The release given here has had some micro-optimizations (26 June 1997)
which with luck don't break anything. In particular regexen are
used more intelligently and the generation phase will use
kjbuckets
if it's available.
This is moderately heavy computer science. Not for the timid.
| COPYRIGHT
|
Do what you like, just don't sue me (roughly translated).
|
| DLispShort.py
|
A very simple example language specification.
|
| DumbLispGen.py
|
Another example specification
|
| arefize.py
|
(not related: simple program used to generate this page)
|
| idl.py
|
A very complex example: CORBA IDL parser generator.
This is a good example of a complex grammar. The
interpretation functions and terminal regexes are
all stubbed.
|
| pygram.py
|
A very complex example: The python grammar.
This module uses a hand written lexer to handle
Python's beautiful peculiarities. Used by kypylint.py.
Look at the top of the module for editable parameters.
|
| kjpylint.py
|
An attempt to use the pygram parser to do simple
checking on python source files. Reports references
not set, assignments not used, etcetera. not all warnings
indicate real problems of course. Used like this:
% python kjpylint.py /home/app/arw/Python-1.5a1/Lib/SocketServer.py
setup
loading
now parsing
(verify_request) 'request' defined before 225 not used
(verify_request) 'self' defined before 225 not used
(verify_request) 'client_address' defined before 225 not used
(handle_error) 'request' defined before 245 not used
(handle_error) 'self' defined before 245 not used
(collect_children) 'status' defined before 293 not used
(setup) 'self' defined before 368 not used
(__del__) 'self' defined before 371 not used
(handle) 'self' defined before 374 not used
(finish) 'self' defined before 377 not used
() '__version__' defined before 104 not used
269: (qref) 'max_packet_size' not defined in module?
Here only the last line indicates a possible real bug
in SocketServer.py
Bugs/Features:
Barfs on inconsistent indentation (ie space-tab is not the same as tab).
Grumpy about one line for loops and lambdas.
For very weird cases may not parse strings correctly.
See top of module for more info.
Right now kjpylint might like to see an extra newline
at the end of the file. I'll look into this, sorry.
Latest: Thu Jul 17 13:50:03 EDT 1997
|
| kjParseBuild.py
|
The Parser generator module, used only to generate
the data structures required for parsing.
|
| kjParser.py
|
The Parser module, used both during the generation
phase and also after generation when the generated
parser is used.
|
| kwParsing.html
|
HTML documentation for the package.
|
| kjSet.py
|
support module for parser generation.
(uses kjbuckets
builtin if available, or uses a straight
Python implementation if not.)
|
| kwP.tar.gz
|
The package (tarred, gzipped)
|
kwParsing parser generator related files.
