Home » Eclipse Projects » Equinox » It's quiet..
| It's quiet.. [message #24760] |
Wed, 21 May 2003 19:38  |
Eclipse User |
|
|
|
Originally posted by: kduffey.marketron.com
Wow, it sure has died down in this thread. Is everyone just overly busy on
other projects, working on Equinox, or what?
Well, I am still having a hell of a time in replacing the core plugin parser
with my own routine. I have no clue how to get it running and testing it. So
I have a question. I have two computers. Would it be acceptable for me to
code on one, and put the class I create over the existing class of the other
Equinox on my 2nd machine, try running it, etc?
If this is ok, how do I "debug" my code? I use System.out.println() but I
have no clue where the output goes. I would like to see if parts of my code
get executed, or not.
Thanks.
|
|
| |
| Re: It's quiet.. [message #24830 is a reply to message #24760] |
Thu, 22 May 2003 09:12  |
| Eclipse User |
|
|
|
I've been doing things similar to Pascal's suggest and have found
Eclispe to be a great development environment. My steps are:
- Checkout out the org.eclipse.core.runtime plugin as a project
directly from CVS.
- Using import fragements and plugins, import the rest of Eclipse
from the workbench by selecting all and then deselecting
org.eclipse.core.runtime. You don't need to copy the plugins
and fragements. Now you can compile the plugin and use the
many features of the Eclipse IDE.
- To test, I launch the workspace as a Runtime Workbench. Eclipse
has a great debugger and you can put breakpoints all over your
code.
BTW - yesterday I did check in the code which split the registry
resolution, registery trimming, and extension/extension point
resolution into 3 separate files.
Keith
Kevin wrote:
> Wow, it sure has died down in this thread. Is everyone just overly busy on
> other projects, working on Equinox, or what?
>
> Well, I am still having a hell of a time in replacing the core plugin parser
> with my own routine. I have no clue how to get it running and testing it. So
> I have a question. I have two computers. Would it be acceptable for me to
> code on one, and put the class I create over the existing class of the other
> Equinox on my 2nd machine, try running it, etc?
>
> If this is ok, how do I "debug" my code? I use System.out.println() but I
> have no clue where the output goes. I would like to see if parts of my code
> get executed, or not.
>
> Thanks.
>
>
,
|
|
|
| Re: It's quiet.. [message #24864 is a reply to message #24830] |
Thu, 22 May 2003 13:22 |
| Eclipse User |
|
|
|
Thanks guys. I did exactly this Keith. The thing is, when I put my code in,
Eclipse doesn't work. Then I tried the old code, and "commented out" my
code, and I am still seeing it not work right. It just shows the splash
screen for a long time, and doesn't go away. I'll keep trying it again. I
did see a list of plugins to start the runtime workbench with, and I
unchecked the core.runtime one, so I assumed that it would use mine, which I
checked out as project? I don't know for sure though if it is using it. So,
what I did was, I took the old code, and added a few System.out.println()
calls to it, and yet I don't see the output at all. Is there some place
System.out.printl() is routed to?
Thanks.
"Keith Kimball" <keith2@us.ibm.com> wrote in message
news:baiiac$i1p$1@rogue.oti.com...
> I've been doing things similar to Pascal's suggest and have found
> Eclispe to be a great development environment. My steps are:
>
> - Checkout out the org.eclipse.core.runtime plugin as a project
> directly from CVS.
>
> - Using import fragements and plugins, import the rest of Eclipse
> from the workbench by selecting all and then deselecting
> org.eclipse.core.runtime. You don't need to copy the plugins
> and fragements. Now you can compile the plugin and use the
> many features of the Eclipse IDE.
>
> - To test, I launch the workspace as a Runtime Workbench. Eclipse
> has a great debugger and you can put breakpoints all over your
> code.
>
> BTW - yesterday I did check in the code which split the registry
> resolution, registery trimming, and extension/extension point
> resolution into 3 separate files.
>
> Keith
>
>
> Kevin wrote:
> > Wow, it sure has died down in this thread. Is everyone just overly busy
on
> > other projects, working on Equinox, or what?
> >
> > Well, I am still having a hell of a time in replacing the core plugin
parser
> > with my own routine. I have no clue how to get it running and testing
it. So
> > I have a question. I have two computers. Would it be acceptable for me
to
> > code on one, and put the class I create over the existing class of the
other
> > Equinox on my 2nd machine, try running it, etc?
> >
> > If this is ok, how do I "debug" my code? I use System.out.println() but
I
> > have no clue where the output goes. I would like to see if parts of my
code
> > get executed, or not.
> >
> > Thanks.
> >
> >
>
> ,
>
|
|
|
| Re: It's quiet.. [message #24894 is a reply to message #24760] |
Thu, 22 May 2003 13:51 |
| Eclipse User |
|
|
|
Originally posted by: dj_houghton.nospam.oti.com
Kevin,
It is interesting that you are working on replacing the XML parsing in Core.
One of the Platform/Core team's plan items for Eclipse 3.0 is to remove the
dependancy of the org.eclipse.core.runtime and org.eclipse.core.resources
plug-ins on the org.apache.xerces plug-in. We are planning to investigate
XML Pull technology and perhaps this is a good chance for us to consolidate
our efforts. These are the relavent bug reports in the Eclipse Bugzilla
database:
https://bugs.eclipse.org/bugs/show_bug.cgi?id=36112
https://bugs.eclipse.org/bugs/show_bug.cgi?id=37696
How far along are you with this work? Would you be interested in
contributing your code to Eclipse? If so, the best way is to add a patch
(the new PluginParser.java source file) to either of the above bug reports.
Thanks.
"Kevin" <kduffey@marketron.com> wrote in message
news:bah2lj$eek$1@rogue.oti.com...
> Wow, it sure has died down in this thread. Is everyone just overly busy on
> other projects, working on Equinox, or what?
>
> Well, I am still having a hell of a time in replacing the core plugin
parser
> with my own routine. I have no clue how to get it running and testing it.
So
> I have a question. I have two computers. Would it be acceptable for me to
> code on one, and put the class I create over the existing class of the
other
> Equinox on my 2nd machine, try running it, etc?
>
> If this is ok, how do I "debug" my code? I use System.out.println() but I
> have no clue where the output goes. I would like to see if parts of my
code
> get executed, or not.
>
> Thanks.
>
>
|
|
| |
| Re: It's quiet.. [message #24985 is a reply to message #24894] |
Fri, 23 May 2003 01:44 |
| Eclipse User |
|
|
|
I don't know if someone else originated it, but a few months back I proposed
it and was asked to go ahead and figure it out. I have had very little time.
However, I should say I got a little discouraged when I tried getting my
code to work and could not figure it out after several hours.
What I have done so far is basically replace the calls that start the
parsing process, and also a lot of the private methods that parse different
pieces, such as the <requires> node, the <extension > node, etc. I have all
those in place to use xml pull parsing. Where I am stuck, besides figuring
out a way to get my code to work with the platform, is figuring out how it
is the pop/pull stuff works, if it is needed (I don't think it is), and how
to properly build the PluginModel using xmlpull. I am very familiar with
xmlpull, I am using it in my work project. It is very fast, very small
library, and uses very little memory. I have found it much faster (7x) over
using xerces with SAX or DOM, and a magnitude faster than JDOM. It also uses
less memory from what I can tell than DOM and even SAX. Plus, the code is
super easy to figure out and read. If people could give up on the validation
portion of xml, I think a lot of people would find that xmlpull is much
easier and faster for most needs. You can even validate inline as you are
parsing, such that you can "require" a given node at any point to be of a
type (start or end tag) and a name. If a match is not found, an
XmlPullParserException is thrown, thus invalidating the xml file.
Anywho, I am happy to help. I have my name on at least one bug (on the cc
list) regarding this issue. I was hoping to have it working by now, but due
to time and frustration in getting the IDE to accept my code, I have not got
much more done.
Where are you (and others) at with it? Maybe we can start collaborating a
bit on this?
"DJ Houghton" <dj_houghton@nospam.oti.com> wrote in message
news:baj2nb$2da$1@rogue.oti.com...
> Kevin,
>
> It is interesting that you are working on replacing the XML parsing in
Core.
> One of the Platform/Core team's plan items for Eclipse 3.0 is to remove
the
> dependancy of the org.eclipse.core.runtime and org.eclipse.core.resources
> plug-ins on the org.apache.xerces plug-in. We are planning to investigate
> XML Pull technology and perhaps this is a good chance for us to
consolidate
> our efforts. These are the relavent bug reports in the Eclipse Bugzilla
> database:
> https://bugs.eclipse.org/bugs/show_bug.cgi?id=36112
> https://bugs.eclipse.org/bugs/show_bug.cgi?id=37696
>
> How far along are you with this work? Would you be interested in
> contributing your code to Eclipse? If so, the best way is to add a patch
> (the new PluginParser.java source file) to either of the above bug
reports.
>
> Thanks.
>
>
> "Kevin" <kduffey@marketron.com> wrote in message
> news:bah2lj$eek$1@rogue.oti.com...
> > Wow, it sure has died down in this thread. Is everyone just overly busy
on
> > other projects, working on Equinox, or what?
> >
> > Well, I am still having a hell of a time in replacing the core plugin
> parser
> > with my own routine. I have no clue how to get it running and testing
it.
> So
> > I have a question. I have two computers. Would it be acceptable for me
to
> > code on one, and put the class I create over the existing class of the
> other
> > Equinox on my 2nd machine, try running it, etc?
> >
> > If this is ok, how do I "debug" my code? I use System.out.println() but
I
> > have no clue where the output goes. I would like to see if parts of my
> code
> > get executed, or not.
> >
> > Thanks.
> >
> >
>
>
|
|
|
| Re: It's quiet.. [message #25028 is a reply to message #24985] |
Fri, 23 May 2003 14:12 |
| Eclipse User |
|
|
|
Originally posted by: jeff_mcaffer_REMOVE.ca.ibm.com
Kevin,
Sounds like you have made some progress. As others have pointed out,
getting your new code to run is something directly supported by PDE. I
strongly suggest flipping the doc open to that page and saving yourself any
further stress. They do a way better job of describing what to do than
anyone here could do.
As for the approach to parsing, here is what I would expect.
- completely delete the original PluginParser class
- implement a new one based on XMLpull which sucks XML and spews
PluginDescriptor(Model?)s
- the DTD in the doc is very simple so it should be clear which Plugin
related objects correspond to which DTD elements.
- just do the simple thing and populate the relevant objects with the
related elements from the XML.
- stuff this all into the same form the is currently returned from the
parser
I recommend not making this problem too complex. You are just parsing XML
and building a representative data structure.
If all else fails, attach your new parser class to one of the bug reports
mentioned by DJ
Jeff
"Kevin" <supreme_java_guru_1@yahoo.com> wrote in message
news:bakckv$ucv$1@rogue.oti.com...
> I don't know if someone else originated it, but a few months back I
proposed
> it and was asked to go ahead and figure it out. I have had very little
time.
> However, I should say I got a little discouraged when I tried getting my
> code to work and could not figure it out after several hours.
>
> What I have done so far is basically replace the calls that start the
> parsing process, and also a lot of the private methods that parse
different
> pieces, such as the <requires> node, the <extension > node, etc. I have
all
> those in place to use xml pull parsing. Where I am stuck, besides figuring
> out a way to get my code to work with the platform, is figuring out how it
> is the pop/pull stuff works, if it is needed (I don't think it is), and
how
> to properly build the PluginModel using xmlpull. I am very familiar with
> xmlpull, I am using it in my work project. It is very fast, very small
> library, and uses very little memory. I have found it much faster (7x)
over
> using xerces with SAX or DOM, and a magnitude faster than JDOM. It also
uses
> less memory from what I can tell than DOM and even SAX. Plus, the code is
> super easy to figure out and read. If people could give up on the
validation
> portion of xml, I think a lot of people would find that xmlpull is much
> easier and faster for most needs. You can even validate inline as you are
> parsing, such that you can "require" a given node at any point to be of a
> type (start or end tag) and a name. If a match is not found, an
> XmlPullParserException is thrown, thus invalidating the xml file.
>
> Anywho, I am happy to help. I have my name on at least one bug (on the cc
> list) regarding this issue. I was hoping to have it working by now, but
due
> to time and frustration in getting the IDE to accept my code, I have not
got
> much more done.
>
> Where are you (and others) at with it? Maybe we can start collaborating a
> bit on this?
>
> "DJ Houghton" <dj_houghton@nospam.oti.com> wrote in message
> news:baj2nb$2da$1@rogue.oti.com...
> > Kevin,
> >
> > It is interesting that you are working on replacing the XML parsing in
> Core.
> > One of the Platform/Core team's plan items for Eclipse 3.0 is to remove
> the
> > dependancy of the org.eclipse.core.runtime and
org.eclipse.core.resources
> > plug-ins on the org.apache.xerces plug-in. We are planning to
investigate
> > XML Pull technology and perhaps this is a good chance for us to
> consolidate
> > our efforts. These are the relavent bug reports in the Eclipse Bugzilla
> > database:
> > https://bugs.eclipse.org/bugs/show_bug.cgi?id=36112
> > https://bugs.eclipse.org/bugs/show_bug.cgi?id=37696
> >
> > How far along are you with this work? Would you be interested in
> > contributing your code to Eclipse? If so, the best way is to add a patch
> > (the new PluginParser.java source file) to either of the above bug
> reports.
> >
> > Thanks.
> >
> >
> > "Kevin" <kduffey@marketron.com> wrote in message
> > news:bah2lj$eek$1@rogue.oti.com...
> > > Wow, it sure has died down in this thread. Is everyone just overly
busy
> on
> > > other projects, working on Equinox, or what?
> > >
> > > Well, I am still having a hell of a time in replacing the core plugin
> > parser
> > > with my own routine. I have no clue how to get it running and testing
> it.
> > So
> > > I have a question. I have two computers. Would it be acceptable for me
> to
> > > code on one, and put the class I create over the existing class of the
> > other
> > > Equinox on my 2nd machine, try running it, etc?
> > >
> > > If this is ok, how do I "debug" my code? I use System.out.println()
but
> I
> > > have no clue where the output goes. I would like to see if parts of my
> > code
> > > get executed, or not.
> > >
> > > Thanks.
> > >
> > >
> >
> >
>
>
|
|
|
Goto Forum:
Current Time: Tue Nov 04 02:36:31 EST 2025
Powered by FUDForum. Page generated in 0.35093 seconds |
Home