jb -- Parser and Lexer Generation for Java

Last Updated: 7 November 2000

Latest Version: jb-7.0

[ Description | Retrieving | Change Log | Mailing List | Acknowledgments ]

Description

The jb system allows Java (tm) parsers to be generated using the Gnu Bison parser generator system. The parser is written using the standard Bison language, but the action parts are written in Java. Jb takes the mixed C-Java file output by Bison and scans it to extract the parse tables, constants, and actions (in Java). Jb then scans various template files specified by the user and inserts the extracted information at specified points in the templates.

In addition to generating parsers, jb provides three methods of generating corresponding lexers.

Flex -- the jb system (starting with version 3.0) can also take lexers generated using the Gnu flex generator and translate them to execute in Java; this is accomplished with a program called jf that operates in a manner analogous to jb.
yylex.generic -- A generic ad-hoc lexer that can be modified to produce lexers for typical programming languages.
yytokenizer.generic -- A wrapper for the Java StreamTokenizer Class.

Retrieving jb

SERL

Various versions of jb can be obtained from the Software Engineering Research Laboratory (SERL) software download web page: http://www.cs.colorado.edu/serl/software

FTP

Various versions of jb are available at the University of Colorado anonymous ftp site (ftp://ftp.cs.colorado.edu/pub/cs/distribs/serl/). Within that site, the following ftp references are symbolic links to the latest versions of jb.

Source:: ftp://ftp.cs.colorado.edu/pub/cs/distribs/serl/jb.tar
Information file:: ftp://ftp.cs.colorado.edu/pub/cs/distribs/serl/jb.txt

Other Locations

The jb system has been listed at the following web sites.

Dependencies:

Bison -- Jb has been tested with bison version 1.24 and 1.25. Other versions probably will work as well as long as the bison output parser is not wildly changed, but jb will issue a warning when using other versions. Bison can be obtained through the Gnu web page (http://www.gnu.org).
Flex -- Jb has been tested with flex version 2.5.2 and 2.5.3. Flex can be obtained through the Gnu web page (http://www.gnu.org).
Java -- As of Jb version 6, jb produces code targeted only at Java version 1.1.5 or later.
Starwave Regular Expression Package -- I use this package (version 1.10) to provide regular expression support. This package was produced by Jonathan Payne (jpayne@starwave.com). and is included in the jb distribution.

Mailing List

If you are interested in receiving occasional mailings about this system, please send your preferred email address to the contact address below and mentioning the name of this system.

Acknowledgments

This work is sponsored by the Defense Advanced Research Projects Agency and the Air Force Material Command, Rome Laboratory, under Contract Number F30602-00-2-0608.

URL for this web page: http://www.cs.colorado.edu/serl/misc/jb.html

Author: Dennis Heimbigner <dennis@cs.colorado.edu>

SERL: jb -- Bison Parser and Lexer Generating Java

Change Log

Note that this list of changes is a partial summary taken from the more complete list in the changelog file.

[ Version 7 | Version 6]

For changes before version 6, refer to the changelog file.

Changes Incorporated into Version 7

Minor version levels are indicated in parentheses.

(0) Start Date: 1 July 2000
(0) Make jbf store its classes into a single jar file that includes the regexp class files as well.
(0) Modified jexec command to take arguments that look like a 1.2 java command so that when using JDK 1.2 (or later), jexec need not be used.
(0) Modified jb and jb to accept a "--substitution" (or "--subst") flag that allows the substitutions to be placed into a separate file instead of on the command line.
(0) Completely revamped the handling of the parser-lexer communications mechanism to be closer to the way Bison does it. This involves two things:
1. YYlex.yylex(), the method called by the parser to get a token, now returns an object of type YYtoken. YYtoken contains all the normal information needed by the parser from the lexer for the current lexeme.
2. Decided that a separate YYtokentypes class was unnecessary and made use of multiple parsers more difficult, so it contents were folded into the parser and lexer classes.
(0) Created a separate doc directory, and broke README.html into several parts.
(0) Moved jexec into a new bin directory.
(0) Made a number of procedural changes to the examples files and to the Makefile.
(0) Added two new examples to test out instantiation of multiple parsers within the same program.

Changes Incorporated into Version 6

Minor version levels are indicated in parentheses.

(0) Removed all pretense of being able to work with JDK 1.0.x; now only JDK 1.1.5 or later is supported.
(0) Added yytokenizer.template to wrap java.io.StreamTokenizer as a third mechanism for producing lexers; also produced a third calculator version (calcy) that uses yytokenizer.
(0) Modified calc.y to use simpler typing in which instances of Double are placed on the value stack rather than YYterminal.
(0) Modified yylex.template, yylex.generic, yyparse.template, and Main to use java.io.PrintWriter instead of java.io.PrintStream, and to use java.io.Reader instead of java.io.InputStream.
(0) Modified the interface between YYlex and YYparse to provide correct access to the location of the token most recently returned from the lexer.
(0) Replaced the java.y and java.lex files with better ones provided (with permission) by Dmitri Bronnikov.
(0) Dropped the example version of the java parser that uses the generic lexer, and added an example version of the IDL parser that uses the generic lexer.
(0) Modified the Makefiles to install the class files into a separate classes directory. Also modified the top level Makefile to construct a Java jar file.
(1) Made jb compatible with JDK 2.0 (was 1.2) by recompiling everything and fixing any encountered errors.

SERL <serl@cs.colorado.edu>

The jb system takes parsers generated using the Gnu Bison parsergenerator system and translates them to execute in Java. The C file output by Bison is scanned to extract the parse tables andconstants. Jb then scans various template Java files specified by the user and inserts the extracted information at specified points in the templates.