEntry(new ZipEntry(path+f.getName()));
FileInputStream in = new FileInputStream(f);
int len;
while((len=in.read(buf))>0)
zip.write(buf,0,len);
in.close();
zip.closeEntry();
}
}
}
private static final long serialVersionUID = 1L;
});
}
/**
* Creates a zip file from this directory by only including the files that match the given glob.
*
* @param glob
* Ant style glob, like "**&#x2F;*.xml". If empty or null, this method
* works like {@link #createZipArchive(OutputStream)}
*
* @since 1.129
*/
public void createZipArchive(OutputStream os, final String glob) throws IOException, InterruptedException {
if(glob==null || glob.length()==0) {
createZipArchive(os);
return;
}
final OutputStream out = (channel!=null)?new RemoteOutputStream(os):os;
act(new FileCallable<Void>() {
public Void invoke(File dir, VirtualChannel channel) throws IOException {
byte[] buf = new byte[8192];
ZipOutputStream zip = new ZipOutputStream(out);
zip.setEncoding(System.getProperty("file.encoding"));
for( String entry : glob(dir,glob) ) {
File file = new File(dir,entry);
if (file.canRead()) {
zip.putNextEntry(new ZipEntry(dir.getName()+'/'+entry));
FileInputStream in = new FileInputStream(file);
int len;
while((len=in.read(buf))>0)
zip.write(buf,0,len);
in.close();
zip.closeEntry();
}
}
zip.close();
return null;
}
private static final long serialVersionUID = 1L;
});
}
/**
* When this {@link FilePath} represents a zip file, extracts that zip file.
*
* @param target
* Target directory to expand files to. All the necessary directories will be created.
* @since 1.248
* @see #unzipFrom(InputStream)
*/
public void unzip(final FilePath target) throws IOException, InterruptedException {
target.act(new FileCallable<Void>() {
public Void invoke(File dir, VirtualChannel channel) throws IOException {
unzip(dir,FilePath.this.read());
return null;
}
private static final long serialVersionUID = 1L;
});
}
/**
* When this {@link FilePath} represents a tar file, extracts that tar file.
*
* @param target
* Target directory to expand files to. All the necessary directories will be created.
* @param compression
* Compression mode of this tar file.
* @since 1.292
* @see #untarFrom(InputStream, TarCompression)
*/
public void untar(final FilePath target, final TarCompression compression) throws IOException, InterruptedException {
target.act(new FileCallable<Void>() {
public Void invoke(File dir, VirtualChannel channel) throws IOException {
readFromTar(FilePath.this.getName(),dir,compression.extract(FilePath.this.read()));
return null;
}
private static final long serialVersionUID = 1L;
});
}
/**
* Reads the given InputStream as a zip file and extracts it into this directory.
*
* @param _in
* The stream will be closed by this method after it's fully read.
* @since 1.283
* @see #unzip(FilePath)
*/
public void unzipFrom(InputStream _in) throws IOException, InterruptedException {
final InputStream in = new RemoteInputStream(_in);
act(new FileCallable<Void>() {
public Void invoke(File dir, VirtualChannel channel) throws IOException {
unzip(dir, in);
return null;
}
private static final long serialVersionUID = 1L;
});
}
private void unzip(File dir, InputStream in) throws IOException {
dir = dir.getAbsoluteFile(); // without absolutization, getParentFile below seems to fail
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
ZipInputStream zip = new ZipInputStream(new BufferedInputStream(in));
java.util.zip.ZipEntry e;
try {
while((e=zip.getNextEntry())!=null) {
File f = new File(dir,e.getName());
if(e.isDirectory()) {
f.mkdirs();
} else {
File p = f.getParentFile();
if(p!=null) p.mkdirs();
FileOutputStream out = new FileOutputStream(f);
try {
IOUtils.copy(zip, out);
} finally {
out.close();
}
f.setLastModified(e.getTime());
zip.closeEntry();
}
}
} finally {
zip.close();
}
}
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
/**
* Absolutizes this {@link FilePath} and returns the new one.
*/
public FilePath absolutize() throws IOException, InterruptedException {
return new FilePath(channel,act(new FileCallable<String>() {
public String invoke(File f, VirtualChannel channel) throws IOException {
return f.getAbsolutePath();
}
}));
}
/**
* Supported tar file compression methods.
*/
public enum TarCompression {
NONE {
public InputStream extract(InputStream in) {
return in;
}
public OutputStream compress(OutputStream out) {
return out;
}
},
GZIP {
public InputStream extract(InputStream _in) throws IOException {
HeadBufferingStream in = new HeadBufferingStream(_in,SIDE_BUFFER_SIZE);
try {
return new GZIPInputStream(in,8192);
} catch (IOException e) {
// various people reported "java.io.IOException: Not in GZIP format" here, so diagnose this problem better
in.fillSide();
throw new IOException2(e.getMessage()+"\nstream="+Util.toHexString(in.getSideBuffer()),e);
}
}
public OutputStream compress(OutputStream out) throws IOException {
return new GZIPOutputStream(new BufferedOutputStream(out));
}
};
public abstract InputStream extract(InputStream in) throws IOException;
public abstract OutputStream compress(OutputStream in) throws IOException;
}
/**
* Reads the given InputStream as a tar file and extracts it into this directory.
*
* @param _in
* The stream will be closed by this method after it's fully read.
* @param compression
* The compression method in use.
* @since 1.292
*/
public void untarFrom(InputStream _in, final TarCompression compression) throws IOException, InterruptedException {
try {
final InputStream in = new RemoteInputStream(_in);
act(new FileCallable<Void>() {
public Void invoke(File dir, VirtualChannel channel) throws IOException {
readFromTar("input stream",dir, compression.extract(in));
return null;
}
private static final long serialVersionUID = 1L;
});
} finally {
IOUtils.closeQuietly(_in);
}
}
/**
* Given a tgz/zip file, extracts it to the given target directory, if necessary.
*
* <p>
* This method is a convenience method designed for installing a binary package to a location
* that supports upgrade and downgrade. Specifically,
*
* <ul>
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
* <li>If the target directory doesn't exist {@linkplain #mkdirs() it'll be created}.
* <li>The timestamp of the .tgz file is left in the installation directory upon extraction.
* <li>If the timestamp left in the directory doesn't match with the timestamp of the current archive file,
* the directory contents will be discarded and the archive file will be re-extracted.
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
* <li>If the connection is refused but the target directory already exists, it is left alone.
* </ul>
*
* @param archive
* The resource that represents the tgz/zip file. This URL must support the "Last-Modified" header.
* (Most common usage is to get this from {@link ClassLoader#getResource(String)})
* @param listener
* If non-null, a message will be printed to this listener once this method decides to
* extract an archive.
* @return
* true if the archive was extracted. false if the extraction was skipped because the target directory
* was considered up to date.
* @since 1.299
*/
public boolean installIfNecessaryFrom(URL archive, TaskListener listener, String message) throws IOException, InterruptedException {
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
URLConnection con;
try {
con = archive.openConnection();
con.connect();
} catch (IOException x) {
if (this.exists()) {
// Cannot connect now, so assume whatever was last unpacked is still OK.
if (listener != null) {
listener.getLogger().println("Skipping installation of " + archive + " to " + remote + ": " + x);
}
return false;
} else {
throw x;
}
}
long sourceTimestamp = con.getLastModified();
FilePath timestamp = this.child(".timestamp");
if(this.exists()) {
if(timestamp.exists() && sourceTimestamp ==timestamp.lastModified())
return false; // already up to date
this.deleteContents();
}
if(listener!=null)
listener.getLogger().println(message);
if(archive.toExternalForm().endsWith(".zip"))
unzipFrom(con.getInputStream());
else
untarFrom(con.getInputStream(),GZIP);
timestamp.touch(sourceTimestamp);
return true;
}
/**
* Reads the URL on the current VM, and writes all the data to this {@link FilePath}
* (this is different from resolving URL remotely.)
*
* @since 1.293
*/
public void copyFrom(URL url) throws IOException, InterruptedException {
InputStream in = url.openStream();
try {
copyFrom(in);
} finally {
in.close();
}
}
/**
* Replaces the content of this file by the data from the given {@link InputStream}.
*
* @since 1.293
*/
public void copyFrom(InputStream in) throws IOException, InterruptedException {
OutputStream os = write();
try {
IOUtils.copy(in, os);
} finally {
os.close();
}
}
/**
* Place the data from {@link FileItem} into the file location specified by this {@link FilePath} object.
*/
public void copyFrom(FileItem file) throws IOException, InterruptedException {
if(channel==null) {
try {
file.write(new File(remote));
} catch (IOException e) {
throw e;
} catch (Exception e) {
throw new IOException2(e);
}
} else {
InputStream i = file.getInputStream();
OutputStream o = write();
try {
IOUtils.copy(i,o);
} finally {
o.close();
i.close();
}
}
}
/**
* Code that gets executed on the machine where the {@link FilePath} is local.
* Used to act on {@link FilePath}.
*
* @see FilePath#act(FileCallable)
*/
public static interface FileCallable<T> extends Serializable {
/**
* Performs the computational task on the node where the data is located.
*
* @param f
* {@link File} that represents the local file that {@link FilePath} has represented.
* @param channel
* The "back pointer" of the {@link Channel} that represents the communication
* with the node from where the code was sent.
*/
T invoke(File f, VirtualChannel channel) throws IOException;
}
/**
* Executes some program on the machine that this {@link FilePath} exists,
* so that one can perform local file operations.
*/
public <T> T act(final FileCallable<T> callable) throws IOException, InterruptedException {
if(channel!=null) {
// run this on a remote system
try {
return channel.call(new FileCallableWrapper<T>(callable));
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
} catch (AbortException e) {
throw e; // pass through so that the caller can catch it as AbortException
} catch (IOException e) {
// wrap it into a new IOException so that we get the caller's stack trace as well.
throw new IOException2("remote file operation failed",e);
}
} else {
// the file is on the local machine.
return callable.invoke(new File(remote), Hudson.MasterComputer.localChannel);
}
}
/**
* Executes some program on the machine that this {@link FilePath} exists,
* so that one can perform local file operations.
*/
public <T> Future<T> actAsync(final FileCallable<T> callable) throws IOException, InterruptedException {
try {
return (channel!=null ? channel : Hudson.MasterComputer.localChannel)
.callAsync(new FileCallableWrapper<T>(callable));
} catch (IOException e) {
// wrap it into a new IOException so that we get the caller's stack trace as well.
throw new IOException2("remote file operation failed",e);
}
}
/**
* Executes some program on the machine that this {@link FilePath} exists,
* so that one can perform local file operations.
*/
public <V,E extends Throwable> V act(Callable<V,E> callable) throws IOException, InterruptedException, E {
if(channel!=null) {
// run this on a remote system
return channel.call(callable);
} else {
// the file is on the local machine
return callable.call();
}
}
/**
* Converts this file to the URI, relative to the machine
* on which this file is available.
*/
public URI toURI() throws IOException, InterruptedException {
return act(new FileCallable<URI>() {
public URI invoke(File f, VirtualChannel channel) {
return f.toURI();
}
});
}
/**
* Creates this directory.
*/
public void mkdirs() throws IOException, InterruptedException {
if(!act(new FileCallable<Boolean>() {
public Boolean invoke(File f, VirtualChannel channel) throws IOException {
if(f.mkdirs() || f.exists())
return true; // OK
// following Ant <mkdir> task to avoid possible race condition.
try {
Thread.sleep(10);
} catch (InterruptedException e) {
// ignore
}
return f.mkdirs() || f.exists();
}
}))
throw new IOException("Failed to mkdirs: "+remote);
}
/**
* Deletes this directory, including all its contents recursively.
*/
public void deleteRecursive() throws IOException, InterruptedException {
act(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
Util.deleteRecursive(f);
return null;
}
});
}
/**
* Deletes all the contents of this directory, but not the directory itself
*/
public void deleteContents() throws IOException, InterruptedException {
act(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
Util.deleteContentsRecursive(f);
return null;
}
});
}
/**
* Gets just the file name portion.
*
* This method assumes that the file name is the same between local and remote.
*/
public String getName() {
String r = remote;
if(r.endsWith("\\") || r.endsWith("/"))
r = r.substring(0,r.length()-1);
int len = r.length()-1;
while(len>=0) {
char ch = r.charAt(len);
if(ch=='\\' || ch=='/')
break;
len--;
}
return r.substring(len+1);
}
/**
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
* The same as {@link FilePath#FilePath(FilePath,String)} but more OO.
* @param rel a relative or absolute path
* @return a file on the same channel
*/
public FilePath child(String rel) {
return new FilePath(this,rel);
}
/**
* Gets the parent file.
*/
public FilePath getParent() {
int len = remote.length()-1;
while(len>=0) {
char ch = remote.charAt(len);
if(ch=='\\' || ch=='/')
break;
len--;
}
return new FilePath( channel, remote.substring(0,len) );
}
/**
* Creates a temporary file in the directory that this {@link FilePath} object designates.
*/
public FilePath createTempFile(final String prefix, final String suffix) throws IOException, InterruptedException {
try {
return new FilePath(this,act(new FileCallable<String>() {
public String invoke(File dir, VirtualChannel channel) throws IOException {
File f = File.createTempFile(prefix, suffix, dir);
return f.getName();
}
}));
} catch (IOException e) {
throw new IOException2("Failed to create a temp file on "+remote,e);
}
}
/**
* Creates a temporary file in this directory and set the contents by the
* given text (encoded in the platform default encoding)
*/
public FilePath createTextTempFile(final String prefix, final String suffix, final String contents) throws IOException, InterruptedException {
return createTextTempFile(prefix,suffix,contents,true);
}
/**
* Creates a temporary file in this directory and set the contents by the
* given text (encoded in the platform default encoding)
*/
public FilePath createTextTempFile(final String prefix, final String suffix, final String contents, final boolean inThisDirectory) throws IOException, InterruptedException {
try {
return new FilePath(channel,act(new FileCallable<String>() {
public String invoke(File dir, VirtualChannel channel) throws IOException {
if(!inThisDirectory)
dir = new File(System.getProperty("java.io.tmpdir"));
else
dir.mkdirs();
File f;
try {
f = File.createTempFile(prefix, suffix, dir);
} catch (IOException e) {
throw new IOException2("Failed to create a temporary directory in "+dir,e);
}
Writer w = new FileWriter(f);
w.write(contents);
w.close();
return f.getAbsolutePath();
}
}));
} catch (IOException e) {
throw new IOException2("Failed to create a temp file on "+remote,e);
}
}
/**
* Deletes this file.
*/
public boolean delete() throws IOException, InterruptedException {
return act(new FileCallable<Boolean>() {
public Boolean invoke(File f, VirtualChannel channel) throws IOException {
return f.delete();
}
});
}
/**
* Checks if the file exists.
*/
public boolean exists() throws IOException, InterruptedException {
return act(new FileCallable<Boolean>() {
public Boolean invoke(File f, VirtualChannel channel) throws IOException {
return f.exists();
}
});
}
/**
* Gets the last modified time stamp of this file, by using the clock
* of the machine where this file actually resides.
*
* @see File#lastModified()
* @see #touch(long)
*/
public long lastModified() throws IOException, InterruptedException {
return act(new FileCallable<Long>() {
public Long invoke(File f, VirtualChannel channel) throws IOException {
return f.lastModified();
}
});
}
/**
* Creates a file (if not already exist) and sets the timestamp.
*
* @since 1.299
*/
public void touch(final long timestamp) throws IOException, InterruptedException {
act(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
if(!f.exists())
new FileOutputStream(f).close();
if(!f.setLastModified(timestamp))
throw new IOException("Failed to set the timestamp of "+f+" to "+timestamp);
return null;
}
});
}
/**
* Checks if the file is a directory.
*/
public boolean isDirectory() throws IOException, InterruptedException {
return act(new FileCallable<Boolean>() {
public Boolean invoke(File f, VirtualChannel channel) throws IOException {
return f.isDirectory();
}
});
}
/**
* Returns the file size in bytes.
*
* @since 1.129
*/
public long length() throws IOException, InterruptedException {
return act(new FileCallable<Long>() {
public Long invoke(File f, VirtualChannel channel) throws IOException {
return f.length();
}
});
}
Merged revisions 17324-17325,17428,17430,17473-17475,17481-17482,17555-17563,17571,17593-17595,17619-17620,17637-17644,17656-17657,17663-17665,17667-17670,17730,17753,17755-17758,17763-17768,17770-17771,17901,17906,17911-17912,17958,17961-17962,17970-17977 via svnmerge from https://www.dev.java.net/svn/hudson/branches/tool-autoinstallation ........ r17324 | jglick | 2009-04-20 18:24:32 -0700 (Mon, 20 Apr 2009) | 2 lines Defensive cloning and Javadoc corrections. (Though it seems no one is using the default impls of these methods yet anyway.) ........ r17325 | jglick | 2009-04-20 18:29:06 -0700 (Mon, 20 Apr 2009) | 1 line Deprecation of delegate method is intentional. ........ r17428 | jglick | 2009-04-23 08:07:14 -0700 (Thu, 23 Apr 2009) | 1 line Noting ToolLocationTranslator more prominently. ........ r17430 | jglick | 2009-04-23 09:16:40 -0700 (Thu, 23 Apr 2009) | 1 line Unused imports. ........ r17473 | jglick | 2009-04-23 19:21:11 -0700 (Thu, 23 Apr 2009) | 1 line Allow Plugin.configure to see the StaplerRequest, needed for some kinds of calls. ........ r17474 | jglick | 2009-04-23 19:25:54 -0700 (Thu, 23 Apr 2009) | 3 lines Initial sketch of plugin to manage tool installations across slaves. (Should really be under ../plugins/ but that was not branched, so putting here for the moment.) ........ r17475 | jglick | 2009-04-23 20:29:36 -0700 (Thu, 23 Apr 2009) | 1 line Typo in Javadoc. ........ r17481 | jglick | 2009-04-23 22:46:26 -0700 (Thu, 23 Apr 2009) | 1 line Unzip of an unbuffered stream is painfully slow. ........ r17482 | jglick | 2009-04-23 22:48:01 -0700 (Thu, 23 Apr 2009) | 1 line ZIP-based tool installer now functional. ........ r17555 | jglick | 2009-04-27 17:50:38 -0700 (Mon, 27 Apr 2009) | 1 line Build against newest trunk. ........ r17556 | jglick | 2009-04-27 18:38:28 -0700 (Mon, 27 Apr 2009) | 1 line Adding help. ........ r17557 | jglick | 2009-04-27 18:57:46 -0700 (Mon, 27 Apr 2009) | 1 line Moving toolName & label config GUI into generic files. ........ r17558 | jglick | 2009-04-27 19:37:30 -0700 (Mon, 27 Apr 2009) | 1 line Typo. ........ r17559 | jglick | 2009-04-27 20:07:47 -0700 (Mon, 27 Apr 2009) | 1 line Second tool installer type based on a freeform command. ........ r17560 | jglick | 2009-04-27 20:24:36 -0700 (Mon, 27 Apr 2009) | 1 line installIfNecessaryFrom will now just return false in case the target dir exists but the URL cannot be opened. ........ r17561 | jglick | 2009-04-27 20:25:22 -0700 (Mon, 27 Apr 2009) | 2 lines chmod only if something was actually installed. Use a+x, not o+x. ........ r17562 | jglick | 2009-04-27 20:26:46 -0700 (Mon, 27 Apr 2009) | 1 line Comment only. ........ r17563 | jglick | 2009-04-27 20:41:12 -0700 (Mon, 27 Apr 2009) | 1 line I18N. ........ r17571 | jglick | 2009-04-28 08:15:31 -0700 (Tue, 28 Apr 2009) | 2 lines Wiki link. ........ r17593 | jglick | 2009-04-28 15:38:33 -0700 (Tue, 28 Apr 2009) | 1 line Turn off echo mode, it's just noise. ........ r17594 | jglick | 2009-04-28 15:39:29 -0700 (Tue, 28 Apr 2009) | 2 lines Synchronizing access to installers for a given node/tool combo. Sending log messages to Hudson log in real time. ........ r17595 | jglick | 2009-04-28 15:46:29 -0700 (Tue, 28 Apr 2009) | 1 line Copyrights. ........ r17619 | jglick | 2009-04-29 06:25:23 -0700 (Wed, 29 Apr 2009) | 1 line Improved Javadoc. ........ r17620 | jglick | 2009-04-29 06:51:11 -0700 (Wed, 29 Apr 2009) | 1 line Demonstraing how to use DLJ to mechanically install official JDK releases on supported operating systems. ........ r17637 | kohsuke | 2009-04-30 16:15:24 -0700 (Thu, 30 Apr 2009) | 1 line fixed what appears to be a C&P mistake ........ r17638 | kohsuke | 2009-04-30 16:16:21 -0700 (Thu, 30 Apr 2009) | 1 line show YUI logger if YUI debugging mode is set. ........ r17639 | kohsuke | 2009-04-30 16:52:28 -0700 (Thu, 30 Apr 2009) | 7 lines Fixed a UI script problem when <hetero-list> nests inside <repeatable>. The cause of the problem was that the behavior for inner <hetero-list> kicks in first, altering HTML (by removing "prototypes"), before behavior for outer <repetable> kicks in and removes the master copy from DOM tree. So <repeatable> ends up computing a master copy after some elements are removed, and hence when we try to reinsert them later by clicking "Add", it'll get incorrect copy, and re-execution of the <hetero-list> behavior bombs out. I fixed this by skipping those HTML-altering behaviors if they show up in the part of DOM tree that's supposed to be snapshotted and removed. ........ r17640 | kohsuke | 2009-04-30 16:57:58 -0700 (Thu, 30 Apr 2009) | 7 lines Added ToolProperty to ToolInstallation (which acts like NodeProperty to Node --- basically decorates the base object with additional properties.) The original motivation is to implement ToolInstaller this way, if Jesse is OK. But decorating tools seem to have some other interesting use, such as a plugin for configuring $M2_HOME/conf/settings.xml, or a plugin to insert unlimited cryptography extension to JDK, or a plugin to add a few more jars into $ANT_HOME/lib. For this to work, ToolInstallations that want properties needs to be modified to use a new constructor. For a starter, I modified the JDK class to do this. I also took the opportunity to modernize JDK/config.jelly. ........ r17641 | kohsuke | 2009-04-30 17:02:04 -0700 (Thu, 30 Apr 2009) | 1 line this appears to be a left-over from early days when we didn't have a means to do form validation ........ r17642 | kohsuke | 2009-04-30 17:02:35 -0700 (Thu, 30 Apr 2009) | 1 line we don't need no.such.JDK but we need name. rebranching ........ r17643 | kohsuke | 2009-04-30 17:03:22 -0700 (Thu, 30 Apr 2009) | 1 line rebranched from Hudson's configure.properties ........ r17644 | kohsuke | 2009-04-30 17:05:31 -0700 (Thu, 30 Apr 2009) | 1 line removed unnecessary translations ........ r17656 | kohsuke | 2009-05-01 11:27:36 -0700 (Fri, 01 May 2009) | 1 line Less hack for improved type safety. Asking people to write more code is not necessarily bad as long as what they need to do is obvious to them. ........ r17657 | kohsuke | 2009-05-01 12:07:20 -0700 (Fri, 01 May 2009) | 5 lines The toolautoinst part of the change that corresponds to rev.17640. ToolInstallers are now hooked under ToolInstallation through InstallSourceProperty as a ToolProperty. This allows the relevant UIs to be shown in one place. Still pending Jesse's feedback. Making it a single commit to simplify a revert if necessary ........ r17663 | kohsuke | 2009-05-01 13:23:51 -0700 (Fri, 01 May 2009) | 1 line doc improvement ........ r17664 | kohsuke | 2009-05-01 13:24:10 -0700 (Fri, 01 May 2009) | 1 line needs this to allow Stapler to databind hetero-list automatically ........ r17665 | jglick | 2009-05-01 13:31:27 -0700 (Fri, 01 May 2009) | 1 line Warnings. ........ r17667 | kohsuke | 2009-05-01 14:20:53 -0700 (Fri, 01 May 2009) | 1 line added a method to do chmod ........ r17668 | kohsuke | 2009-05-01 14:30:35 -0700 (Fri, 01 May 2009) | 1 line prefer JDK6 version to avoid unnecessary JNA loading ........ r17669 | kohsuke | 2009-05-01 14:33:09 -0700 (Fri, 01 May 2009) | 1 line no, I must have confused this. ........ r17670 | kohsuke | 2009-05-01 15:57:11 -0700 (Fri, 01 May 2009) | 2 lines - Automated JDK installer and its test. - Minor update to the signature of ToolInstaller for supplying the expected default location. ........ r17730 | kohsuke | 2009-05-04 11:27:34 -0700 (Mon, 04 May 2009) | 1 line avoid NPE as reported in HUDSON-3608 ........ r17753 | jglick | 2009-05-04 19:34:08 -0700 (Mon, 04 May 2009) | 2 lines Moved toolautoinst functionality into core. ........ r17755 | jglick | 2009-05-04 20:28:34 -0700 (Mon, 04 May 2009) | 8 lines Added TaskListener parameter and IOException and InterruptedException to several methods: Run.getEnvironment NodeSpecific.forNode ToolLocationTranslator.getToolHome ToolLocationNodeProperty.getToolHome ToolInstallation.translateFor ProcessCache.Factory.getMavenInstallation ProcessCache.Factory.getJava ........ r17756 | jglick | 2009-05-04 20:36:56 -0700 (Mon, 04 May 2009) | 1 line expectedLocation was never strictly necessary, and will soon no longer be used even by JDKInstaller. ........ r17757 | jglick | 2009-05-04 20:40:41 -0700 (Mon, 04 May 2009) | 1 line Not sure yet what @since should be. ........ r17758 | jglick | 2009-05-04 20:41:51 -0700 (Mon, 04 May 2009) | 1 line getDefaultInstallers method added. Not yet in use. ........ r17763 | kohsuke | 2009-05-05 10:50:36 -0700 (Tue, 05 May 2009) | 1 line deprecated redundant getJavaHome method ........ r17764 | kohsuke | 2009-05-05 12:10:45 -0700 (Tue, 05 May 2009) | 1 line added general-purpose property defaulting mechanism, and added default tool installer for JDK ........ r17765 | jglick | 2009-05-05 12:35:56 -0700 (Tue, 05 May 2009) | 1 line capitalize("") -> "" ........ r17766 | jglick | 2009-05-05 12:49:52 -0700 (Tue, 05 May 2009) | 1 line Missing @since. ........ r17767 | jglick | 2009-05-05 13:01:41 -0700 (Tue, 05 May 2009) | 1 line Unchecked warning. ........ r17768 | jglick | 2009-05-05 13:19:49 -0700 (Tue, 05 May 2009) | 2 lines Two unsuccessful attempts at form validation. ........ r17770 | jglick | 2009-05-05 13:48:37 -0700 (Tue, 05 May 2009) | 1 line Better class names. ........ r17771 | kohsuke | 2009-05-05 14:12:21 -0700 (Tue, 05 May 2009) | 5 lines brought the form validation check back. The trick is to use the 'value' parameter. In the end we'd expect this check to go away anyway, as we change input field to drop-down list, but in the mean time, this is why it didn't work. Maybe we probably need more static code checkers? Like FindBugs rules? ........ r17901 | kohsuke | 2009-05-09 19:15:40 -0700 (Sat, 09 May 2009) | 1 line having two toolHome is odd. ........ r17906 | kohsuke | 2009-05-10 11:09:13 -0700 (Sun, 10 May 2009) | 1 line adding the generalization of UpdateCenter, but for compatibility reasons, we are leaving UpdateCenter as-is. ........ r17911 | kohsuke | 2009-05-10 17:02:53 -0700 (Sun, 10 May 2009) | 1 line integrated json-lib with a bit of bug fix ........ r17912 | kohsuke | 2009-05-10 17:13:05 -0700 (Sun, 10 May 2009) | 1 line JDK list is now fetched from hudson.dev.java.net ........ r17958 | kohsuke | 2009-05-11 13:44:25 -0700 (Mon, 11 May 2009) | 48 lines Handle AbortException better to avoid error like this: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Failed to delete directory: C:\files\hudson\workspace\hudson\hudson\main\remoting\target. Reason: Unable to delete file C:\files\hudson\workspace\hudson\hudson\main\remoting\target\remoting-1.304-SNAPSHOT.jar [INFO] ------------------------------------------------------------------------ [INFO] For more information, run Maven with the -e switch [INFO] ------------------------------------------------------------------------ [INFO] Total time: 31 seconds [INFO] Finished at: Sun May 10 18:15:33 PDT 2009 [INFO] Final Memory: 25M/48M [INFO] ------------------------------------------------------------------------ Recording fingerprints Archiving artifacts Recording test results ERROR: Failed to archive JUnit reports hudson.util.IOException2: remote file operation failed at hudson.FilePath.act(FilePath.java:596) at hudson.tasks.junit.JUnitResultArchiver.perform(JUnitResultArchiver.java:82) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:372) at hudson.model.AbstractBuild$AbstractRunner.performAllBuildStep(AbstractBuild.java:360) at hudson.model.Build$RunnerImpl.post2(Build.java:183) at hudson.model.AbstractBuild$AbstractRunner.post(AbstractBuild.java:345) at hudson.model.Run.run(Run.java:943) at hudson.model.Build.run(Build.java:112) at hudson.model.ResourceController.execute(ResourceController.java:93) at hudson.model.Executor.run(Executor.java:119) Caused by: hudson.AbortException: Test reports were found but none of them are new. Did tests run? For example, \files\hudson\workspace\hudson\hudson\main\core\target\surefire-reports\TEST-hudson.BulkChangeTest.xml is 2 days 23 hr old at hudson.tasks.junit.TestResult.parse(TestResult.java:134) at hudson.tasks.junit.TestResult.<init>(TestResult.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:95) at hudson.tasks.junit.JUnitResultArchiver$1.invoke(JUnitResultArchiver.java:82) at hudson.FilePath$FileCallableWrapper.call(FilePath.java:1583) at hudson.remoting.UserRequest.perform(UserRequest.java:92) at hudson.remoting.UserRequest.perform(UserRequest.java:46) at hudson.remoting.Request$2.run(Request.java:236) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:619) ........ r17961 | kohsuke | 2009-05-11 15:16:39 -0700 (Mon, 11 May 2009) | 1 line adding Windows JDK silent installation ........ r17962 | kohsuke | 2009-05-11 16:08:13 -0700 (Mon, 11 May 2009) | 1 line Fixing JDK offline installatino for Windows ........ r17970 | kohsuke | 2009-05-11 19:37:59 -0700 (Mon, 11 May 2009) | 1 line Fixing a problem that prevents the usage of this tag in <repeatable> block, which reinserts the same fragment over and over. So statically assigning IDs won't work ........ r17971 | kohsuke | 2009-05-11 22:17:09 -0700 (Mon, 11 May 2009) | 1 line continuing to remove static ID generation, which breaks the tag when used inside <repetable> tag. ........ r17972 | kohsuke | 2009-05-11 22:19:11 -0700 (Mon, 11 May 2009) | 1 line I don't know when this started, but even in the hudson-dev:run mode, the versionis no longer '?', so we need a different way to force reload ........ r17973 | kohsuke | 2009-05-11 22:26:34 -0700 (Mon, 11 May 2009) | 1 line adjusted the text a bit since multiple "add" and "delete"s were rather confusing ........ r17974 | kohsuke | 2009-05-11 22:31:58 -0700 (Mon, 11 May 2009) | 1 line added help screen ........ r17975 | kohsuke | 2009-05-11 22:41:39 -0700 (Mon, 11 May 2009) | 1 line form validation now works against the checkbox, too ........ r17976 | kohsuke | 2009-05-11 22:43:00 -0700 (Mon, 11 May 2009) | 1 line with auto installation, we expect users to specify a directory that doesn't exist yet, so don't report it as an error ........ r17977 | kohsuke | 2009-05-11 22:43:24 -0700 (Mon, 11 May 2009) | 1 line improving the text ........ git-svn-id: https://hudson.dev.java.net/svn/hudson/trunk/hudson/main@17985 71c3de6d-444a-0410-be80-ed276b4c234a
2009-05-12 21:59:10 +08:00
/**
* Sets the file permission.
*
* On Windows, no-op.
*
* @since 1.303
*/
public void chmod(final int mask) throws IOException, InterruptedException {
if(!isUnix()) return;
act(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
if(LIBC.chmod(f.getAbsolutePath(),mask)!=0)
throw new IOException("Failed to chmod "+f+" : "+LIBC.strerror(Native.getLastError()));
return null;
}
});
}
/**
* List up files and directories in this directory.
*
* <p>
* This method returns direct children of the directory denoted by the 'this' object.
*/
public List<FilePath> list() throws IOException, InterruptedException {
return list((FileFilter)null);
}
/**
* List up subdirectories.
*
* @return can be empty but never null. Doesn't contain "." and ".."
*/
public List<FilePath> listDirectories() throws IOException, InterruptedException {
return list(new DirectoryFilter());
}
private static final class DirectoryFilter implements FileFilter, Serializable {
public boolean accept(File f) {
return f.isDirectory();
}
private static final long serialVersionUID = 1L;
}
/**
* List up files in this directory, just like {@link File#listFiles(FileFilter)}.
*
* @param filter
* The optional filter used to narrow down the result.
* If non-null, must be {@link Serializable}.
* If this {@link FilePath} represents a remote path,
* the filter object will be executed on the remote machine.
*/
public List<FilePath> list(final FileFilter filter) throws IOException, InterruptedException {
return act(new FileCallable<List<FilePath>>() {
public List<FilePath> invoke(File f, VirtualChannel channel) throws IOException {
File[] children = f.listFiles(filter);
if(children ==null) return null;
ArrayList<FilePath> r = new ArrayList<FilePath>(children.length);
for (File child : children)
r.add(new FilePath(child));
return r;
}
});
}
/**
* List up files in this directory that matches the given Ant-style filter.
*
* @param includes
* See {@link FileSet} for the syntax. String like "foo/*.zip" or "foo/*&#42;/*.xml"
* @return
* can be empty but always non-null.
*/
public FilePath[] list(final String includes) throws IOException, InterruptedException {
return act(new FileCallable<FilePath[]>() {
public FilePath[] invoke(File f, VirtualChannel channel) throws IOException {
String[] files = glob(f,includes);
FilePath[] r = new FilePath[files.length];
for( int i=0; i<r.length; i++ )
r[i] = new FilePath(new File(f,files[i]));
return r;
}
});
}
/**
* Runs Ant glob expansion.
*
* @return
* A set of relative file names from the base directory.
*/
private static String[] glob(File dir, String includes) throws IOException {
if(isAbsolute(includes))
throw new IOException("Expecting Ant GLOB pattern, but saw '"+includes+"'. See http://ant.apache.org/manual/CoreTypes/fileset.html for syntax");
FileSet fs = Util.createFileSet(dir,includes);
DirectoryScanner ds = fs.getDirectoryScanner(new Project());
String[] files = ds.getIncludedFiles();
return files;
}
/**
* Reads this file.
*/
public InputStream read() throws IOException {
if(channel==null)
return new FileInputStream(new File(remote));
final Pipe p = Pipe.createRemoteToLocal();
channel.callAsync(new Callable<Void,IOException>() {
public Void call() throws IOException {
FileInputStream fis=null;
try {
fis = new FileInputStream(new File(remote));
Util.copyStream(fis,p.getOut());
return null;
} finally {
IOUtils.closeQuietly(fis);
IOUtils.closeQuietly(p.getOut());
}
}
});
return p.getIn();
}
/**
* Reads this file into a string, by using the current system encoding.
*/
public String readToString() throws IOException {
InputStream in = read();
try {
return IOUtils.toString(in);
} finally {
in.close();
}
}
/**
* Writes to this file.
* If this file already exists, it will be overwritten.
* If the directory doesn't exist, it will be created.
*/
public OutputStream write() throws IOException, InterruptedException {
if(channel==null) {
File f = new File(remote).getAbsoluteFile();
f.getParentFile().mkdirs();
return new FileOutputStream(f);
}
return channel.call(new Callable<OutputStream,IOException>() {
public OutputStream call() throws IOException {
File f = new File(remote).getAbsoluteFile();
f.getParentFile().mkdirs();
FileOutputStream fos = new FileOutputStream(f);
return new RemoteOutputStream(fos);
}
});
}
/**
* Overwrites this file by placing the given String as the content.
*
* @param encoding
* Null to use the platform default encoding.
* @since 1.105
*/
public void write(final String content, final String encoding) throws IOException, InterruptedException {
act(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
f.getParentFile().mkdirs();
FileOutputStream fos = new FileOutputStream(f);
Writer w = encoding != null ? new OutputStreamWriter(fos, encoding) : new OutputStreamWriter(fos);
try {
w.write(content);
} finally {
w.close();
}
return null;
}
});
}
/**
* Computes the MD5 digest of the file in hex string.
*/
public String digest() throws IOException, InterruptedException {
return act(new FileCallable<String>() {
public String invoke(File f, VirtualChannel channel) throws IOException {
return Util.getDigestOf(new FileInputStream(f));
}
});
}
/**
* Rename this file/directory to the target filepath. This FilePath and the target must
* be on the some host
*/
public void renameTo(final FilePath target) throws IOException, InterruptedException {
if(this.channel != target.channel) {
throw new IOException("renameTo target must be on the same host");
}
act(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
f.renameTo(new File(target.remote));
return null;
}
});
}
/**
* Moves all the contents of this directory into the specified directory, then delete this directory itself.
*
* @since 1.308.
*/
public void moveAllChildrenTo(final FilePath target) throws IOException, InterruptedException {
if(this.channel != target.channel) {
throw new IOException("pullUpTo target must be on the same host");
}
act(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
File t = new File(target.getRemote());
for(File child : f.listFiles()) {
File target = new File(t, child.getName());
if(!child.renameTo(target))
throw new IOException("Failed to rename "+child+" to "+target);
}
f.delete();
return null;
}
});
}
/**
* Copies this file to the specified target.
*/
public void copyTo(FilePath target) throws IOException, InterruptedException {
OutputStream out = target.write();
try {
copyTo(out);
} finally {
out.close();
}
}
/**
* Sends the contents of this file into the given {@link OutputStream}.
*/
public void copyTo(OutputStream os) throws IOException, InterruptedException {
final OutputStream out = new RemoteOutputStream(os);
act(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
FileInputStream fis = null;
try {
fis = new FileInputStream(f);
Util.copyStream(fis,out);
return null;
} finally {
IOUtils.closeQuietly(fis);
IOUtils.closeQuietly(out);
}
}
});
}
/**
* Remoting interface used for {@link FilePath#copyRecursiveTo(String, FilePath)}.
*
* TODO: this might not be the most efficient way to do the copy.
*/
interface RemoteCopier {
/**
* @param fileName
* relative path name to the output file. Path separator must be '/'.
*/
void open(String fileName) throws IOException;
void write(byte[] buf, int len) throws IOException;
void close() throws IOException;
}
public int copyRecursiveTo(String fileMask, FilePath target) throws IOException, InterruptedException {
return copyRecursiveTo(fileMask,null,target);
}
/**
* Copies the files that match the given file mask to the specified target node.
*
* @param fileMask
* Ant GLOB pattern.
* String like "foo/bar/*.xml" Multiple patterns can be separated
* by ',', and whitespace can surround ',' (so that you can write
* "abc, def" and "abc,def" to mean the same thing.
* @param excludes
* Files to be excluded. Can be null.
* @return
* the number of files copied.
*/
public int copyRecursiveTo(final String fileMask, final String excludes, final FilePath target) throws IOException, InterruptedException {
if(this.channel==target.channel) {
// local to local copy.
return act(new FileCallable<Integer>() {
public Integer invoke(File base, VirtualChannel channel) throws IOException {
if(!base.exists()) return 0;
assert target.channel==null;
try {
class CopyImpl extends Copy {
private int copySize;
public CopyImpl() {
setProject(new org.apache.tools.ant.Project());
}
protected void doFileOperations() {
copySize = super.fileCopyMap.size();
super.doFileOperations();
}
public int getNumCopied() {
return copySize;
}
}
CopyImpl copyTask = new CopyImpl();
copyTask.setTodir(new File(target.remote));
copyTask.addFileset(Util.createFileSet(base,fileMask,excludes));
copyTask.setIncludeEmptyDirs(false);
copyTask.execute();
return copyTask.getNumCopied();
} catch (BuildException e) {
throw new IOException2("Failed to copy "+base+"/"+fileMask+" to "+target,e);
}
}
});
} else
if(this.channel==null) {
// local -> remote copy
final Pipe pipe = Pipe.createLocalToRemote();
Future<Void> future = target.actAsync(new FileCallable<Void>() {
public Void invoke(File f, VirtualChannel channel) throws IOException {
try {
readFromTar(remote+'/'+fileMask, f,TarCompression.GZIP.extract(pipe.getIn()));
return null;
} finally {
pipe.getIn().close();
}
}
});
int r = writeToTar(new File(remote),fileMask,excludes,TarCompression.GZIP.compress(pipe.getOut()));
try {
future.get();
} catch (ExecutionException e) {
throw new IOException2(e);
}
return r;
} else {
// remote -> local copy
final Pipe pipe = Pipe.createRemoteToLocal();
Future<Integer> future = actAsync(new FileCallable<Integer>() {
public Integer invoke(File f, VirtualChannel channel) throws IOException {
try {
return writeToTar(f,fileMask,excludes,TarCompression.GZIP.compress(pipe.getOut()));
} finally {
pipe.getOut().close();
}
}
});
try {
readFromTar(remote+'/'+fileMask,new File(target.remote),TarCompression.GZIP.extract(pipe.getIn()));
} catch (IOException e) {// BuildException or IOException
try {
future.get(3,TimeUnit.SECONDS);
throw e; // the remote side completed successfully, so the error must be local
} catch (ExecutionException x) {
// report both errors
throw new IOException2(Functions.printThrowable(e),x);
} catch (TimeoutException _) {
// remote is hanging
throw e;
}
}
try {
return future.get();
} catch (ExecutionException e) {
throw new IOException2(e);
}
}
}
/**
* Writes to a tar stream and stores obtained files to the base dir.
*
* @return
* number of files/directories that are written.
*/
private Integer writeToTar(File baseDir, String fileMask, String excludes, OutputStream out) throws IOException {
FileSet fs = Util.createFileSet(baseDir,fileMask,excludes);
byte[] buf = new byte[8192];
TarOutputStream tar = new TarOutputStream(new BufferedOutputStream(out));
tar.setLongFileMode(TarOutputStream.LONGFILE_GNU);
String[] files;
if(baseDir.exists()) {
DirectoryScanner ds = fs.getDirectoryScanner(new org.apache.tools.ant.Project());
files = ds.getIncludedFiles();
} else {
files = new String[0];
}
for( String f : files) {
if(Functions.isWindows())
f = f.replace('\\','/');
File file = new File(baseDir, f);
TarEntry te = new TarEntry(f);
te.setModTime(file.lastModified());
if(!file.isDirectory())
te.setSize(file.length());
tar.putNextEntry(te);
if (!file.isDirectory()) {
FileInputStream in = new FileInputStream(file);
int len;
while((len=in.read(buf))>=0)
tar.write(buf,0,len);
in.close();
}
tar.closeEntry();
}
tar.close();
return files.length;
}
/**
* Reads from a tar stream and stores obtained files to the base dir.
*/
private static void readFromTar(String name, File baseDir, InputStream in) throws IOException {
TarInputStream t = new TarInputStream(in);
try {
TarEntry te;
while ((te = t.getNextEntry()) != null) {
File f = new File(baseDir,te.getName());
if(te.isDirectory()) {
f.mkdirs();
} else {
File parent = f.getParentFile();
if (parent != null) parent.mkdirs();
OutputStream fos = new FileOutputStream(f);
try {
IOUtils.copy(t,fos);
} finally {
fos.close();
}
f.setLastModified(te.getModTime().getTime());
int mode = te.getMode()&0777;
if(mode!=0 && !Hudson.isWindows()) // be defensive
try {
LIBC.chmod(f.getPath(),mode);
} catch (NoClassDefFoundError e) {
// be defensive. see http://www.nabble.com/-3.0.6--Site-copy-problem%3A-hudson.util.IOException2%3A--java.lang.NoClassDefFoundError%3A-Could-not-initialize-class--hudson.util.jna.GNUCLibrary-td23588879.html
}
}
}
} catch(IOException e) {
throw new IOException2("Failed to extract "+name,e);
} finally {
t.close();
}
}
/**
* Creates a {@link Launcher} for starting processes on the node
* that has this file.
* @since 1.89
*/
public Launcher createLauncher(TaskListener listener) throws IOException, InterruptedException {
if(channel==null)
return new LocalLauncher(listener);
else
return new RemoteLauncher(listener,channel,channel.call(new IsUnix()));
}
private static final class IsUnix implements Callable<Boolean,IOException> {
public Boolean call() throws IOException {
return File.pathSeparatorChar==':';
}
private static final long serialVersionUID = 1L;
}
/**
* Validates the ant file mask (like "foo/bar/*.txt, zot/*.jar")
* against this directory, and try to point out the problem.
*
* <p>
* This is useful in conjunction with {@link FormValidation}.
*
* @return
* null if no error was found. Otherwise returns a human readable error message.
* @since 1.90
* @see #validateFileMask(FilePath, String)
*/
public String validateAntFileMask(final String fileMasks) throws IOException, InterruptedException {
return act(new FileCallable<String>() {
public String invoke(File dir, VirtualChannel channel) throws IOException {
if(fileMasks.startsWith("~"))
return Messages.FilePath_TildaDoesntWork();
StringTokenizer tokens = new StringTokenizer(fileMasks,",");
while(tokens.hasMoreTokens()) {
final String fileMask = tokens.nextToken().trim();
if(hasMatch(dir,fileMask))
continue; // no error on this portion
// in 1.172 we introduced an incompatible change to stop using ' ' as the separator
// so see if we can match by using ' ' as the separator
if(fileMask.contains(" ")) {
boolean matched = true;
for (String token : Util.tokenize(fileMask))
matched &= hasMatch(dir,token);
if(matched)
return Messages.FilePath_validateAntFileMask_whitespaceSeprator();
}
// a common mistake is to assume the wrong base dir, and there are two variations
// to this: (1) the user gave us aa/bb/cc/dd where cc/dd was correct
// and (2) the user gave us cc/dd where aa/bb/cc/dd was correct.
{// check the (1) above first
String f=fileMask;
while(true) {
int idx = findSeparator(f);
if(idx==-1) break;
f=f.substring(idx+1);
if(hasMatch(dir,f))
return Messages.FilePath_validateAntFileMask_doesntMatchAndSuggest(fileMask,f);
}
}
{// check the (1) above next as this is more expensive.
// Try prepending "**/" to see if that results in a match
FileSet fs = Util.createFileSet(dir,"**/"+fileMask);
DirectoryScanner ds = fs.getDirectoryScanner(new Project());
if(ds.getIncludedFilesCount()!=0) {
// try shorter name first so that the suggestion results in least amount of changes
String[] names = ds.getIncludedFiles();
Arrays.sort(names,SHORTER_STRING_FIRST);
for( String f : names) {
// now we want to decompose f to the leading portion that matched "**"
// and the trailing portion that matched the file mask, so that
// we can suggest the user error.
//
// this is not a very efficient/clever way to do it, but it's relatively simple
String prefix="";
while(true) {
int idx = findSeparator(f);
if(idx==-1) break;
prefix+=f.substring(0,idx)+'/';
f=f.substring(idx+1);
if(hasMatch(dir,prefix+fileMask))
return Messages.FilePath_validateAntFileMask_doesntMatchAndSuggest(fileMask, prefix+fileMask);
}
}
}
}
{// finally, see if we can identify any sub portion that's valid. Otherwise bail out
String previous = null;
String pattern = fileMask;
while(true) {
if(hasMatch(dir,pattern)) {
// found a match
if(previous==null)
return String.format("'%s' doesn't match anything, although '%s' exists",
fileMask, pattern );
else
return String.format("'%s' doesn't match anything: '%s' exists but not '%s'",
fileMask, pattern, previous );
}
int idx = findSeparator(pattern);
if(idx<0) {// no more path component left to go back
if(pattern.equals(fileMask))
return String.format("'%s' doesn't match anything", fileMask );
else
return String.format("'%s' doesn't match anything: even '%s' doesn't exist",
fileMask, pattern );
}
// cut off the trailing component and try again
previous = pattern;
pattern = pattern.substring(0,idx);
}
}
}
return null; // no error
}
private boolean hasMatch(File dir, String pattern) {
FileSet fs = Util.createFileSet(dir,pattern);
DirectoryScanner ds = fs.getDirectoryScanner(new Project());
return ds.getIncludedFilesCount()!=0 || ds.getIncludedDirsCount()!=0;
}
/**
* Finds the position of the first path separator.
*/
private int findSeparator(String pattern) {
int idx1 = pattern.indexOf('\\');
int idx2 = pattern.indexOf('/');
if(idx1==-1) return idx2;
if(idx2==-1) return idx1;
return Math.min(idx1,idx2);
}
});
}
/**
* Shortcut for {@link #validateFileMask(String)} in case the left-hand side can be null.
*/
public static FormValidation validateFileMask(FilePath pathOrNull, String value) throws IOException {
if(pathOrNull==null) return FormValidation.ok();
return pathOrNull.validateFileMask(value);
}
/**
* Short for {@code validateFileMask(value,true)}
*/
public FormValidation validateFileMask(String value) throws IOException {
return validateFileMask(value,true);
}
/**
* Checks the GLOB-style file mask. See {@link #validateAntFileMask(String)}
* @since 1.294
*/
public FormValidation validateFileMask(String value, boolean errorIfNotExist) throws IOException {
value = fixEmpty(value);
if(value==null)
return FormValidation.ok();
try {
if(!exists()) // no workspace. can't check
return FormValidation.ok();
String msg = validateAntFileMask(value);
if(errorIfNotExist) return FormValidation.error(msg);
else return FormValidation.warning(msg);
} catch (InterruptedException e) {
return FormValidation.ok();
}
}
/**
* Validates a relative file path from this {@link FilePath}.
*
* @param value
* The relative path being validated.
* @param errorIfNotExist
* If true, report an error if the given relative path doesn't exist. Otherwise it's a warning.
* @param expectingFile
* If true, we expect the relative path to point to a file.
* Otherwise, the relative path is expected to be pointing to a directory.
*/
public FormValidation validateRelativePath(String value, boolean errorIfNotExist, boolean expectingFile) throws IOException {
AbstractProject subject = Stapler.getCurrentRequest().findAncestorObject(AbstractProject.class);
subject.checkPermission(Item.CONFIGURE);
value = fixEmpty(value);
// none entered yet, or something is seriously wrong
if(value==null || (AbstractProject<?,?>)subject ==null) return FormValidation.ok();
// a common mistake is to use wildcard
if(value.contains("*")) return FormValidation.error("Wildcard is not allowed here");
try {
if(!exists()) // no base directory. can't check
return FormValidation.ok();
FilePath path = child(value);
if(path.exists()) {
if (expectingFile) {
if(!path.isDirectory())
return FormValidation.ok();
else
return FormValidation.error(value+" is not a file");
} else {
if(path.isDirectory())
return FormValidation.ok();
else
return FormValidation.error(value+" is not a directory");
}
}
String msg = "No such "+(expectingFile?"file":"directory")+": " + value;
if(errorIfNotExist) return FormValidation.error(msg);
else return FormValidation.warning(msg);
} catch (InterruptedException e) {
return FormValidation.ok();
}
}
/**
* A convenience method over {@link #validateRelativePath(String, boolean, boolean)}.
*/
public FormValidation validateRelativeDirectory(String value, boolean errorIfNotExist) throws IOException {
return validateRelativePath(value,errorIfNotExist,false);
}
public FormValidation validateRelativeDirectory(String value) throws IOException {
return validateRelativeDirectory(value,true);
}
@Deprecated
public String toString() {
// to make writing JSPs easily, return local
return remote;
}
public VirtualChannel getChannel() {
if(channel!=null) return channel;
else return Hudson.MasterComputer.localChannel;
}
/**
* Returns true if this {@link FilePath} represents a remote file.
*/
public boolean isRemote() {
return channel!=null;
}
private void writeObject(ObjectOutputStream oos) throws IOException {
Channel target = Channel.current();
if(channel!=null && channel!=target)
throw new IllegalStateException("Can't send a remote FilePath to a different remote channel");
oos.defaultWriteObject();
oos.writeBoolean(channel==null);
}
private void readObject(ObjectInputStream ois) throws IOException, ClassNotFoundException {
Channel channel = Channel.current();
assert channel!=null;
ois.defaultReadObject();
if(ois.readBoolean()) {
this.channel = channel;
} else {
this.channel = null;
}
}
private static final long serialVersionUID = 1L;
public static int SIDE_BUFFER_SIZE = 1024;
/**
* Adapts {@link FileCallable} to {@link Callable}.
*/
private class FileCallableWrapper<T> implements DelegatingCallable<T,IOException> {
private final FileCallable<T> callable;
public FileCallableWrapper(FileCallable<T> callable) {
this.callable = callable;
}
public T call() throws IOException {
return callable.invoke(new File(remote), Channel.current());
}
public ClassLoader getClassLoader() {
return callable.getClass().getClassLoader();
}
private static final long serialVersionUID = 1L;
}
private static final Comparator<String> SHORTER_STRING_FIRST = new Comparator<String>() {
public int compare(String o1, String o2) {
return o1.length()-o2.length();
}
};
}