Version 3.5.5

issues resolved

  • SourceRoot is now silent by default - look at the Log class if you want to change that.
  • BETA: the below work on Java Symbol Solver is still ongoing.

Version 3.5.4

issues resolved

  • BETA: the below work on Java Symbol Solver is still ongoing.

Version 3.5.3

issues resolved

  • Unicode escapes (\u1234) are now retained in the AST, but they are now only allowed in comments, string and character literals, and identifiers.
  • BETA: the below work on Java Symbol Solver is still ongoing.

Version 3.5.2

issues resolved

  • The pretty printer now cleans up Javadoc comments.
  • BETA: the below work on Java Symbol Solver is still ongoing.

Version 3.5.1

issues resolved

  • BETA: the below work on Java Symbol Solver is still ongoing.

Version 3.5.0

issues resolved

  • A functional visitor API has been added. See PR 1195 for now.
  • Build is working again on Windows thanks to Leonardo Herrera.
  • The pretty printer now has an option to order imports, also thanks to Leonardo Herrera.
  • Receiver parameters are now well-supported instead of being a hack. See issue 1194 for a description.
  • BETA: the below work on Java Symbol Solver is still ongoing.

Version 3.4.4

issues resolved

  • BETA: the below work on Java Symbol Solver is still ongoing.

Version 3.4.3

issues resolved

  • BETA: we're still doing work to integrate parts of Java Symbol Solver to simplify its API.
  • VisitorMap is joined by VisitorSet and VisitorList, for when you want to store Nodes in collection but don't want its default equals/hascode behaviour

Version 3.4.2

issues resolved

Version 3.4.1

issues resolved

  • Two visitors were added: NoCommentEqualsVisitor and NoCommentHashCodeVisitor - as the name implies you can use these to compare nodes without considering comments. Thanks Ryan Beckett!
  • isSomeType() methods have been added for many types that help avoid instanceof.
  • asSomeType() methods have been added for many types that help avoid casting to that type.
  • ifSomeType() methods have been added for many types, giving a nice functional way of doing if-is-type-then-cast-to-type-then-use.
  • The LexicalPreservingPrinter had its API changed a little: setup and printing are now separate things, so you don't have to drag an instance of LexicalPreservingPrinter through your code anymore.
  • traverseScope was added to all nodes with a scope, so you can travel through the scope without tripping over (non-)optionality.

Version 3.4.0

issues resolved

  • BREAKING: We missed a Java 9 feature which is now on board: try with resources can now refer to a resource declared outside of the statement. This means that the node type you get for those resources is now Expression instead of VariableDeclarationExpr. For Java 8 and below you can simply cast it to VariableDeclarationExpr again. See also the Javadoc for TryStmt.
  • You can now inspect the AST by exporting it to XML, JSON, YAML, or a Graphviz's dot diagram, thanks to Ryan Beckett!
  • GenericVisitorWithDefaults and VoidVisitorWithDefaults were added which function like empty visitors, but all the visit methods call a default method by default.
  • Annotation support was cleaned up, adding some obscure locations where you can have annotations.
  • EnumDeclaration regained its constructor builder methods. They were accidentally lost around 3.2.2.
  • ArrayType now has an origin field which indicates in which position the array brackets were found.

Version 3.3.5

issues resolved

Version 3.3.4

issues resolved

  • SourceZip has been added. Use it to read source code from jars or zip files. Thank you @ryan-beckett !
  • JavaCC was upgraded to 7.0.2
  • A new option for the pretty printer was added. You can now wrap-and-column-align parameters of method calls. Thank you @tarilabs !

Version 3.3.3

issues resolved

  • Parsing a partial java file (like an expression or statement) no longer ignores trailing code.
  • New memory saving option: turn off token list.

Version 3.3.2

issues resolved

  • VisitorMap lets you override hashcode/equals for nodes when used as a key for a map.

Version 3.3.1

issues resolved

  • The token list is now mutable - see methods on JavaToken. This caused mild breakage - some fields have become Optional.

Version 3.3.0

issues resolved

  • Breaking: TryStmt::tryBlock and EnclosedExpr::inner were optional for no good reason. Now they are required.
  • You can now ask a JavaToken for its category, which is useful for examining the token list or doing syntax highlighting or so.
  • enum and strictfp can now be used as identifiers on lower Java versions.

Version 3.2.12

issues resolved

Version 3.2.11

issues resolved

  • We're up to date with the latest Java 9 module system again.

Version 3.2.10

issues resolved

  • Node.replace(old, new) was added, including property-specific X.replaceY(newY) methods

Version 3.2.9

Scrapped due to release problem.

Version 3.2.8

issues resolved

  • Added isInnerClass() that checks if a ClassOrInterfaceDeclaration is an inner class (note: this is different from a nested class!)
  • @ryan-beckett contributed a huge Eclipse setup guide

Version 3.2.7

issues resolved

Version 3.2.6

issues resolved

  • EmptyMemberDeclaration is gone! It was deprecated for a while because it it was in the AST, but doesn't have any meaning in a Java program. EmptyStmt was also deprecated, but that has been reverted. This node does have meaning.

Version 3.2.5

issues resolved

  • NodeWithCondition was added on all nodes containing a condition.
  • Lots of work on improving lexical preservation.
  • If a file was parsed from a file system, you can now get path information etc. from CompilationUnit
  • API BREAKING: every node now points to its start and end token. Some of the API has started returning TokenRange instead of Range - you can call toRange to get the old object type. We may still change the naming of some of this code in the following month.

Version 3.2.4

New style changelog, no more issue numbers, but a link: issues resolved and any notable changes:

  • the new method Node.removeForced() by removing it, or removing the first parent that is optional. This is different from Node.remove(), remove() only tries to remove the node from the parent and fails if it can't.
  • FieldAccessExpr.scope is now a required property. You might find some get()s in your code that are no longer necessary.
  • ReferenceType no longer has a type parameter, so every ReferenceType<?> can be replaced by ReferenceType now.

Version 3.2.3

  • 907 906 905 903 911 910 909 908 smaller improvements and fixes

Version 3.2.2

Beta: TreeStructureVisitor.

  • 770 902 904 901 smaller improvements and fixes

Version 3.2.1

Beta: TreeStructureVisitor.

  • Maven dependencies were updated to their latest versions
  • 890 the concept of "method signature" now exists in JavaParser
  • 896 891 889 887 882 789 smaller improvements and fixes

Version 3.2.0

The lexical preservation code is stable!

Beta: TreeStructureVisitor.

  • 885 884 879 878 smaller improvements and fixes

Version 3.1.4

This is the first version to parse Java 9.

Beta: TreeStructureVisitor, and LexicalPreservingPrinter.

  • 872 873 874 787 Remaining Java 9 work.

Version 3.1.3

Beta: TreeStructureVisitor, and LexicalPreservingPrinter.

A start has been made on source level support. The default level is Java 8. It can be set to Java 9 like this for a parser instance:

private final JavaParser parser = new JavaParser(new ParserConfiguration().setValidator(new Java9Validator()));

and like this for the static parse methods:

JavaParser.getStaticConfiguration().setValidator(new Java9Validator());
  • 862 552 "_" is an illegal identifier on source level 9.
  • 869 867 109 855 857 smaller improvements and fixes

Version 3.1.2

Beta: TreeStructureVisitor, ConcreteSyntaxModel, and LexicalPreservingPrinter.

  • 594 849 831 a validation framework was introduced to inform about problems in the AST without needing to change the grammar, and without requiring parsing code. It is open for extension by users.
  • 852 853 826 832 846 839 smaller improvements and fixes

Version 3.1.1

Beta: TreeStructureVisitor, ConcreteSyntaxModel, and LexicalPreservingPrinter.

  • 654 124 lexical preservation (printing source code with the same formatting it had when parsing) has been added. Thank you @ftomassetti for a lot of work!
  • 554 800 first (big) step towards Java 9 support: JavaParser can read project Jigsaw module definitions.
  • 795 786 751 extend the TreeVisitor with more traversal options. Thanks @ryan-beckett!
  • 791 GenericListVisitorAdapter has been added which collects its results in a list. Thanks @Gottox!
  • 815 798 797 813 clean up Problem text
  • 819 818 817 816 441 809 808 807 fix various absurd annotation related issues.
  • 777 805 802 796 790 792 793 781 784 785 783 782 779 357 799 763 smaller improvements and fixes

Version 3.1.0

Beta: TreeStructureVisitor and ConcreteSyntaxModel.

  • 705 755 Add the concrete syntax model, which will give you information about the exact syntax a certain nodes matches.

  • 777 smaller improvements and fixes

Version 3.1.0-beta.2

This version is a beta because TreeStructureVisitor is not in its definite state yet.

  • 762 761 772 merge javaparser-metamodel and javaparser-generator-utils into javaparser-core.
  • 766 the ModifierVisitor is now created by a code generator. Its behaviour has been made logical, and may give different results than before.
  • 755 ConstructorDeclaration and MethodDeclaration now share a parent: CallableDeclaration
  • 687 759 773 769 768 767 765 759 smaller improvements and fixes

Version 3.1.0-beta.1

This version is a beta because there are a lot of new features that may still change.

This version needs a minor version increase because of a backwards compatability issue:

  • 719 getJavadoc, getJavadocComment and getComment could return null. Our promise was to return Optional, so that is what they do now.

New:

  • 658 718 736 737 we have created a metamodel. It gives information about the structure of the various AST nodes, as if you are introspecting them. You can find it in javaparser-metamodel, the main class is JavaParserMetaModel
  • 353 365 visitors are no longer hand made, they are now generated from the metamodel. This should make them 100% reliable. Affected visitors are: GenericVisitorAdapter, EqualsVisitor, VoidVisitorAdapter, VoidVisitor, GenericVisitor, HashCodeVisitor, CloneVisitor.

If you want to generate your own visitors, you can use the VisitorGenerator class from javaparser-core-generators

If you want to reuse the code generation utilities, look at module javaparser-generator-utils - there is a very useful SourceRoot class in there that takes away a lot of file management troubles.

  • 538 735 TreeStructureVisitor has been added, which should be considered beta.
  • 220 733 717 749 745 750 743 748 666 732 746 734 733 smaller improvements and fixes

Version 3.0.1

  • 699 433 325 Javadoc can now be parsed
  • 703 696 added NodeWithOptionalScope
  • 702 FieldAccessExpr now implements NodeWithSimpleName, which means that "field" has been renamed to "name"
  • 707 706 improve range of array types and levels
  • 709 smaller improvements and fixes

Version 3.0.0

  • 695 697 689 680 693 691 682 690 677 679 688 684 683 smaller improvements and fixes

Version 3.0.0-RC.4

  • 668 669 TypeDeclarationStmt became LocalClassDeclarationStmt
  • 347 665 every node now has some documentation
  • 660 670 673 four types of import declaration have been merged back into the old ImportDeclaration
  • 659 The pretty printer can now take customized visitors
  • 650 671 672 674 524 smaller improvements and fixes

Version 3.0.0-RC.3

  • 639 622 632 657 656 652 653 647 648 645 194 643 630 624 628 627 626 625 623 cleanups, small fixes, and general housekeeping

Version 3.0.0-RC.2

  • 593 EmptyImportDeclaration and NonEmptyImportDeclaration have been removed
  • 612 VariableDeclaratorId has been removed. It has been substituted by "SimpleName name"
  • 614 617 the list of tokens has been linearized and simplified
  • 615 support for arrays has once more been changed. See the issue
  • 580 453 380 618 580 611 610 424 608 smaller improvements and fixes

Version 3.0.0-RC.1

  • 499 601 renames many fields to be more consistent
  • 596 605 602 604 smaller improvements and fixes

Version 3.0.0-alpha.11

  • 547 595 Node.range is now using Optional instead of Range.UNKNOWN
  • 584 588 548 585 bug fixes and improvements

Version 3.0.0-alpha.10

  • 578 579 577 575 290 570 568 567 562 564 551 bug fixes and improvements

Version 3.0.0-alpha.9

  • 403 358 549 Make all names nodes: either SimpleName or Name. This makes every name in the AST visitable. NameExpr is now a wrapper to use SimpleName in an expression.
  • 516 536 use Optional<> for return values.
  • 556 557 558 550 small improvements and fixes.
  • 560 559 make nodes observable.

Version 3.0.0-alpha.8

  • 344 529 turn DumpVisitor into an official PrettyPrinter
  • 532 508 427 530 531 513 528 cleanups

Version 3.0.0-alpha.7

  • 515 roll back attempt at using Optional
  • 522 504 make NodeList not a Node (restores parent/children behaviour to before alpha.4)
  • 527 526 rename getChildrenNodes to getChildNodes
  • 525 495 520 bug fix

Version 3.0.0-alpha.6

  • 503 modified ImportDeclaration hierarchy to have getters for static and "asterisk" again
  • 506 bug fix

Version 3.0.0-alpha.5

Version 3.0.0-alpha.4

  • 463 471 nodes can now be removed easily
  • 491 import handling changed. Instead of "ImportDeclaration", we now have the four types of import as described in the JLS. See issue
  • 452 355 474 494 various improvements
  • 493 492 485 Simplify the JavaParser interface

Version 3.0.0-alpha.3

  • 112 237 466 465 461 460 458 457 fundamentally changes how we deal with arrays. It is explained in the last post here
  • 472 456 makes the "data" field on every node more structured.
  • 477 468 refactor TypeArguments. You will find that TypeArguments is no longer a type, it is just a list in some nodes.
  • 482 adds the "nodeTypes" packages to the osgi export.
  • 479 476 makes all setters on nodes return this so they become chainable.
  • 473 437 clean up CloneVisitor.

Version 3.0.0-alpha.2

Version 3.0.0-alpha.1

  • 400 405 introduce many "builder" style methods for constructing code. Thanks DeepSnowNeeL!
  • 409 remove ASTHelper (methods are now on specific Node subclasses)
  • 414 JavaParser can now be instantiated and reused. InstanceJavaParser removed
  • 418 417 411 408 bugs
  • 367 420 407 402 various cleanups

Version 2.5.1

  • 394 OSGi manifest added
  • 391 fix ModifierVisitor NullPointerException bug
  • 385 a few new parse methods
  • 386 fix dumping an empty import (a single ; after the package declaration)

Version 2.5.0

API breaking changes:

  • 191: moved TreeVisitor to visitor package

  • 368, 162, 303, 302, 360: use correct type in some places.

  • 371: code is now compiled with Java 8

  • 342, 331: ModifierVisitorAdapter detects and removes broken nodes

  • 328, 270: upgrade JavaCC (use TokenMgrException now) Other changes:

  • 297: enable access to tokens.

  • 341, 361: node positions are now OO

  • 211, 373: escaping of \n \r for string literals

  • 336, 276, 141: JavaDoc support now works

  • 337, 281: reorganize the stream reading code

  • 343, 309, 332, 57: take advantage of common interfaces

  • 329, 326, 327: deal with platform issues

  • 163, 236, 252, 296, 269, 339, 321, 322, 252, 253, 293, 295: various fixes

  • 310, 311, 313, 301, 294: some code clean-ups

Version 2.4.0

  • several fixes in DumpVisitor for bugs due to lazy initialization
  • make TypeDeclaration implements DocumentableNode directly
  • TypedNode interface introduced
  • introduce MultiBoundType
  • refactored IntersectionType and UnionType
  • refactored CatchClause
  • parsing annotations in throws declarations
  • parse orphan semicolons in import statements
  • added PackageDeclaration.getPackageName
  • solved issue with newlines in string literals
  • fixed handling of UnknownType in EqualsVisitor
  • improvements to CommentsParser
  • removing old grammar

Version 2.3.0

  • ClassOrInterfaceType implements NamedNode
  • DumpVisitor can now be extended
  • Improved documentation
  • AST: lists are now lazy initialized

Version 2.1.0

Code changes

Version 2.0.0

  • support Java 8

Version 1.0.8 (2010-01-17)

  • Fixed issues:
    • Issue 17: A refactor suggestion for AnnotationExpr and its subclasses
    • Issue 21: Java 5 JavaParser compiled JARs
    • Issue 22: Please use java.lang.reflect.Modifier constants in japa.parser.ast.body.ModifierSet
    • Issue 27: Implement the "equal" method
    • Issue 30: equals and hashCode methods

Version 1.0.7 (2009-04-12)

  • Issue 19 fixed:
  • Tests changed to run with junit 4

Version 1.0.6 (2009-01-11)

  • Issue 11 fixed: changed method get/setPakage to get/setPackage in the class CompilationUnit
  • Created new visitor adapter to help AST modification: ModifierVisitorAdapter
  • Changed visitor adapters to abstract

Version 1.0.5 (2008-10-26)

  • Created simplified constructors in the nodes of the AST (without positional arguments)
  • Created ASTHelper class with some helpful methods (more methods are still needed)

Version 1.0.4 (2008-10-07)

  • Moved to javacc 4.1.
  • The java_1_5.jj can be build alone without compilation errors

Version 1.0.3 (2008-09-06)

  • Removed SuperMemberAccessExpr class, it was no longer used
  • Removed the methods get/setTypeArgs() from ArrayCreationExpr, this node shouldn't have these methods.
  • Fixed the bug with start/end position of the nodes IntegerLiteralMinValueExpr and LongLiteralMinValueExpr
  • The methods get/setAnnotations() from all BodyDeclaration subclasses were pushed down to BodyDeclaration class

Version 1.0.2 (2008-07-20)

  • Issue fixed: Issue 1: Add support for editing AST nodes or create new ones

Version 1.0.1 (2008-07-01)

  • Issue fixed: Issue 5: end line and end column equal to begin line and begin column

Version 1.0.0 (2008-06-25)

  • Changed version numbering, starting version 1.0.0
  • Javadoc done for packages:
    • japa.parser
    • japa.parser.ast
  • Corrected bug when parsing in multithread:
    • JavaParser.setCacheParser(false) must be called before to use the parser concurrent

2008-06-19

  • No code changes, added binary distribution to download page

2008-06-11

2008-06-09

  • Added Adapters for de visitors

2008-05-28

2008-05-25

  • Added support for comments and javadoc to the tree.
    • Javadocs are stored directly to members (BodyDeclaration and all deriveds (classes, methods, fields, etc.)), accessible by the method getJavadoc().
    • All comments are stored in the CompilationUnit, accessible by the method getComments().

2008-04-01

  • Changed all nodes public attributes to be private and created getters to access them
  • Changed the methods of the Node getLine e getColumn to getBeginLine and getBeginColumn
  • Added the methods getEndLine and getEndColumn to the Node class (works only in the BlockNode)

2007-12-22

  • Corrected ConditionalExpression bug

2007-10-21

  • Added LGPL License

2007-10-21

  • Bugs corrected:
    • Created PackageDeclaration member of CompilationUnit to add suport for annotations in the package declaration
    • Parameterized anonymous constructor invocation
    • Explicit constructor invotation Type Arguments
    • ctrl+z ("\u001A") ar end of compilation unit

2007-10-09

  • EnumConstantDeclaration annotation support corrected
  • Parssing Java Unicode escape characters suport added

2007-10-03

  • Bug corrected: "MotifComboPopup.this.super()" statement was generating parser error

2007-10-01

  • Bug corrected: Casting signed primitive values
	double d = (double) -1;
	                    ^

2007-08-06

  • Bug with the single line comments in the final of the unit corrected

2007-07-31

  • Fixed the bug with the following expression: Class c = (int.class);

2007-06-26

  • Bug fixes from Leon Poyyayil work
    • suport for hex floating point
    • unicode digits in indentifier
    • MemberValueArrayInitializer

2007-03-09

  • Long and Integer literal MIN_VALUE bug

2007-02-24

  • '\0' bug fixed

2007-02-01

  • Many bug fixes
  • Added line/column to nodes