CEDET

>>> "David PONCE" <David.Ponce@...> seems to think that:
  [ ... ]
>>   Anyway, I have some doc for extending semanticdb I need to write
>>   also, and was thinking it would be good to move it all into a
>>   subdirectory such as "semantic/info".  It would also be good to
>>   split up the existing manual, as it is somewhat hard to edit at
>>   3000+ lines.  As many functions it refers too are now obsolete, I
>>   really need to go through and update it.
>[...]
>
>GNU Emacs uses emacs/man for manual sources (.texi) and emacs/info for
>manual in info format.
>
>Maybe we could use a such organization?  However I would prefer
>"semantic/doc" to "semantic/man" because "man" is already reserved for
>Un*x manuals ;-)

I think semantic/doc makes sense.

  [ ... ]
>>     EL>   Perhaps Richard has some suggestions on this topic.
>> 
>> I would like to suggest that we merge all semantic
>> documentation info one manual.  Of course this doesn't mean
>> that we can't split the manual into several files.
  [ ... ]
>
>Thanks Richard, I love that idea!  I found your outline very good.
>Maybe could you check it in in a new semantic/doc (or what we decide
>to use) directory.  It would be an excellent starting point.
>
>Eric, what do you think?

I agree.  I certainly think that getting things started is a good
idea.

>Could you suggest me where I could split the wisent manual up, and
>where to insert parts in the global outline?  I would really
>appreciate it :-)

Your outline so far is great.  I have the follow suggestions, most of
which are just enhancements on things you probably didn't get to.

It would be great if the outline contained the include statements
where you suggest sections of the manual could be written.  This
would answer David's question most succinctly.

For the parser developer, I can think of two extra sections.  One for
semanticdb extensions,  (If a system database is needed.)  A second
for the `semantic-ctxt' extensions.  Many of the most interesting
tools will completely fail to work without local context parsing
support.

Perhaps even a section on foreign tokens.  For example, putting a
Java token into a C++ file could auto-gen a native method, just as
putting a token into a Texinfo file converts it into documentation.

In addition, in the "writing grammars" section should have
subsections as listed in the examples of the overview section.  It
might be useful to have a fourth section describing the similarities
between the two file types (by and wy) and how to use the grammar
mode.  (I'm not sure if that should be covered elsewhere.)

With those sections, perhaps instead of "Parser Developer", a better
term is "Language Support Developer" where the parser is one subset
of a greater whole of completely abstracting the language so that it
can be used by generic tools.

An application developer section needs to know how to get token lists,
get details from tokens, extract and search those lists with and
without the database, and convert tokens into displayable text
strings in one of the many different formats.

Thanks!
Eric

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

>>>>> "EL" == Eric M Ludlam <eric@...> writes:
    EL> 
    EL> I think semantic/doc makes sense.

I took the liberty of adding semantic/doc directory to CVS.
I also created one file per each chapter.
A simple Makefile is also provided.

    >> Could you suggest me where I could split the wisent manual up, and
    >> where to insert parts in the global outline?  I would really
    >> appreciate it :-)

David, most of wisent manual provbably belongs in
lang-support-guide.texi file.  However please feel free to
experiement with the overal organization. 

    EL> It would be great if the outline contained the include statements
    EL> where you suggest sections of the manual could be written.  This
    EL> would answer David's question most succinctly.

For now each chapter is included by the master file.

I also cut and pasted rest of your comments into the texinfo
files.  The comments were fenced out with @ignore .... @end ignore.
It is probably a good idea to comment out comments meant for
the authors with such fencing of @c comment lines.

In the next few days, I'll try to add more comments and more
sections so that we get settle down on the overall
structure.

Remember everything is open to changes.  Please feel free to
try out different names, different organizations, etc.

On Sun, 23 Feb 2003, ryk@... wrote:

>  Hi all,
>  
>  I use pcl-cvs (aka pcvs in emacs 21) daily.  I use the cvs-mode-idiff
>  differencing command dozens if not hundreds of times each day.
>  Hence the extra 3 or 4 second delay that semantic causes while it
>  parses files cvs-mode-idiff loads for differencing purpose was too
>  much for me especially since these files are used only for
>  differencing and nothing else.
>  
>  So I wrote the code below to conditionally suppress semantic on
>  temporary files.  I tracked semantic-activate-mode-bindings down by
>  starting with semantic-post-change-major-mode-function which was in
>  find-file-hook list.
>  
>  What do you think?  Is there a better way?  Should something like this
>  be a part of semantic?

IMHO there is a better way, means an already builtin way ;-)
Take a look at `semantic-before-toplevel-bovination-hook'.
If i understand your problem and your solution right then the following should
be enough:

,----
| (add-hook 'semantic-before-toplevel-bovination-hook
|           (lambda ()
|             (not (string-match "\\.\\(HEAD$\\|BASE\\)$" (buffer-name)))))
`----

Any hook-function returning nil prevents semantic from parsing.

At least this should do what your solution does...

Thoughts?

Klaus

>  
>  
>  (defadvice semantic-activate-mode-bindings
>    (around suppress-tmp-files activate compile)
>    "Don't setup semantic for tmp files such as files CVS loads for the 
>  purpose of differencing."
>    (cond
>     ;; Bypass files loaded by CVS for differencing.
>     ;; This doesn't work on files retrieved with explicit version numbers.
>     ((string-match "\\.\\(HEAD$\\|BASE\\)$" (buffer-name))
>      ;; Hmm.  I thought the line below would not be needed at first,
>      ;; because bypassing the body of semantic-activate-mode-bindings
>      ;; would not setup semantic variables.  But this advice does not
>      ;; work without the line below. -ryk2/23/03.
>      (setq semantic-toplevel-bovine-table nil))
>     (t ad-do-it)))
>  
>  
>  
>  -------------------------------------------------------
>  This SF.net email is sponsored by: SlickEdit Inc. Develop an edge.
>  The most comprehensive and flexible code editor you can use.
>  Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial.
>  http://www.slickedit.com/sourceforge

-- 
Klaus Berndl			mailto: klaus.berndl@...
sd&m AG				http://www.sdm.de
software design & management	
Thomas-Dehler-Str. 27, 81737 München, Germany
Tel +49 89 63812-392, Fax -220

>>> Klaus Berndl <klaus.berndl@...> seems to think that:
>On Sun, 23 Feb 2003, ryk@... wrote:
>
>>  Hi all,
>>  
>>  I use pcl-cvs (aka pcvs in emacs 21) daily.  I use the cvs-mode-idiff
>>  differencing command dozens if not hundreds of times each day.
>>  Hence the extra 3 or 4 second delay that semantic causes while it
>>  parses files cvs-mode-idiff loads for differencing purpose was too
>>  much for me especially since these files are used only for
>>  differencing and nothing else.
>>  
>>  So I wrote the code below to conditionally suppress semantic on
>>  temporary files.  I tracked semantic-activate-mode-bindings down by
>>  starting with semantic-post-change-major-mode-function which was in
>>  find-file-hook list.
>>  
>>  What do you think?  Is there a better way?  Should something like this
>>  be a part of semantic?
>
>IMHO there is a better way, means an already builtin way ;-)
>Take a look at `semantic-before-toplevel-bovination-hook'.
>If i understand your problem and your solution right then the following should
>be enough:
>
>,----
>| (add-hook 'semantic-before-toplevel-bovination-hook
>|           (lambda ()
>|             (not (string-match "\\.\\(HEAD$\\|BASE\\)$" (buffer-name)))))
>`----
>
>Any hook-function returning nil prevents semantic from parsing.
>
>At least this should do what your solution does...
  [ ... ]

Hi Klaus,

  This is indeed the right kind of solution.  I just went back to
  look at the implementation, and discovered the documentation is
  incorrect for that hook.  It claims that if a hook value returns
  nil, then the cache is returned.  This is no longer true.  Parsing
  will now also returns nil if a hook fails.

  I don't think this matters for Richard's case, but it may for other
  uses.

Eric

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Klaus,

You are absolutely correct.
After I sent out my email, I had a feeling that I should
have done something more to figure this out on my own.

This hook should definitely be added to the manual.
Thanks.

>>>>> "KB" == Klaus Berndl <klaus.berndl@...> writes:
    KB> 
    KB> On Sun, 23 Feb 2003, ryk@... wrote:
    >> Hi all,
    >> 
    >> I use pcl-cvs (aka pcvs in emacs 21) daily.  I use the cvs-mode-idiff
    >> differencing command dozens if not hundreds of times each day.
    >> Hence the extra 3 or 4 second delay that semantic causes while it
    >> parses files cvs-mode-idiff loads for differencing purpose was too
    >> much for me especially since these files are used only for
    >> differencing and nothing else.
    >> 
    >> So I wrote the code below to conditionally suppress semantic on
    >> temporary files.  I tracked semantic-activate-mode-bindings down by
    >> starting with semantic-post-change-major-mode-function which was in
    >> find-file-hook list.
    >> 
    >> What do you think?  Is there a better way?  Should something like this
    >> be a part of semantic?
    KB> 
    KB> IMHO there is a better way, means an already builtin way ;-)
    KB> Take a look at `semantic-before-toplevel-bovination-hook'.
    KB> If i understand your problem and your solution right then the following should
    KB> be enough:
    KB> 
    KB> ,----
    KB> | (add-hook 'semantic-before-toplevel-bovination-hook
    KB> |           (lambda ()
    KB> |             (not (string-match "\\.\\(HEAD$\\|BASE\\)$" (buffer-name)))))
    KB> `----
    KB> 
    KB> Any hook-function returning nil prevents semantic from parsing.
    KB> 
    KB> At least this should do what your solution does...
    KB> 
    KB> Thoughts?
    KB> 
    KB> Klaus
    KB> 
    >> 
    >> 
    >> (defadvice semantic-activate-mode-bindings
    >> (around suppress-tmp-files activate compile)
    >> "Don't setup semantic for tmp files such as files CVS loads for the 
    >> purpose of differencing."
    >> (cond
    >> ;; Bypass files loaded by CVS for differencing.
    >> ;; This doesn't work on files retrieved with explicit version numbers.
    >> ((string-match "\\.\\(HEAD$\\|BASE\\)$" (buffer-name))
    >> ;; Hmm.  I thought the line below would not be needed at first,
    >> ;; because bypassing the body of semantic-activate-mode-bindings
    >> ;; would not setup semantic variables.  But this advice does not
    >> ;; work without the line below. -ryk2/23/03.
    >> (setq semantic-toplevel-bovine-table nil))
    >> (t ad-do-it)))
    >> 
    >> 
    >> 
    >> -------------------------------------------------------
    >> This SF.net email is sponsored by: SlickEdit Inc. Develop an edge.
    >> The most comprehensive and flexible code editor you can use.
    >> Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial.
    >> http://www.slickedit.com/sourceforge
    KB> 
    KB> -- 
    KB> Klaus Berndl			mailto: klaus.berndl@...
    KB> sd&m AG				http://www.sdm.de
    KB> software design & management	
    KB> Thomas-Dehler-Str. 27, 81737 München, Germany
    KB> Tel +49 89 63812-392, Fax -220
    KB> 
    KB> 
    KB> 
    KB> -------------------------------------------------------
    KB> This SF.net email is sponsored by: SlickEdit Inc. Develop an edge.
    KB> The most comprehensive and flexible code editor you can use.
    KB> Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial.
    KB> http://www.slickedit.com/sourceforge
    KB> _______________________________________________
    KB> Cedet-devel mailing list
    KB> Cedet-devel@...
    KB> https://lists.sourceforge.net/lists/listinfo/cedet-devel
    KB>

>>> "Berndl, Klaus" <klaus.berndl@...> seems to think that:
>Hi Eric
>
>>  This is indeed the right kind of solution.  I just went back to
>>  look at the implementation, and discovered the documentation is
>>  incorrect for that hook.  It claims that if a hook value returns
>>  nil, then the cache is returned.  This is no longer true.  Parsing
>>  will now also returns nil if a hook fails.
>
>Oh, does this mean: If one of these hooks fails then always nil is returned,
>regardless of the cache-contents??
>Why this change? What are the advantages compared to the previous
>behavior where simply the parsing is not performed and the cache returned??
  [ ... ]

I think it is a bug that went unnoticed when we refactored the
parsing APIs.  I'll add it to my list of things to fix.

Eric

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

>>> David Ponce <david.ponce@...> seems to think that:
>Hi All,
>
>I checked an updated version of the Wisent's manual in.
>
>I took into account latest changes in both Wisent and Semantic, and
>cleaned up the texinfo code to follow what I saw in Emacs manuals ;-)
>
>Free free to proofread it, and maybe use better English ;-)
  [ ... ]

Hi,

  It looks good.  I tried to read the new sections but am not sure I
  read everything.

  Anyway, I have some doc for extending semanticdb I need to write
  also, and was thinking it would be good to move it all into a
  subdirectory such as "semantic/info".  It would also be good to
  split up the existing manual, as it is somewhat hard to edit at
  3000+ lines.  As many functions it refers too are now obsolete, I
  really need to go through and update it.

  Perhaps Richard has some suggestions on this topic.

Eric

ps.  Here is a short diff

*** wisent.texi.~1.18.~	Sun Feb 23 13:09:09 2003
--- wisent.texi	Sun Feb 23 13:40:20 2003
***************
*** 134,140 ****
  Wisent can generate compilers compatible with the Semantic tool set.
  See the @ref{Top, Semantic Manual, , semantic}.
  
! It benefits of the main Bison's features:
  
  @itemize @bullet
  @item 
--- 134,140 ----
  Wisent can generate compilers compatible with the Semantic tool set.
  See the @ref{Top, Semantic Manual, , semantic}.
  
! It benefits from these Bison features:
  
  @itemize @bullet
  @item 
***************
*** 383,389 ****
  @item $nterm
  This variable is initialized with the nonterminal symbol
  (@var{nonterm}) the rule belongs to.  It could be useful to improve
! error reporting or debugging.  It is used too, to automatically
  provide the @code{reparse-symbol} property of a Semantic token
  generated by the function @code{wisent-cooked-token}.  See
  @ref{Wisent Semantic}, for details.
--- 383,389 ----
  @item $nterm
  This variable is initialized with the nonterminal symbol
  (@var{nonterm}) the rule belongs to.  It could be useful to improve
! error reporting or debugging.  It is also used to automatically
  provide the @code{reparse-symbol} property of a Semantic token
  generated by the function @code{wisent-cooked-token}.  See
  @ref{Wisent Semantic}, for details.
***************
*** 1116,1122 ****
  
  It is important to understand that the parser does not parse
  characters, but lexical tokens, and does not know anything about
! characters and streams!
  
  @cindex lexical analysis
  @cindex lexer
--- 1116,1122 ----
  
  It is important to understand that the parser does not parse
  characters, but lexical tokens, and does not know anything about
! characters in text streams!
  
  @cindex lexical analysis
  @cindex lexer

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

>>>>> "EL" == Eric M Ludlam <eric@...> writes:
    EL> 
    >>>> David Ponce <david.ponce@...> seems to think that:
    >> Hi All,
    >> 
    >> I checked an updated version of the Wisent's manual in.
    >> 
    >> I took into account latest changes in both Wisent and Semantic, and
    >> cleaned up the texinfo code to follow what I saw in Emacs manuals ;-)
    >> 
    >> Free free to proofread it, and maybe use better English ;-)
    EL>   [ ... ]
    EL> 
    EL> Hi,
    EL> 
    EL>   It looks good.  I tried to read the new sections but am not sure I
    EL>   read everything.
    EL> 
    EL>   Anyway, I have some doc for extending semanticdb I need to write
    EL>   also, and was thinking it would be good to move it all into a
    EL>   subdirectory such as "semantic/info".  It would also be good to
    EL>   split up the existing manual, as it is somewhat hard to edit at
    EL>   3000+ lines.  As many functions it refers too are now obsolete, I
    EL>   really need to go through and update it.
    EL> 
    EL>   Perhaps Richard has some suggestions on this topic.

I would like to suggest that we merge all semantic
documentation info one manual.  Of course this doesn't mean
that we can't split the manual into several files.

I provide a very rough outline below.

If we agree on this approach, we can work on refining the
structure (chapters and sections) and brief summaries of
each chapters and sections.  Then we can merge existing
semantic/wisent documents first.  Refining I'm sure would
follow.  

I will be happy to act as the secretary in making pushing
things along so that we have semantic 2.0 manual to go along
with the first release.


\input texinfo  @c -*-texinfo-*-
@c %**start of header
@set TITLE  Semantic Manual
@set AUTHOR Semantic Developers
@setfilename semantic-manual.info
@settitle @value{TITLE}

@node top
@top @value{TITLE}

@menu
* Overview::                    
* Installation::                
* Semantic User's Guide::       
* Application Developer's Guide::  
* Parser Developer's Guide::    
* Semantic Internals::          
* Index::                       
@end menu

@node Overview
@chapter Overview

This chapter gives the overview of semantic and its goals.
Things to mention here could be:

@example
* regular expressions (and syntax tables) are the basis of identifying
  components of a buffer for purposes such as color highlighting
* regexp is useful but has limitations
* semantic provides new intrastructure that goes far beyond regexp
  based text analysis by
  - parse each buffer based on the grammar of the language
  - define a uniform structure of the parse tree mostly independent of
    any one language
  - define an API so that programmers can develop applications that
    work for all languages
  - provide bison-like parser generator so that new languages can be
    supported easily
  - provide a set of applications that demonstrate the usefulness
    of semantic.

The following diagram illustrates the benefits of using semantic:

* Semantic parses a language (for which a parser is available) and
  produces a common parse tree regardless of the original source
  language.

* Semantic provides a common API based on which applications can be
  written in a language independent way.  This means that the
  application will work for ALL languages for which a semantic parser
  can be witten.

* Semantic provids good set of tools and examples for the appplication
  writers.

* Semantic also provies extensive tools for the parser developers as well.
  For many languages, those that can be described using a context-free
  grammer, all that one needs to do is write a grammar file along with
  appropriate semantic rules.


The words in all-capital are those that semantic itself provides.
Others are current or future languages or applications that are not
distributed along with semantic.


                                                             Applications
                                                                 and
                                                              Utilities
                                                                -------
                                                               /       \
               +---------------+    +--------+    +--------+
         C --->| C      PARSER |--->|        |    |        |
               +---------------+    |        |    |        |
               +---------------+    | COMMON |    | COMMON |<--- SPEEDBAR
      Java --->| JAVA   PARSER |--->|        |    |        |
               +---------------+    | PARSE  |    | PARSE  |<--- SENATOR
               +---------------+    |        |    |        |
    Python --->| PYTHON PARSER |--->| TREE   |    | TREE   |<--- DOCUMENT
               +---------------+    |        |    |        |
               +---------------+    | FORMAT |    | API    |<--- SEMANTICDB
    Scheme --->| SCHEME PARSER |--->|        |    |        |
               +---------------+    |        |    |        |<--- jdee
               +---------------+    |        |    |        |
   Texinfo --->| TEXI.  PARSER |--->|        |    |        |<--- ecb
               +---------------+    |        |    |        |

                    ...                ...           ...         ...

               +---------------+    |        |    |        |<--- app. 1
   Lang. A --->| A      Parser |--->|        |    |        |
               +---------------+    |        |    |        |<--- app. 2
               +---------------+    |        |    |        |
   Lang. B --->| B      Parser |--->|        |    |        |<--- app. 3
               +---------------+    |        |    |        |

                     ...        ...     ...          ...       ...

               +---------------+    |        |    |        |
   Lang. Y --->| Y      Parser |--->|        |    |        |<--- app. ?
               +---------------+    |        |    |        |
               +---------------+    |        |    |        |<--- app. ?
   Lang. Z --->| Z      Parser |--->|        |    |        |
               +---------------+    +--------+    +--------+
@end example

@node Installation
@chapter Installation

@node Semantic User's Guide
@chapter Semantic User's Guide

This chapter describes how to use semantic for the end user.
Hopefully not much elisp knowledge is required for the
target audience of this chapter.

Also describe each of the applications that come with semantic
as well as mention third party applications such as jdee and ecb.

@menu
* semanticdb::                  
* speedbar::                    
* senator::                     
* analyzer::                    
* class browser::               
* document::                    
* charts::                      
@end menu

@node semanticdb
@section semanticdb
@cindex semanticdb

Describe what this is and how the user can control its behavior.

@node speedbar
@section speedbar
@cindex speedbar

Give brief overview and refer to the speedbar documentation.

@node senator
@section senator
@cindex senator

@node analyzer
@section analyzer
@cindex analyzer

@node class browser
@section class browser
@cindex class browser

@node document
@section document
@cindex document

@node charts
@section charts
@cindex charts

@node Application Developer's Guide
@chapter Application Developer's Guide

This chapter is for developers of semantic applications
such as JDEE and ECB.
Describe all things that such developers would need to know
such as semantic token structures and API.

Most of the stuff from @ref{Programming,,,semantic}
would go here.

@node Parser Developer's Guide
@chapter Parser Developer's Guide

This chapter is for those who would like to add a new parser to the
growing family of parsers.
This chapter should
@example
* describe wisent LALR parser, bovine LL parser, and give examples of
  custom parsers.
* describe grammar file formats, semantic actions.
* describe how to write lexers
* give examples such as wisent-awk, wisent-calc, etc
@end example    

@menu
* Parser Overview::             
* Writing Grammars::            
* Writing Lexers::              
* Parser Error Handling::       
@end menu

@node Parser Overview
@section Parser Overview
@cindex Parser Overview

@example
wisent parsers
bovine parsers
custom parsers
@end example

@node Writing Grammars
@section Writing Grammars
@cindex Writing Grammars

@node Writing Lexers
@section Writing Lexers
@cindex Writing Lexers

@example
semantic-flex?
new lexer based on macros
@end example

@node Parser Error Handling
@section Parser Error Handling
@cindex Parser Error Handling

@node Semantic Internals
@chapter Semantic Internals

This chapter explains the internals of semantic.
This information would not be needed by neither applicatioan developers
nor grammar developers.

It would be useful mostly for the hackers who would like to learn
more about how semantic works.

@node Index
@unnumbered Index
@printindex cp

@iftex
@contents
@summarycontents
@end iftex

@bye

Hi Eric,

 > Emacs developers are in the process of eliminating all occurrences of
 > the word "path" in references to anything that is not a list of
 > directories.  To accommodate this I've synchronized GNU Emacs
 > speedbar.el changes (as made by the FSF) into the CEDET version of
 > speedbar.

In drame.el I noticed the following change:

***************
*** 374,380 ****
         (setq temp-buffer-show-function 'dframe-temp-buffer-show-function)
         ;; If this buffer is killed, we must make sure that we destroy
         ;; the frame the dedicated window is in.
-       (make-local-hook 'kill-buffer-hook)
         (add-hook 'kill-buffer-hook `(lambda ()
   				     (let ((skilling (boundp 'skilling)))
   				       (if skilling
--- 372,377 ----

Maybe it could cause problem, unless you use the development version
of GNU Emacs 21.4.

IMO, in that latter version only, `make-local-hook' automatically make
a hook buffer local, when called with a non-nil LOCAL argument.

 > Second, I changed all occurrences of `path' to `directory', as I was
 > using `path' to describe a directory anyway.
[...]
 > The most intrusive change, that of `speedbar-line-path' unfortunately
 > also changes a key override API for custom displays, changing
 > `line-path' to `line-directory'.  An alias cannot fix this problem.

In case it will help, I did the following changes in my local copy of
cedet:

Index: ede/ede-speedbar.el
===================================================================
RCS file: /cvsroot/cedet/cedet/ede/ede-speedbar.el,v
retrieving revision 1.22
diff -c -r1.22 ede-speedbar.el
*** ede/ede-speedbar.el	5 Dec 2001 01:25:05 -0000	1.22
--- ede/ede-speedbar.el	22 Feb 2003 08:49:28 -0000
***************
*** 262,268 ****
     "Find the file TEXT at path TOKEN.
   INDENT is the current indentation level."
     (speedbar-find-file-in-frame
!    (concat (speedbar-line-path) token)))

   (defun ede-create-tag-buttons (filename indent)
     "Create the tag buttons associated with FILENAME at INDENT."
--- 262,268 ----
     "Find the file TEXT at path TOKEN.
   INDENT is the current indentation level."
     (speedbar-find-file-in-frame
!    (concat (speedbar-line-directory) token)))

   (defun ede-create-tag-buttons (filename indent)
     "Create the tag buttons associated with FILENAME at INDENT."
Index: eieio/eieio-speedbar.el
===================================================================
RCS file: /cvsroot/cedet/cedet/eieio/eieio-speedbar.el,v
retrieving revision 1.14
diff -c -r1.14 eieio-speedbar.el
*** eieio/eieio-speedbar.el	13 Mar 2002 12:14:54 -0000	1.14
--- eieio/eieio-speedbar.el	22 Feb 2003 08:49:28 -0000
***************
*** 168,174 ****
     (speedbar-add-mode-functions-list
      (list modename
   	 '(speedbar-item-info . eieio-speedbar-item-info)
! 	 '(speedbar-line-path . eieio-speedbar-line-path))))

   (defun eieio-speedbar-buttons (dir-or-object depth fetcher)
     "Create buttons for the speedbar display.
--- 168,174 ----
     (speedbar-add-mode-functions-list
      (list modename
   	 '(speedbar-item-info . eieio-speedbar-item-info)
! 	 '(speedbar-line-directory . eieio-speedbar-line-path))))

   (defun eieio-speedbar-buttons (dir-or-object depth fetcher)
     "Create buttons for the speedbar display.
Index: semantic/semantic-ia-sb.el
===================================================================
RCS file: /cvsroot/cedet/cedet/semantic/semantic-ia-sb.el,v
retrieving revision 1.8
diff -c -r1.8 semantic-ia-sb.el
*** semantic/semantic-ia-sb.el	9 Aug 2002 23:21:38 -0000	1.8
--- semantic/semantic-ia-sb.el	22 Feb 2003 08:49:29 -0000
***************
*** 63,69 ****
   (speedbar-add-mode-functions-list
    (list "Analyze"
          ;;'(speedbar-item-info . eieio-speedbar-item-info)
!        '(speedbar-line-path . semantic-ia-sb-line-path)))

   ;;;###autoload
   (defun semantic-speedbar-analysis ()
--- 63,69 ----
   (speedbar-add-mode-functions-list
    (list "Analyze"
          ;;'(speedbar-item-info . eieio-speedbar-item-info)
!        '(speedbar-line-directory . semantic-ia-sb-line-path)))

   ;;;###autoload
   (defun semantic-speedbar-analysis ()
Index: semantic/semantic-sb.el
===================================================================
RCS file: /cvsroot/cedet/cedet/semantic/semantic-sb.el,v
retrieving revision 1.36
diff -c -r1.36 semantic-sb.el
*** semantic/semantic-sb.el	21 Feb 2003 18:05:30 -0000	1.36
--- semantic/semantic-sb.el	22 Feb 2003 08:49:29 -0000
***************
*** 245,251 ****
   (defun semantic-sb-token-jump (text token indent)
     "Jump to the location specified in token.
   TEXT TOKEN and INDENT are the details."
!   (let ((file (speedbar-line-path indent))
   	(parent (semantic-sb-detail-parent)))
       (let ((f (selected-frame)))
         (dframe-select-attached-frame speedbar-frame)
--- 245,251 ----
   (defun semantic-sb-token-jump (text token indent)
     "Jump to the location specified in token.
   TEXT TOKEN and INDENT are the details."
!   (let ((file (speedbar-line-directory indent))
   	(parent (semantic-sb-detail-parent)))
       (let ((f (selected-frame)))
         (dframe-select-attached-frame speedbar-frame)

[...]

David

Hi David,

>>> David Ponce <david.ponce@...> seems to think that:
>Hi Eric,
>
> > Emacs developers are in the process of eliminating all occurrences of
> > the word "path" in references to anything that is not a list of
> > directories.  To accommodate this I've synchronized GNU Emacs
> > speedbar.el changes (as made by the FSF) into the CEDET version of
> > speedbar.
>
>In drame.el I noticed the following change:
>
>***************
>*** 374,380 ****
>         (setq temp-buffer-show-function 'dframe-temp-buffer-show-function)
>         ;; If this buffer is killed, we must make sure that we destroy
>         ;; the frame the dedicated window is in.
>-       (make-local-hook 'kill-buffer-hook)
>         (add-hook 'kill-buffer-hook `(lambda ()
>   				     (let ((skilling (boundp 'skilling)))
>   				       (if skilling
>--- 372,377 ----
>
>Maybe it could cause problem, unless you use the development version
>of GNU Emacs 21.4.
>
>IMO, in that latter version only, `make-local-hook' automatically make
>a hook buffer local, when called with a non-nil LOCAL argument.

This was eliminated in Emacs's CVS copy of speedbar.  Add hook's 4th
argument also specifies it may be local.

I was hoping I could get away with it, since every change of theirs
I ignore becomes one more merge hassle later.  I hadn't tested this
in older versions of Emacs yet.

> > Second, I changed all occurrences of `path' to `directory', as I was
> > using `path' to describe a directory anyway.
>[...]
> > The most intrusive change, that of `speedbar-line-path' unfortunately
> > also changes a key override API for custom displays, changing
> > `line-path' to `line-directory'.  An alias cannot fix this problem.
>
>In case it will help, I did the following changes in my local copy of
>cedet:

Thanks for looking into all those spots for me.  If you can check
them all it, that would be a great first step in making the rest of
CEDET compatible.  Thanks!

  [ ... ]
>Index: semantic/semantic-sb.el

I will be doing some more with this, and the other two speedbar using
semantic files as well.

This change that any future release of semantic (beta or otherwise)
now requires also a new release of speedbar and a new release of
eieio.  Using those means also a new release of EDE.  To make it all
work we should release a CEDET commons with inversion.  Zoiks.  Lots
to do now.

Thanks
Eric

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Eric,

[...]
 >>In case it will help, I did the following changes in my local copy of
 >>cedet:
 >
 >
 > Thanks for looking into all those spots for me.  If you can check
 > them all it, that would be a great first step in making the rest of
 > CEDET compatible.  Thanks!

Done.  I also committed a small fix into wisent-python.wy and a more
important compatibility fix into semantic-grammar.el.

Here is a change log:

2003-02-22  David Ponce  <david@...>

         * ede/ede-speedbar.el

         (ede-file-find): Replaced `speedbar-line-path' with
         `speedbar-line-directory' used since speedbar 0.15.

         * eieio/eieio-speedbar.el
         (eieio-speedbar-create-engine): Replaced `speedbar-line-path' with
         `speedbar-line-directory' used since speedbar 0.15.

         * semantic/semantic-grammar.el

         No more depend on `with-syntax-table' implementation.  For
         example, since GNU Emacs 21.4 `with-syntax-table' no more use a
         copy of the given syntax table.

         (semantic-grammar-skip-quoted-syntax-table): New variable.
         (semantic-grammar-goto-grammar-indent-anchor): Use it.
         (semantic-grammar-brackets-as-parens-syntax-table): New variable.
         (semantic-grammar-do-lisp-indent): Use it.

         * semantic/semantic-ia-sb.el

         (speedbar-add-mode-functions-list "Analyse"): Replaced
         `speedbar-line-path' with `speedbar-line-directory' used since
         speedbar 0.15.

         * semantic/semantic-sb.el

         (semantic-sb-token-jump): Replaced `speedbar-line-path' with
         `speedbar-line-directory' used since speedbar 0.15.

         * semantic/wisent/wisent-python.wy

         (goal): Fixed comment delimiter.

[...]
 > This change that any future release of semantic (beta or otherwise)
 > now requires also a new release of speedbar and a new release of
 > eieio.  Using those means also a new release of EDE.  To make it all
 > work we should release a CEDET commons with inversion.  Zoiks.  Lots
 > to do now.

Maybe it would be worth providing global cedet releases, as I already
proposed sometimes ago ;-)

David

Hi Richard,
        
> I just updated my local copies to latest CVS versions.
> Nothing seems to be broken with wisent-python.el.

Great!

> The only minor thing is that pp-to-string is not autoloaded
> in GNU/Emacs 21.2, so I added (require 'pp) to get things to
> work.  Other than that, no problems.  Very smooth upgrade
> for such a large change.  Good work!

Thanks!  That should be fixed now.  Functions that uses
`pp-to-string' ensure that `pp' is loaded at runtime.

David


2003-02-20  David Ponce  <david@...>

        * semantic-grammar.el

        (semantic-grammar-as-string): Requires `pp' feature.

        * wisent/wisent-debug.el
        
        Requires `pp' if `pprint' feature was not found.

That sounds like a really clever idea.  The bovinator could use this
too.  Right now (or at least, last time I looked at it) the parser was
a flat table, and `assoc' is used to lookup rules values.  The same
trick could apply.  Nifty.

Eric

>>> "David PONCE" <David.Ponce@...> seems to think that:
>Hi All,
>
>I am about to check-in a big change to Wisent!
>
>Here is a summary:
>
>- The internal data structure of LALR automatons is improved.
>
>  * The goto table is filled with nonterminal symbols instead of item
>    numbers.
>
>  * The vector of semantic action lambda expressions is replaced by an
>    obarray.  Each semantic action is now represented by a symbol
>    interned in that obarray.  `symbol-function' on a semantic action
>    symbol return the semantic action lambda expression.
>
>    A semantic action symbol name has the form NONTERMINAL:INDEX, where
>    NONTERMINAL is the name of the nonterminal symbol the action
>    belongs to, and INDEX is an action sequence number within the scope
>    of NONTERMINAL.  For example, this nonterminal definition:
>
>    input:
>        line                     [input:0]
>      | input line
>        (format "%s %s" $1 $2)   [input:1]
>      ;
>
>    will produce two semantic actions, and associated symbols:
>
>      `input:0', a default action that returns $1.
>      `input:1', that returns `(format "%s %s" $1 $2)'.
>
>  * The action table is filled with semantic action symbols instead of
>    negative item numbers.
>
>  * An automaton is now a vector:
>    [action-table goto-table start-alist semantic-actions-obarray].
>
>- The LR parser engine benefits of the above changes.  It no more
>  needs to handle translation of item numbers, and it can directly
>  call semantic actions.  Consequently it should be faster!
>
>- Grammar compilation follows normal Emacs Lisp byte-compilation
>  strategy.  That is, when a `wisent-compile-grammar' form is
>  evaluated it produces a LALR automaton without semantic actions
>  byte-compiled.  When byte-compiled it produces byte-code for the
>  whole automaton, including semantic actions.
>
>- I introduced a new library: wisent-debug.el, that takes advantage of
>  new automaton data structure and byte-compilation strategy, to
>  provide some debugging features for wisent.  For now, I implemented
>  low level debugging of semantic actions, using the standard Emacs
>  debugger.
>
>  Mainly three commands are available:
>
>  `wisent-debug-on-entry'/`wisent-cancel-debug-on-entry', that mimic
>  `debug-on-entry'/`cancel-debug-on-entry' for semantic actions.
>
>  `wisent-debug-show-entry' that shows the code of a semantic action.
>
>Unfortunately, the new automaton format is incompatible with the
>current one.  It will be necessary to update .el files from .wy
>grammars, and to recompile them :-(
>
>If you want to try it, I attached a tarball that includes a big patch
>to semantic (from today CVS version), and the new wisent-debug.el
>file.
>
>Finally a change log is at end.
  [ ... ]

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

Hi David,

I just updated my local copies to latest CVS versions.
Nothing seems to be broken with wisent-python.el.

The only minor thing is that pp-to-string is not autoloaded
in GNU/Emacs 21.2, so I added (require 'pp) to get things to
work.  Other than that, no problems.  Very smooth upgrade
for such a large change.  Good work!


Index: semantic-grammar.el
===================================================================
RCS file: /cvsroot/cedet/cedet/semantic/semantic-grammar.el,v
retrieving revision 1.11
diff -u -r1.11 semantic-grammar.el
--- semantic-grammar.el	19 Feb 2003 16:25:29 -0000	1.11
+++ semantic-grammar.el	20 Feb 2003 05:35:26 -0000
@@ -36,6 +36,7 @@
 (require 'wisent-bovine)
 
 (eval-when-compile
+  (require 'pp)	;; for pp-to-string
   (require 'font-lock)
   (require 'semantic-edit))
 



>>>>> "DP" == David PONCE <David.Ponce@...> writes:
    DP> 
    DP> Hi All,
    DP> I am about to check-in a big change to Wisent!
    DP> 
    DP> Here is a summary:
    DP> 
    DP> - The internal data structure of LALR automatons is improved.
    DP> 
    DP>   * The goto table is filled with nonterminal symbols instead of item
    DP>     numbers.
    DP> 
    DP>   * The vector of semantic action lambda expressions is replaced by an
    DP>     obarray.  Each semantic action is now represented by a symbol
    DP>     interned in that obarray.  `symbol-function' on a semantic action
    DP>     symbol return the semantic action lambda expression.
    DP> 
    DP>     A semantic action symbol name has the form NONTERMINAL:INDEX, where
    DP>     NONTERMINAL is the name of the nonterminal symbol the action
    DP>     belongs to, and INDEX is an action sequence number within the scope
    DP>     of NONTERMINAL.  For example, this nonterminal definition:
    DP> 
    DP>     input:
    DP>         line                     [input:0]
    DP>       | input line
    DP>         (format "%s %s" $1 $2)   [input:1]
    DP>       ;
    DP> 
    DP>     will produce two semantic actions, and associated symbols:
    DP> 
    DP>       `input:0', a default action that returns $1.
    DP>       `input:1', that returns `(format "%s %s" $1 $2)'.
    DP> 
    DP>   * The action table is filled with semantic action symbols instead of
    DP>     negative item numbers.
    DP> 
    DP>   * An automaton is now a vector:
    DP>     [action-table goto-table start-alist semantic-actions-obarray].
    DP> 
    DP> - The LR parser engine benefits of the above changes.  It no more
    DP>   needs to handle translation of item numbers, and it can directly
    DP>   call semantic actions.  Consequently it should be faster!
    DP> 
    DP> - Grammar compilation follows normal Emacs Lisp byte-compilation
    DP>   strategy.  That is, when a `wisent-compile-grammar' form is
    DP>   evaluated it produces a LALR automaton without semantic actions
    DP>   byte-compiled.  When byte-compiled it produces byte-code for the
    DP>   whole automaton, including semantic actions.
    DP> 
    DP> - I introduced a new library: wisent-debug.el, that takes advantage of
    DP>   new automaton data structure and byte-compilation strategy, to
    DP>   provide some debugging features for wisent.  For now, I implemented
    DP>   low level debugging of semantic actions, using the standard Emacs
    DP>   debugger.
    DP> 
    DP>   Mainly three commands are available:
    DP> 
    DP>   `wisent-debug-on-entry'/`wisent-cancel-debug-on-entry', that mimic
    DP>   `debug-on-entry'/`cancel-debug-on-entry' for semantic actions.
    DP> 
    DP>   `wisent-debug-show-entry' that shows the code of a semantic action.
    DP> 
    DP> Unfortunately, the new automaton format is incompatible with the
    DP> current one.  It will be necessary to update .el files from .wy
    DP> grammars, and to recompile them :-(
    DP> 
    DP> If you want to try it, I attached a tarball that includes a big patch
    DP> to semantic (from today CVS version), and the new wisent-debug.el
    DP> file.
    DP> 
    DP> Finally a change log is at end.
    DP> 
    DP> Thoughts?
    DP> Enjoy!
    DP> 
    DP> David
    DP> 
    DP> ------- Change Log
    DP> 
    DP>   * semantic-grammar.el
    DP> 
    DP>   (semantic-grammar-automaton)
    DP>   (semantic-grammar-keywords, semantic-grammar-tokens)
    DP>   (emantic-grammar-setup-semantic): Re-generated.
    DP> 
    DP>   * wisent/wisent-awk.el
    DP> 
    DP>   (wisent-awk-automaton)
    DP>   (wisent-awk-keywords, wisent-awk-tokens)
    DP>   (wisent-awk-setup-parser): Re-generated.
    DP> 
    DP>   * wisent/wisent-calc.el
    DP> 
    DP>   (wisent-calc-automaton, wisent-calc-tokens)
    DP>   (wisent-calc-setup-parser): Re-generated.
    DP> 
    DP>   * wisent/wisent-cim.el
    DP> 
    DP>   (wisent-cim-automaton)
    DP>   (wisent-cim-keywords, wisent-cim-tokens)
    DP>   (wisent-cim-default-setup): Re-generated.
    DP> 
    DP>   * wisent/wisent-comp.el
    DP> 
    DP>   (wisent-defcontext semantic-actions): Moved before first use.
    DP>   (wisent-token-actions): Create new semantic actions obarray.  Fill
    DP>   the action table with semantic action symbols instead of negative
    DP>   item numbers.
    DP>   (wisent-goto-actions): Fill the goto table with nonterminal
    DP>   symbols instead of item numbers.
    DP>   (wisent-semantic-action): Store code of action in the function
    DP>   cell of a symbol interned in the semantic actions obarray.  Return
    DP>   that symbol.  The new semantic action variable `$action' is bound
    DP>   to the name of that symbol.
    DP>   (wisent-semantic-actions, wisent-token-translations): Removed.
    DP>   (wisent-parser-automaton): Return a vector of 4 elements: the
    DP>   action table, the goto table, the alist of start symbols, and the
    DP>   semantic actions obarray.
    DP>   (wisent-parse-nonterminals): Produce symbolic location of semantic
    DP>   actions.
    DP>   (wisent-parse-grammar): Intern generated symbols.
    DP>   (wisent-compiled-grammar-p): Removed.
    DP>   (wisent-compile-grammar): Doc fix.  Use `wisent-automaton-p'.
    DP>   (wisent-byte-compile-grammar)
    DP>   (wisent-byte-compile-automaton): New functions.
    DP> 
    DP>   * wisent/wisent-grammar.el
    DP> 
    DP>   (wisent-grammar-parsetable-builder): Ensure that the grammar
    DP>   [byte-]compiler is available to handle byte-compilation of the
    DP>   grammar.
    DP> 
    DP>   * wisent/wisent-java-tags.el
    DP>   * wisent/wisent-java.el
    DP> 
    DP>   (wisent-java-parser-tables)
    DP>   (wisent-java-keywords, wisent-java-tokens)
    DP>   (wisent-java-default-setup): Re-generated.
    DP> 
    DP>   * wisent/wisent-python.el
    DP> 
    DP>   (wisent-python-parser-tables)
    DP>   (wisent-python-keywords, wisent-python-tokens)
    DP>   (wisent-python-default-setup): Re-generated.
    DP> 
    DP>   * wisent/wisent.el
    DP> 
    DP>   (autoload 'wisent-compile-grammar): Removed.
    DP>   (wisent-automaton-p): New.
    DP>   (wisent-translate, wisent-untranslate): Removed.
    DP>   (wisent-parse): Use new automaton format.
    DP> 
    DP>   * wisent/wisent-debug.el
    DP> 
    DP>   New file.
    DP> 
    DP>   * wisent/Project.ede
    DP>   (ede-proj-target-elisp "wisent"): Added wisent-debug.el.
    DP> 
    DP>   * wisent/Makefile
    DP> 
    DP>   Re-generated.
    DP> 
    DP>

Hi Eric,

I missed changing "depth" to "semantic-lex-maximum-depth", because I
blindly looked for "maximum-depth" in my code to replace with
"semantic-lex-maximum-depth" instead of "depth".  For every
semantic-lex-XXX that you mentioned in your email, I looked for XXX to
replace.

I fixed it.  Also I managed to get rid of all warnings except one
reference to "end".  I need to know the end point especially when
re-parsing due to EXPANDFULL so that proper number of DEDENT tokens
can be generated when the end point is reached.  If not the "end" free
variable, what else can I use to detect that the cursor is at the end
of region currently being parsed?

I checked in the changes.
So python parse now works again.

Thanks for your help.

>>>>> "EL" == Eric M Ludlam <eric@...> writes:
    EL> 
    EL> I tried out the python parser with the latest lexer.  I ran into a
    EL> problem because when I use "make", it uses emacs 20, but I was
    EL> running Emacs 21.  There is some sort of incompatibility in th cl
    EL> package that caused it to error out.
    EL> 
    EL> Once I got past that, a byte compile threw these warnings:
    EL> 
    EL> Compiling file /home/zappo/cedet/semantic/wisent/wisent-python.el at Tue Feb 18 22:42:21 2003
    EL> Entering directory `/home/zappo/cedet/semantic/wisent/'
    EL> 
    EL> In semantic-lex-python-pop-indent-stack:
    EL> wisent-python.el:74:19:Warning: reference to free variable end
    EL> 
    EL> In toplevel form:
    EL> wisent-python.el:108:62:Warning: reference to free variable depth
    EL> wisent-python.el:132:30:Warning: reference to free variable start
    EL> wisent-python.el:132:36:Warning: reference to free variable end
    EL> wisent-python.el:193:42:Warning: reference to free variable opos
    EL> wisent-python.el:229:13:Warning: reference to free variable opos
    EL> 
    EL> which may be causing some problems.
    EL> 
    EL> Anyway, I parsed the buffer, and the only unmatched syntax that came
    EL> up was on some of the indentation whitespace, and then again for a
    EL> couple symbols here and there.  The "x=1" type line parsed fine.
    EL> 
    EL> Eric

>>> ryk@... seems to think that:
>Hi Eric,
>
>I missed changing "depth" to "semantic-lex-maximum-depth", because I
>blindly looked for "maximum-depth" in my code to replace with
>"semantic-lex-maximum-depth" instead of "depth".  For every
>semantic-lex-XXX that you mentioned in your email, I looked for XXX to
>replace.

Oh, did I mess mentioning some variables in my email?  I apologize.

>I fixed it.  Also I managed to get rid of all warnings except one
>reference to "end".  I need to know the end point especially when
>re-parsing due to EXPANDFULL so that proper number of DEDENT tokens
>can be generated when the end point is reached.  If not the "end" free
>variable, what else can I use to detect that the cursor is at the end
>of region currently being parsed?
  [ ... ]

The `start' and `end' variables are still there, but it is better to
use the new state variable which is created like this:

	    ;; Bounds needed for unterminated syntax
	    (semantic-lex-analysis-bounds (cons start end))

Thus, you can replace `end' with (cdr semantic-lex-analysis-bounds).
When protecting against unterminated syntax, you can use the
`semantic-lex-unterminated-syntax-protection' macro which hides
that.  I don't know if that is appropriate for your INDENT/DETEND loop
or not, but if you have your match-code throw an error, it would work.

Good Luck
Eric

-- 
          Eric Ludlam:                 zappo@..., eric@...
   Home: http://www.ludlam.net            Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net               GNU: http://www.gnu.org

>>>>> "EL" == Eric M Ludlam <eric@...> writes:
    EL> 
    >>>> ryk@... seems to think that:
    >> Hi Eric,
    >> 
    >> I missed changing "depth" to "semantic-lex-maximum-depth", because I
    >> blindly looked for "maximum-depth" in my code to replace with
    >> "semantic-lex-maximum-depth" instead of "depth".  For every
    >> semantic-lex-XXX that you mentioned in your email, I looked for XXX to
    >> replace.
    EL> 
    EL> Oh, did I mess mentioning some variables in my email?  I apologize.

No need.  If I was thinking even a little bit, I should have
figured it out.

    EL> The `start' and `end' variables are still there, but it is better to
    EL> use the new state variable which is created like this:
    EL> 
    EL> 	    ;; Bounds needed for unterminated syntax
    EL> 	    (semantic-lex-analysis-bounds (cons start end))
    EL> 
    EL> Thus, you can replace `end' with (cdr semantic-lex-analysis-bounds).
    EL> When protecting against unterminated syntax, you can use the
    EL> `semantic-lex-unterminated-syntax-protection' macro which hides
    EL> that.  I don't know if that is appropriate for your INDENT/DETEND loop
    EL> or not, but if you have your match-code throw an error, it would work.
    EL> 
    EL> Good Luck
    EL> Eric

I justed checked in wisent-python.el where I replaced 'end'
free variable with (cdr semantic-lex-analysis-bounds) as you
suggested.  I verified that the python parser still works
afterwards.  Byte compiler does not report any warnings for
wisent-python.el anymore.  Thanks.