lxr-developer — Don't use, active developers only

Re: [Lxr-developer] Code documentation

[Neil S. <ns...@ci...>]

> I couldn't actually find a doxygen page with a link to the lxr'd source
> code,

...no, neither could I, but they seem to have had the idea :-) ...

> but I'm sure there are many possibilites for doxygen and lxr to work
> together.

I agree.

> What I want to do is to understand the lxr source code and document it
using
> doxygen. A first step....

I can't help here, I'm not an LXR contributor - just now, at least. All I
can offer is moral support.

So good luck with it! I think LXR (it's probably a little late to rename =
it,
but it has uses far beyond browsing the Linux source) is a killer app for
developers on large projects.

Cheers,

Neil.

> Zubin
>
> -----Original Message-----
> From: Neil Salter [mailto:ns...@ci...]
> Sent: Tuesday, 12 June 2001 5:36 PM
> To: lxr...@li...
> Subject: Re: [Lxr-developer] Code documentation
>
>
> Hello,
>
> The guys here: http://unstable.elemental.com/mozilla/ had the idea of
cross
> linking between doxygen and LXR. I think this is a good idea. Currently
I'm
> using doxygen to document a (fairly large) source tree, and I plan to u=
se
> LXR on the same tree. I'd say it'd be pretty useful to be able to jump
from,
> say, a classname in LXR to the doxygen page for the class - and vice
versa.
> It could all be done manually by the user just now, but it might be wor=
th
> thinking about as potential future feature for incorporation into the
> software.
>
> Just my =A30.01.
>
> Cheers,
>
> Neil.
>
> ----- Original Message -----
> From: "Malcolm Box" <ma...@br...>
> To: <lxr...@li...>
> Sent: Tuesday, June 12, 2001 3:03 AM
> Subject: Re: [Lxr-developer] Code documentation
>
>
> > Hi Zubin,
> >
> > That would be excellent!  I've used Doxygen before, and it's a great
> > tool.  I wasn't aware that it supported Perl, but if it does then let=
's
> > go for it.
> >
> > As far as I'm concerned, any patch that improves comments or
> > documentation in any way will always be gladly accepted.
> >
> > Cheers,
> >
> > Malcolm
> >
> > Sethna, Zubin (RSA, Brisbane) wrote:
> >
> > >Hi folks
> > >
> > >How to people feel about adding Doxygen comments to the LXR source
code?
> > >Doxygen generates html pages from the comments you insert into the
source
> > >code. Just a header to each function describing what it does and its
> > >input/output? I would like to start this if I get the go ahead from =
u
> all.
> > >
> > >I'm a newbie to perl and I think it would ease the learning curve fo=
r
> people
> > >like me who wish to contribute code to this project.
> > >
> > >
> > >Regards
> > >
> > >
> > >
> > >Zubin
> > >
> > >
> > >_______________________________________________
> > >Lxr-developer mailing list
> > >Lxr...@li...
> > >http://lists.sourceforge.net/lists/listinfo/lxr-developer
> > >
> >
> >
> >
> > _______________________________________________
> > Lxr-developer mailing list
> > Lxr...@li...
> > http://lists.sourceforge.net/lists/listinfo/lxr-developer
>
>
> _______________________________________________
> Lxr-developer mailing list
> Lxr...@li...
> http://lists.sourceforge.net/lists/listinfo/lxr-developer
>
> _______________________________________________
> Lxr-developer mailing list
> Lxr...@li...
> http://lists.sourceforge.net/lists/listinfo/lxr-developer

RE: [Lxr-developer] Code documentation

[Sethna, Z. (R. Brisbane) <ZS...@rs...>]

I couldn't actually find a doxygen page with a link to the lxr'd source
code, but I'm sure there are many possibilites for doxygen and lxr to =
work
together.

What I want to do is to understand the lxr source code and document it =
using
doxygen. A first step....

Zubin

-----Original Message-----
From: Neil Salter [mailto:ns...@ci...]
Sent: Tuesday, 12 June 2001 5:36 PM
To: lxr...@li...
Subject: Re: [Lxr-developer] Code documentation


Hello,

The guys here: http://unstable.elemental.com/mozilla/ had the idea of =
cross
linking between doxygen and LXR. I think this is a good idea. Currently =
I'm
using doxygen to document a (fairly large) source tree, and I plan to =
use
LXR on the same tree. I'd say it'd be pretty useful to be able to jump =
from,
say, a classname in LXR to the doxygen page for the class - and vice =
versa.
It could all be done manually by the user just now, but it might be =
worth
thinking about as potential future feature for incorporation into the
software.

Just my =A30.01.

Cheers,

Neil.

----- Original Message -----
From: "Malcolm Box" <ma...@br...>
To: <lxr...@li...>
Sent: Tuesday, June 12, 2001 3:03 AM
Subject: Re: [Lxr-developer] Code documentation


> Hi Zubin,
>
> That would be excellent!  I've used Doxygen before, and it's a great
> tool.  I wasn't aware that it supported Perl, but if it does then =
let's
> go for it.
>
> As far as I'm concerned, any patch that improves comments or
> documentation in any way will always be gladly accepted.
>
> Cheers,
>
> Malcolm
>
> Sethna, Zubin (RSA, Brisbane) wrote:
>
> >Hi folks
> >
> >How to people feel about adding Doxygen comments to the LXR source =
code?
> >Doxygen generates html pages from the comments you insert into the =
source
> >code. Just a header to each function describing what it does and its
> >input/output? I would like to start this if I get the go ahead from =
u
all.
> >
> >I'm a newbie to perl and I think it would ease the learning curve =
for
people
> >like me who wish to contribute code to this project.
> >
> >
> >Regards
> >
> >
> >
> >Zubin
> >
> >
> >_______________________________________________
> >Lxr-developer mailing list
> >Lxr...@li...
> >http://lists.sourceforge.net/lists/listinfo/lxr-developer
> >
>
>
>
> _______________________________________________
> Lxr-developer mailing list
> Lxr...@li...
> http://lists.sourceforge.net/lists/listinfo/lxr-developer


_______________________________________________
Lxr-developer mailing list
Lxr...@li...
http://lists.sourceforge.net/lists/listinfo/lxr-developer

Re: [Lxr-developer] Code documentation

[Neil S. <ns...@ci...>]

Hello,

The guys here: http://unstable.elemental.com/mozilla/ had the idea of cro=
ss
linking between doxygen and LXR. I think this is a good idea. Currently I=
'm
using doxygen to document a (fairly large) source tree, and I plan to use
LXR on the same tree. I'd say it'd be pretty useful to be able to jump fr=
om,
say, a classname in LXR to the doxygen page for the class - and vice vers=
a.
It could all be done manually by the user just now, but it might be worth
thinking about as potential future feature for incorporation into the
software.

Just my =A30.01.

Cheers,

Neil.

----- Original Message -----
From: "Malcolm Box" <ma...@br...>
To: <lxr...@li...>
Sent: Tuesday, June 12, 2001 3:03 AM
Subject: Re: [Lxr-developer] Code documentation


> Hi Zubin,
>
> That would be excellent!  I've used Doxygen before, and it's a great
> tool.  I wasn't aware that it supported Perl, but if it does then let's
> go for it.
>
> As far as I'm concerned, any patch that improves comments or
> documentation in any way will always be gladly accepted.
>
> Cheers,
>
> Malcolm
>
> Sethna, Zubin (RSA, Brisbane) wrote:
>
> >Hi folks
> >
> >How to people feel about adding Doxygen comments to the LXR source cod=
e?
> >Doxygen generates html pages from the comments you insert into the sou=
rce
> >code. Just a header to each function describing what it does and its
> >input/output? I would like to start this if I get the go ahead from u
all.
> >
> >I'm a newbie to perl and I think it would ease the learning curve for
people
> >like me who wish to contribute code to this project.
> >
> >
> >Regards
> >
> >
> >
> >Zubin
> >
> >
> >_______________________________________________
> >Lxr-developer mailing list
> >Lxr...@li...
> >http://lists.sourceforge.net/lists/listinfo/lxr-developer
> >
>
>
>
> _______________________________________________
> Lxr-developer mailing list
> Lxr...@li...
> http://lists.sourceforge.net/lists/listinfo/lxr-developer

RE: [Lxr-developer] Code documentation

[Sethna, Z. (R. Brisbane) <ZS...@rs...>]

I don't think it directly supports perl but I think it can be tweaked to be
useable. Will try this unless someone can suggest a better/more perl
friendly documentation tool.

Zubin

-----Original Message-----
From: Malcolm Box [mailto:ma...@br...]
Sent: Tuesday, 12 June 2001 12:04 PM
To: lxr...@li...
Subject: Re: [Lxr-developer] Code documentation


Hi Zubin,

That would be excellent!  I've used Doxygen before, and it's a great 
tool.  I wasn't aware that it supported Perl, but if it does then let's 
go for it.

As far as I'm concerned, any patch that improves comments or 
documentation in any way will always be gladly accepted.

Cheers,

Malcolm

Sethna, Zubin (RSA, Brisbane) wrote:

>Hi folks
>
>How to people feel about adding Doxygen comments to the LXR source code?
>Doxygen generates html pages from the comments you insert into the source
>code. Just a header to each function describing what it does and its
>input/output? I would like to start this if I get the go ahead from u all.
>
>I'm a newbie to perl and I think it would ease the learning curve for
people
>like me who wish to contribute code to this project.
>
>
>Regards
>
>
>
>Zubin
>
>
>_______________________________________________
>Lxr-developer mailing list
>Lxr...@li...
>http://lists.sourceforge.net/lists/listinfo/lxr-developer
>



_______________________________________________
Lxr-developer mailing list
Lxr...@li...
http://lists.sourceforge.net/lists/listinfo/lxr-developer

Re: [Lxr-developer] Code documentation

[Malcolm B. <ma...@br...>]

Hi Zubin,

That would be excellent!  I've used Doxygen before, and it's a great 
tool.  I wasn't aware that it supported Perl, but if it does then let's 
go for it.

As far as I'm concerned, any patch that improves comments or 
documentation in any way will always be gladly accepted.

Cheers,

Malcolm

Sethna, Zubin (RSA, Brisbane) wrote:

>Hi folks
>
>How to people feel about adding Doxygen comments to the LXR source code?
>Doxygen generates html pages from the comments you insert into the source
>code. Just a header to each function describing what it does and its
>input/output? I would like to start this if I get the go ahead from u all.
>
>I'm a newbie to perl and I think it would ease the learning curve for people
>like me who wish to contribute code to this project.
>
>
>Regards
>
>
>
>Zubin
>
>
>_______________________________________________
>Lxr-developer mailing list
>Lxr...@li...
>http://lists.sourceforge.net/lists/listinfo/lxr-developer
>

[Lxr-developer] Code documentation

[Sethna, Z. (R. Brisbane) <ZS...@rs...>]

Hi folks

How to people feel about adding Doxygen comments to the LXR source code?
Doxygen generates html pages from the comments you insert into the source
code. Just a header to each function describing what it does and its
input/output? I would like to start this if I get the go ahead from u all.

I'm a newbie to perl and I think it would ease the learning curve for people
like me who wish to contribute code to this project.


Regards



Zubin

[Lxr-developer] greetings to the list, changes to the previous lxr (0.3), and what I'm up to w.r.t. lxr

[Rusty C. <rca...@te...>]

Well, I mistakenly sent the following to the old lxr-interest list
on lister.ping.uio.no, and having been redirected here thought
I'd say it here again... Sorry if you've already seen it over
on the other list!

<<<begin forwarded junk>>>>

Greetings, lxr-interest list!
 
(Sorry for the really long 'Subject:'!)
 
I joined a day or 2 ago and have seen nothing come along,
so figured I'd poke my head up and say "HI!".  
 
<insert cute graphic here>  Hi!   ;-)
 
Anyway, on to more serious matters:
 
I've taken the previous version of lxr and made it work on
solaris using Netscape Commerce Server, if anyone cares ;-)
 
I joined the list to ask if anyone wanted the patches,
then I found out that lxr has changed since my changes, so
I suppose its 'never mind' time :-)
 
 
I also fixed it so that enums, structs, and unions were
correctly identified.  The line in genxref that needs 
changing looks like:
 
        $contents =~ s/((struct|enum|union)\s+($ident|)\s*({}|(;)))/
 
and needs to say:
 
 
        $contents =~ s/((struct|enum|union)\s+($ident)\s*({}|(;)))/
 
 
instead.  I'm working on making it correctly identify the 
identifiers inside the enum, since the below is perfectly legal
c/c++ syntax (but not parsed at all (it appears that everything
inside {} is thrown out - oh, well ;-) )
 
typedef enum {
        ENUM1 = 2,
        ENUM2 = 3
} an_enum;
 
With the above change it correctly highlights 'an_enum' but not 
"ENUM1" or "ENUM2".  As I say, I'm working on that, at the same
time that I'm going to try out the new lxr (from cvs last night)....
 
Anyway, Hello again to all, and now you know what I'm up to,
for what its worth...
 
rc
 
<<<end>>>

Now that I've discovered its at sourceforge, I'll get the latest
CVS from there (and/or maybe the stable) and try again...  Maybe I won't 
have to work on the enum thing ;-)  (but I'm sure I'll end up wanting to add
SOMETHING to it ;-)

rc


Rusty Carruth          Email:     rca...@Te... or rca...@sl...
Voice: (480) 345-3621  SnailMail: Schlumberger ATE
FAX:   (480) 345-8793             7855 S. River Parkway, Suite 116
Ham: N7IKQ @ 146.82+,pl 162.2     Tempe, AZ 85284-1825
ICBM: 33 20' 44"N   111 53' 47"W

[Lxr-developer] RE: Lxr-developer digest, Vol 1 #6 - 2 msgs

[Brook, A. <and...@mc...>]

> >When looking throught he Mozilla Java (bonsai/webtools I think) source I
> see
> >it has hyperlinked identifiers, eg Class A extends B, with B being linked
> to
> >the defining class, saving much manual work to get there.  Now I did post
> a
> >question about 6months back asking if this was a post LXR transformation
> or
> >was part of LXR itself.  I didnt get a response.  Can anyone shed any
> light
> >on this feature?
> >
> Do you mean like line 39 of 
> http://lxr.mozilla.org/grendel/source/grendel/sources/grendel/integrator/S
> toreCtx.java 
> ?
> 
> If so, I think the current (0.7) codebase should implement this, apart 
> from the fact that Java browsing is completely fubared at the moment.  
> I'm working on fixing this, but it's proving to be a long job cleaning 
> up the Perl to make this work right.  As far as I can tell, the actual 
> indexing and cross-referencing is working fine, so it's just a matter of 
> getting it to be displayed.
> 
> Any help you could give on this would of course be much appreciated.
> 
	Thanks malcom, I'll go get the 0.7 code and see whats what, Im OK
with scripts but a novice at
	PERL if thats where the problems are, should be a fun learning
experience :)

	ttfn
	andy.

Re: [Lxr-developer] Java identifier HTML linking

[Malcolm B. <ma...@br...>]

Hi Andy,

Brook, Andy wrote:

>When looking throught he Mozilla Java (bonsai/webtools I think) source I see
>it has hyperlinked identifiers, eg Class A extends B, with B being linked to
>the defining class, saving much manual work to get there.  Now I did post a
>question about 6months back asking if this was a post LXR transformation or
>was part of LXR itself.  I didnt get a response.  Can anyone shed any light
>on this feature?
>
Do you mean like line 39 of 
http://lxr.mozilla.org/grendel/source/grendel/sources/grendel/integrator/StoreCtx.java 
?

If so, I think the current (0.7) codebase should implement this, apart 
from the fact that Java browsing is completely fubared at the moment.  
I'm working on fixing this, but it's proving to be a long job cleaning 
up the Perl to make this work right.  As far as I can tell, the actual 
indexing and cross-referencing is working fine, so it's just a matter of 
getting it to be displayed.

Any help you could give on this would of course be much appreciated.

Cheers,

Malcolm

[Lxr-developer] Java identifier HTML linking

[Brook, A. <and...@mc...>]

Hi all,

When looking throught he Mozilla Java (bonsai/webtools I think) source I see
it has hyperlinked identifiers, eg Class A extends B, with B being linked to
the defining class, saving much manual work to get there.  Now I did post a
question about 6months back asking if this was a post LXR transformation or
was part of LXR itself.  I didnt get a response.  Can anyone shed any light
on this feature?

If this feature doesnt already exist it would be v.nice to have it.  

Cheers,
andy.

import com.mh.theOpinionsExpressedAreMyOwnNotThatOfMyCompany;
 

RE: [Lxr-developer] LXR transition news

[Per K. G. <pk...@ne...>]

>Shows what I know about the code!  I've just been poking through the CVS 
>files, and it appears that there is already a dbm backend called DB.pm 
>in lib/LXR/Index/DB.pm

>However, I haven't tested this - does anyone else out there use it or 
>know if it works?

I wrote the code quick and dirty to mimic the stuff in DBI.pm, I can
almost guarantee that it does not work. It is not tested at all, but
it is a start. 


                                                   Per Kristian

Re: [Lxr-developer] LXR transition news

[Malcolm B. <ma...@br...>]

Neil Salter wrote:
<snip re Dbm backend>
 > Malcolm Box wrote:

>>Of course, if you wanted to contribute a dbm backend then I'd be happy
>>to add it to the project :-)
>>
Shows what I know about the code!  I've just been poking through the CVS 
files, and it appears that there is already a dbm backend called DB.pm 
in lib/LXR/Index/DB.pm

However, I haven't tested this - does anyone else out there use it or 
know if it works?

Cheers,

Malcolm

Re: [Lxr-developer] LXR transition news

[Neil S. <ns...@ci...>]

Hi Malcom,

Thanks for the reponse.

> There are no plans to continue with this.  The new codebase makes
> extensive use of relational dbs to speed up the browsing.  It's pretty
> simple to set up either MySQL or Postgres, so it shouldn't be a big
> deterant.

Fair enough :)

> Of course, if you wanted to contribute a dbm backend then I'd be happy
> to add it to the project :-)

I'm afraid I don't even have time to properly contribute to the main project
I help out on just now :(  - also on sourceforge. But I hope to use LXR for
something at work, so it may well get tested.

> P.S.  In future, can you followup to the list, so others can see the
> discussion?

No problem, I just hit reply assuming it would do this. It was only after
I'd
sent the message that I noticed that it didn't. I think I've seen this
happen
before. I think there's an option on sourceforge to allow you to set things
up
so that replying to a message from the list will, by default, go to the
list.

Cheers,

Neil.

P.S.    I like your domain name, do you live in Newcastle by any chance?

----- Original Message -----
From: "Malcolm Box" <ma...@br...>
To: "Neil Salter" <ns...@ci...>
Sent: Tuesday, May 22, 2001 12:59 PM
Subject: Re: [Lxr-developer] LXR transition news


> Hi Neil,
>
> Neil Salter wrote:
> >
> > Just a thought... the LXR I currently have (0.3) does not require MySQL
or
> > any other DB. It uses dbm files. This makes it relatively simple to set
up,
> > i.e. download LXR, tweak one or two things, and go. Are there any plans
to
> > continue this facility, or will LXR users need to put in place a DB?
>
> There are no plans to continue with this.  The new codebase makes
> extensive use of relational dbs to speed up the browsing.  It's pretty
> simple to set up either MySQL or Postgres, so it shouldn't be a big
> deterant.
>
> Of course, if you wanted to contribute a dbm backend then I'd be happy
> to add it to the project :-)
>
> Cheers,
>
> Malcolm
>
> P.S.  In future, can you followup to the list, so others can see the
> discussion?

Re: [Lxr-developer] CVS archive now online

[Malcolm B. <ma...@br...>]

Jens-Uwe Mager wrote:
> 
> On Tue, May 22, 2001 at 04:39:48PM +0900, Malcolm Box wrote:
> 
> > Just a quick note to let you all know that the CVS archive has now
> > appeared on SF, many thanks to Peder for organising the transfer.
> 
> Does anybody know a way to switch over an existing checkout to the new
> repository?

Your best bet is to go through the CVS directories and edit the
Repository file to point to the new server.  You may also need to edit
other files in the CVS dirs - poke about for ones that refer to the old
server and change as necessary.

I've used this technique in the past successfully.

Cheers,

Malcolm

Re: [Lxr-developer] CVS archive now online

[Jens-Uwe M. <ju...@he...>]

On Tue, May 22, 2001 at 04:39:48PM +0900, Malcolm Box wrote:

> Just a quick note to let you all know that the CVS archive has now 
> appeared on SF, many thanks to Peder for organising the transfer.

Does anybody know a way to switch over an existing checkout to the new
repository?

-- 
Jens-Uwe Mager

HELIOS Software GmbH
Steinriede 3
30827 Garbsen
Germany

Phone:		+49 5131 709320
FAX:		+49 5131 709325
Internet:	ju...@he...

[Lxr-developer] CVS archive now online

[Malcolm B. <ma...@br...>]

Just a quick note to let you all know that the CVS archive has now 
appeared on SF, many thanks to Peder for organising the transfer.

Time to get hacking :-)

Malcolm

[Lxr-developer] LXR transition news

[Malcolm B. <ma...@br...>]

Hi all,

I see from the subscription list that several people have moved over
from the old LXR list to this new list and I have to say I'm heartened :-)

The new setup is coming along nicely.  The CVS tree has been sent over
to Sourceforge, and hopefully the admins there will have it up within a
day or two.  I've made an interim release 0.7 of the current lxr code
base so that there is a visible place for people to get this new lxr.

We also have some more project developers/admins, currently Per Kristian
and Peder have joined the new project, bring with them much needed
expertise on the lxr code!

I thought I'd outline some of the things I want to be working on over
the next few months (other than the transition of course).

The first one of these is to improve the interface of lxr to the
exuberant ctags (ectags) program.  This is used to find the indentifiers
in the source code.  Currently there are several problems with the
interface:

1) lxr doesn't work properly with ectags 5.0, the latest release last time
I looked
2) Because C++ & Java are essentially supported by C.pm, the tagging is
wrong sometimes (e.g. C++ reserved words are indexed)
3) lxr can't make use of all the other languages that ectags supports easily

My plan for fixing this goes something like this:

First, create a new tagger class called ectags which is a generic
interface to ectags - ie it will correctly index any file that ectags
can parse.  Specifically, this would implement indexfile().
Create a "meta-language" class that derives from ectags and is driven by
data file for each language we wish to support.   This would implement
referencefile(). Specifically, the SimpleParse spec and reserved words
should be taken from a data file for each language.  Then basic support
for any language can be provided by the metalang class once some simple
configuration data has been provided.
If specific languages have more complex needs then of course new lang
subclasses can still be defined and used as before.
Finally, replace the hardwired pathname <-> lang subclass mapping in
Lang.pm with a data file, so that no code has to be changed either to
decide which files are what type (ie recognise .c++ as C++ or .j as
Java) or to change which class is used to index which type.

So then the adding of basic support for languages supported by ectags
will consist of writing a few regexps in a data file (which if we're
really cunning we can extract from ectags itself - serious bonus points
for this!).

So that's one part.  What other things do I want to see improved?

1) The mysql db backend.  At the moment the indexes and queries on mysql
are not optimal when generating the browsing output.  Expect patches
soon :-)
2) I have heard reports that the postgresql backend is broken - this
needs fixing
3) I've implemented locally the ability to suppress reference display
when doing an indent query - this makes it *much* faster for the common
case of looking for the definition of something, since generating the
references for a common identifier can take 90+seconds on my database.
This may be worth putting in.
4) On a more bluesky note, I'd like to improve the browsing interface
generally by making better use of CSS and HTML 4.0 to support more
configurable fontification and perhaps mouse/sidebar interaction.
Integrating this with mozilla's XUL/RDF support could result in a killer
source browser.
5) And finally, I want to try to understand the demonic perl that
generates the templates for browsing :-)

So I guess these are my initial thoughts on directions to go in and what
I'm likely to be working on in the coming months, and some of the
directions I'd like to take the project.  If anyone else has things
they'd like to see or hack on, then it would be good to hear about it on
this list.

Cheers,

Malcolm