I'm experiencing some rather annoying problems with scala. The problem is, that I can compile small scala project perfectly, but when the projects are bigger, the compiler crashes with an StackOverflowException.
Clearly, I have to increase the stack size for the compiler, however, that's probably my main problem here, I don't know how.
I'm starting netbeans with these parameters:
netbeans_default_options="-J-client -J-Xmx512m -J-Xss8m -J-Xms512m -J-XX:PermSize=128m -J-XX:MaxPermSize=512m -J-Dapple.laf.useScreenMenuBar=true -J-Dapple.awt.graphics.UseQuartz=true -J-Dsun.java2d.noddraw=true"
So, as far as I'm aware, -J-Xss8m should increase the thread stack size to 8 mb. However, that doesn't seem to affect the compiler. So I tried to pass the same parameter to the compiler directly, using the compiler flags, which I can set in netbeans, resulting in this:
-deprecation -J-Xss8m
But again, that doesn't help, I'm still getting the exception. I searched through the netbeans documentation, but all I found was the netbeans startup parameters, which I had already set. I hope somebody here can give me further information on how to handle this problem.
Further information:
So, after a day I finally had the chance to try everything out on a different machine. I used the same settings and same compiler, but to my surprise, I didn't get the same result. Meaning, on his machine the compiler compiles the whole code without any exception.
The only difference between mine computer and his is, that his has more RAM and CPU power, but that shouldn't make the deal since we both use netbeans with the same startup options.
By now, I even tried out the RC of the 2.9 scala compiler, it didn't help much. Also, I checked if I have the correct scala plugin installed, since there might be problems when using the 2.8 plugin with the 2.9 compiler and vice versa. However, I'm using the 2.9 plugin and 2.9 compiler, so that's fine.
The problem of giving the Scala compiler more stack space is similar to specifying more heap space. Both of these options must be specified as custom JVM arguments when running the Scala compiler. However Netbeans lacks any sort of documentation on how to do it, so here it is.
The way to specify custom JVM arguments for the Scala compiler with Netbeans is by customizing build.xml for each project.
Open nbproject/build-impl.xml in the project's folder.
Search for "scalac" and you will find the following target: -init-macrodef-scalac.
Copy the whole target definition, paste it into your build.xml, and save it.
Close nbproject/build-impl.xml, from now on you will work with build.xml.
In the target you just copied, locate the <scalac> tag, the nesting will be as follows: target.macrodef.sequential.scalac
Add a custom "jvmargs" attribute to the scalac tag, it will look as follows: <scalac jvmargs="-Xss2048k -Xmx2000m" ... >
Save the build.xml. Now whenever you compile your project with netbeans, the compiler will be run with the custom jvm arguments.
Related
I use Delphi 7 and I encountered a very annoying thing. One of my projects has gone quite large and a form I used in one of the units became obsolete. I decided to get rid of it. It seems I can't. The IDE always asks for .dfm file, no matter what i tried (and I think i was very thorough).
Here's what I did:
First, I used the IDE's Remove from Project to remove the whole thing from the project. Then I commented out all sections that used anything in that unit and compiled, bulit and ran the project. It went without any errors or warnings. I saved all files and closed the Delphi IDE, started up a Notepad and removed the form's declaration from the unit, then deleted every associated file, except the .pas (with this, it became like any other regular unit).
Then I searched every file in the project's folder with Total Commander to see if any of them contained the name of the removed unit. None did.
I started Delphi and loaded the project.
With this I felt satisfied and added the unit's name to the uses list and pressed Compile.
IT STILL LOOKS FOR THE DAMNED DFM!!
If someone knows, please explain this to me. What did I miss?
If the compiler is looking for the .dfm file it will be because the compiler is including a unit that contains this:
{$R *.dfm}
That is what tells the compiler to link the .dfm file with the same name as the unit.
It seems likely that you still refer to the unit somewhere in your project. The compiler must be finding either a .pas file or a compiled .dcu file for your unit. Search for all files named <UnitName>.pas and <UnitName>.dcu. Remove these files. Then you will encounter compiler errors whenever you have code that attempts to use those units. Deal with those by removing the unit from the uses clause that names it. Then you should be done.
Compiling Spark gives this compile error :
To fix I modify Utils.classIsLoadable method to just return true:
def classIsLoadable(clazz: String): Boolean = {
// Try { Class.forName(clazz, false, getContextOrSparkClassLoader) }.isSuccess
true
}
I realise this is not a good fix, but so far Spark seems to be running correctly from source. Has this compile error been experienced before and is there a fix? Will returning true suffice for now , I'm not sure what impact modifying this return value may have?
I suggest compiling Spark from the command-line using Maven or SBT instead of trying to use your IDE's compiler. Many of the core Spark developers use IntelliJ for editing Spark's source code but still use the command-line compilers, largely because it's been difficult to get the project to build correctly inside IDEs. Even if you're using an external compiler, you should still be able to benefit from IntelliJ's syntax highlighting, type checking, etc.
Here's a relevant discussion from the Spark developer mailing list: http://apache-spark-developers-list.1001551.n3.nabble.com/IntelliJ-IDEA-cannot-compile-TreeNode-scala-td7090.html
Note that Spark users should be able to use IntelliJ to compile applications that depend on Spark; this issue only affects developers who want to build Spark itself.
If you're interested in fixing the build to work with IntelliJ, I recommend opening a ticket on the Spark issue tracker.
I am developing an eclipse plugin that runs code violation checker on the difference of two versions of a file. Right now I am using diff.exe to get the difference between the two files. But as diff.exe is an extrenal app, I realized that its better to use eclipse built-in compare tool to get the file difference.
So I used org.eclipse.compare and reached up to this point:
public static List<Patch> compare(String old, String recent) {
try{
IRangeComparator left = new TokenComparator(old); //what exactly to be passed in this constructor, a file path, a literal value or something else?
IRangeComparator right = new TokenComparator(recent);
RangeDifference[] diffs = RangeDifferencer.findDifferences(left, right); // This line is throwing NPE
//..
// Process RangeDifferences into Collection of Patch collection
//..
}catch(Exception e){}
//Returns a collection of file differences.
return null;
}
Now the problem is I am not sure what exactly to be passed in the constructor TokenComparator(String). The document says this constructor Creates a TokenComparator for the given string. But it is not written what exactly to be passed in this constructor, a file path, a literal value or something else? When I'm passing a file path or a string literal I am getting NullPointerException on the next line of finding differences.
java.lang.NullPointerException
at org.eclipse.compare.internal.core.LCS.isCappingDisabled(LCS.java:98)
at org.eclipse.compare.internal.core.LCS.longestCommonSubsequence(LCS.java:55)
at org.eclipse.compare.rangedifferencer.RangeComparatorLCS.longestCommonSubsequence(RangeComparatorLCS.java:186)
at org.eclipse.compare.rangedifferencer.RangeComparatorLCS.findDifferences(RangeComparatorLCS.java:31)
at org.eclipse.compare.rangedifferencer.RangeDifferencer.findDifferences(RangeDifferencer.java:98)
at org.eclipse.compare.rangedifferencer.RangeDifferencer.findDifferences(RangeDifferencer.java:82)
at org.eclipse.compare.rangedifferencer.RangeDifferencer.findDifferences(RangeDifferencer.java:67)
at com.dassault_systemes.eclipseplugin.codemonview.util.CodeMonDiff.compare(CodeMonDiff.java:48)
at com.dassault_systemes.eclipseplugin.codemonview.util.CodeMonDiff.main(CodeMonDiff.java:56)
Someone please tell what is right way to proceed.
If the question is What value the token comparators constructor takes then the answer is it takes the input string to compare. Specified in javadoc here http://help.eclipse.org/indigo/index.jsp?topic=%2Forg.eclipse.platform.doc.isv%2Freference%2Fapi%2Forg%2Feclipse%2Fcompare%2Fcontentmergeviewer%2FTokenComparator.html
TokenComparator(String text)
Creates a TokenComparator for the given string.
And the null pointer yo are getting is because in function isCappingDisabled it tries to open the compare plugin which seems to be null. You seem to be missing a direct dependency to the plugin "org.eclipse.compare.core"
The org.eclipse.compare plugin was never meant to be used in standalone : many of its functionalities require a running instance of Eclipse. Furthermore, it mixes core and UI code within the same plugin, which will lead to unexpected behavior if you are not very careful about what you use and what dependencies are actually available in your environment.
You mentionned that you were developping an Eclipse plugin. However, the NPE you get indicates that you are not running your code as an Eclipse plugin, but rather as a standard Java program. In an Eclipse environment, ComparePlugin.getDefault() cannot return null : the plugin needs to be started for that call to return anything but null.... and the mere loading of the ComparePlugin class within Eclipse is enough to start it.
The answer will be a choice :
You need your code to run as a standalone Java program out of Eclipse. In such an event, you cannot use org.eclipse.compare and diff.exe is probably your best choice (or you could switch to an implementation of diff that was implemented in Java in order to be independent of the platform).
You do not need your program to work in a standalone environment, only as an Eclipse plugin. In this case, you can keep the code you're using. However, when you run your code, you have to launch it as a new "Eclipse application" instead of "Java Application". You might want to look at a tutorial on how to develop Eclipse plugins for this, This simple tutorial from Lars Vogel shows how to run a new Eclipse Application to test an Hello World plugin. You will need a similar code, with a menu entry to launch your plugin somewhere (right-click on a file then select "check violations" in your case?).
I'm using GWT in my stuff, and I would like to make it,
international, so I use GWT constants method.
I have a java file with defaults, and I now need to make properties files.
In a remember, there is a special thing to do (or done automagically) to generate
a kind of template where all constants are generated with empty labels for other langages.
Did I dream this ?
(using eclipse indigo to develop webapp with gwt but not gae)
[edit:]
this was not a dream, it's i18ncreator:
http://code.google.com/intl/fr-FR/webtoolkit/doc/latest/RefCommandLineTools.html#i18nCreator
but I can't make it working on windows :-(
[edit again ]
due to this issue : http://code.google.com/p/google-web-toolkit/issues/detail?id=5113
recommended solution is use i18ncreator in gwt 1.7 (!)
you should see the page on locales in GWT
I had the same issue. I was looking all over the place for the answer but could not find an answer; either in the docs or on stackoverflow.
So I asked in the GWT gitter channel and was told to use the compiler argument
-extra <destination-folder-name>
to generate the .properties files from the Interface files.
Steps in eclipse:
Select project you want to compile
[right click] -> Google -> GWT Compile
In the window that opens, open the Advanded options.
Add the following additional compiler argument -extra <destination-folder-name>
Compile
This should generate the *.properties files in the /destination-folder-name.
NOTE: This only generates the .properties files. It does not actually compile the application with all the locales for deploy.
Move the MyInterfaceExtension_*.properties to be right beside the MyInterfaceExtension.java file.
Make copies for each locale i.e. MyInterfaceExtension_fr_CA.properties, MyInterfaceExtension_fr_FR.properties, etc..
Translate them
Then run the compilation process again with out the -extra <destination-folder-name> option. Because it is not needed anymore.
This will compile with all the locales you enabled. You can now deploy the app the usual way.
Quick Tips:
When compiling for the first time in order to generate the .properties file, I commented out the locales in the module definition file so that the compiler will not sit there and compile again and again for every browser and every locale
i.e. supported_browser_count x enabled_locale_count = 5 browsers x 3 locales = 15 compilation Permutations, which is going to increase your compilation time.
Because, all I needed was that one *_en.properties file.
For the second compilation, after you copied and translated the properties files for each locale, you have to enable all the locales you want to support and compile.
Credits:
github #niloc132 : Colin Alworth
github #ibaca : Ignacio Baca Moreno-Torres
For helping me with this.
For my project, I used the i18n-Creator
http://code.google.com/webtoolkit/doc/latest/DevGuideI18n.html#DevGuidePropertiesFiles
It kind of does the opposite of what you are asking for. With the i18n-creator, you create the properties files for the various locales and run the script that is generated with the i18n-creator, and it will generate the constants interface.
I haven't heard yet of this feature in Eclipse but IntelliJ IDEA has this feature, you just create the Constants Interface class and the properties file. If you add a method in the class file it will warn you to add the property or the other way around. HTH.
GWT compiles the Java source into Javascript, and names the files according to a hash of their contents. I'm getting a new set of files every compile, because the javascript contents are changing, even when I don't change the source at all.
The files are different for OBF and PRETTY output, but if I set it to DETAILED, they're no longer different every compile. In PRETTY, I can see that all/most of the differences between compiles are in the value parameters for typeId. For example, a funciton called initValues() is called with different values for it's typeId parameter.
In PRETTY mode, the differences you see are allocation of Java Classes to TypeIds. It's how GWT manages run time type checking. You'll notice a table at the bottom of each script essentially mapping each typeId to all compatible superclasses. This is how GWT can still throw ClassCastException in JavaScript (though you should run into this very rarely!).
In OBF mode, the differences are due to the allocation of minified function names.
In both cases, it's due to the order the compiler is processing the code. Some internal symbol tables might be using a non-ordered collection store symbols for processing. It can happen for lots of reasons.
As far as I know, GWT will compile a new version every time you compile it, this is a feature ;)
You can use ant to control it though, so that it only builds the GWT section of your application if it's actually changed:
http://wiki.shiftyjelly.com/index.php/GWT#Use_The_Power_of_Ant_to_Build_Changes_Only