Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 13 Next »

Here are some tips on how to debug what is happening on the transition from groovy source code into the generated bytecode.

artifact | transformation | artifact
source (Hello.groovy) | - GroovyLexer -> | antlr tokens
antlr tokens | - GroovyRecognizer -> | antlr ast
antlr ast | - AntlrParserPlugin -> | groovy ast
groovy ast | - ??? -> | bytecode (Hello.class)

Note1 groovy.g is used to generate GroovyLexer and GroovyRecognizer

Note2 GroovyRecognizer is sometimes easier to understand in its syntax diagram form

Note3 AntlrParserPlugin source available.

Example

For these examples let's assume the file Hello.groovy contains

GroovyLexer (viewing Antlr Tokens)

To view the antlr tokens that the source code has been broken into you need to do the following in groovy-core subdirectory

GroovyRecognizer (viewing Antlr AST)

To view the antlr AST that the recognized tokens have built

AntlrParserPlugin (viewing Groovy AST)

To view the Groovy AST that is one step closer to the generated bytecode you can generate Hello.groovy.xml using these unix commands.

This can be generated using both classic and jsr parsers, by changing the groovy.jsr system property. By doing this we can diff the generated Groovy AST artifacts for debugging and migration purposes.

jsr parser
classic parser

Viewing Bytecode

One interesting bytecode viewer is jclasslib which renders Hello.class in this manner...

Decompiling bytecode back to Java

If, however, you are bamboozled by bytecode... The easiest to grok mechanism for reading the compiled code is to use a tool like JAD to decompile the Hello.class into a readable Hello.java

  • No labels