genex-dev Mailing List for GeneX Gene Expression Database (Page 107)

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

"Michael Pear" <mic...@ho...> writes:

> Hi Jiaye,
> 
>  CVS where? I don't think I have access to this, do I?

You probably don't, Michael. Harry can set you up if you call him.

jas



"Jason E. Stewart" <ja...@op...> writes:

> Hey Todd,
> 
> Short term to ensure things are working on genex.ncgr.org ASAP:
> * I will modify the version strings on the DTD's in the installation so
>   they appear to be the version you're expecting

This is fixed. Someone please try the curation tool and let me know if
there's any problems. I'm sorry that I didn't realize sooner this was
broken, too much email on a day that I was already frantically busy...

jas.




> Good question. The versions changed for two reasons (neither of which
> are all that convincing):
> * The DTD's were moved from my CVS tree on my local machine to a new
>   CVS repository (Harry's machine), in a 'CVS-friendly' way that
>   unfortunately reset all version numbers to 1.1
> * Many comments were modified in the files to change all uses of GEML
>   to be GeneXML

ahhhh...that explains it.

>
> Now as to how to fix this. Long term, here are my suggestions:
> 1) The version numbers are automatically assigned by CVS, and until we
>    get everything into it's permanent home on SourceForge, these may
>    change again, and as such, are not stable

eww. ok, have to deal with it.

> 2) The dates, while also assigned by CVS are more stable, as they will
>    at least be guaranteed to increase no matter what archive they are
>    in.

unless the date/time is changed.

> 3) I would recommend that the CT uses an approach similar to that of
>    xml2db, which is to accept a list of values and not just a single
>    value. This means that you could simply add to the list when a
>    minor version change takes place (such as the comment changes
>    above).
> 4) I'm not sure whether it's possible for you to have these in some
>    external configuration file instead of in the java source. That way
>    you could modify the configuration file and not need to recompile
>    any .class files

could be put in properties file.

>
> Short term to ensure things are working on genex.ncgr.org ASAP:
> * I will modify the version strings on the DTD's in the installation so
>   they appear to be the version you're expecting

or i could make the mismatch less strict than it is now by just writing a
message to the console window instead of throwing an exception and stopping.
That's what I will do for now since versions might be messed up for a while.
When we settle things down, versions will be checked again somehow (most
likely not the way they are now).

>
> jas.
>
> "Todd Peterson" <tf...@nc...> writes:
>
> > When did versions of DTDs change? Here is what CT is expecting (direct
from
> > GEMLConstants.java):
> >
> >  public static final String EXPECTED_GEML_VERSION = "1.15";
> >  public static final String EXPECTED_USF_VERSION = "1.7";
> >  public static final String EXPECTED_CSF_VERSION = "1.4";
> >  public static final String EXPECTED_AMS_VERSION = "1.7";
> >  public static final String EXPECTED_ALS_VERSION = "1.5";
> >  public static final String EXPECTED_GEML_DATE = "2000/10/31";
> >  public static final String EXPECTED_USF_DATE = "2000/10/17";
> >  public static final String EXPECTED_CSF_DATE = "2000/10/17";
> >  public static final String EXPECTED_AMS_DATE = "2000/10/17";
> >  public static final String EXPECTED_ALS_DATE = "2000/10/17";
> >
> > here's what I get from downloaded control bundle:
> >
> > <GENEXML cvs_id="$Id: genexml.dtd,v 1.1 2001/01/03 17:38:09 jes Exp $"
> > date="2001/01/03" version="1.1">
> >
> > <USF cvs_id="$Id: usf.dtd,v 1.1 2001/01/03 17:38:09 jes Exp $"
> >         date="2001/01/03"
> >         version="1.1">
> >
> > <CSF cvs_id="$Id: csf.dtd,v 1.1 2001/01/03 17:38:09 jes Exp $"
> >         date="2001/01/03"
> >         version="1.1">
> >
> > <ALS cvs_id="$Id: als.dtd,v 1.1 2001/01/03 17:38:09 jes Exp $"
> >         date="2001/01/03"
> >         version="1.1">
> >
> > Todd Peterson
> > NCGR
> > mailto:tf...@nc...
> > (505) 995-4417
>




Hi Jiaye,

 CVS where? I don't think I have access to this, do I?

Regards,

Michael Pear

----- Original Message ----- 
From: "Jiaye Zhou" <ze...@in...>
To: <gen...@li...>; "Michael Pear" <mic...@ho...>
Sent: Friday, January 26, 2001 12:03 PM
Subject: Re: [GeneX-dev] xml2db.pl errors...


> 
> Michael,
> 
>   The fixed version is in CVS now. Thanks again for point this out.
> 
> jiaye
> 
> Quoting Michael Pear <mic...@ho...>:
> 
> > On the key removal for putative_csf_fk and nonunique_flag,
> > I consider what I did as a temporary kludge. I hope there is
> > a better permanent solution....Jiaye knows better than I how
> > these keys are set, from XML, programmatically, etc. and what
> > needs to be done to avoid their original creation.
> > 
> > Regards,
> 
> _______________________________________________
> Genex-dev mailing list
> Gen...@li...
> http://lists.sourceforge.net/lists/listinfo/genex-dev
> 




Hey Todd,

Good question. The versions changed for two reasons (neither of which
are all that convincing):
* The DTD's were moved from my CVS tree on my local machine to a new
  CVS repository (Harry's machine), in a 'CVS-friendly' way that
  unfortunately reset all version numbers to 1.1
* Many comments were modified in the files to change all uses of GEML
  to be GeneXML

Now as to how to fix this. Long term, here are my suggestions:
1) The version numbers are automatically assigned by CVS, and until we
   get everything into it's permanent home on SourceForge, these may
   change again, and as such, are not stable
2) The dates, while also assigned by CVS are more stable, as they will
   at least be guaranteed to increase no matter what archive they are
   in. 
3) I would recommend that the CT uses an approach similar to that of
   xml2db, which is to accept a list of values and not just a single
   value. This means that you could simply add to the list when a
   minor version change takes place (such as the comment changes
   above). 
4) I'm not sure whether it's possible for you to have these in some
   external configuration file instead of in the java source. That way
   you could modify the configuration file and not need to recompile
   any .class files

Short term to ensure things are working on genex.ncgr.org ASAP:
* I will modify the version strings on the DTD's in the installation so
  they appear to be the version you're expecting

jas.

"Todd Peterson" <tf...@nc...> writes:

> When did versions of DTDs change? Here is what CT is expecting (direct from
> GEMLConstants.java):
> 
>  public static final String EXPECTED_GEML_VERSION = "1.15";
>  public static final String EXPECTED_USF_VERSION = "1.7";
>  public static final String EXPECTED_CSF_VERSION = "1.4";
>  public static final String EXPECTED_AMS_VERSION = "1.7";
>  public static final String EXPECTED_ALS_VERSION = "1.5";
>  public static final String EXPECTED_GEML_DATE = "2000/10/31";
>  public static final String EXPECTED_USF_DATE = "2000/10/17";
>  public static final String EXPECTED_CSF_DATE = "2000/10/17";
>  public static final String EXPECTED_AMS_DATE = "2000/10/17";
>  public static final String EXPECTED_ALS_DATE = "2000/10/17";
> 
> here's what I get from downloaded control bundle:
> 
> <GENEXML cvs_id="$Id: genexml.dtd,v 1.1 2001/01/03 17:38:09 jes Exp $"
> date="2001/01/03" version="1.1">
> 
> <USF cvs_id="$Id: usf.dtd,v 1.1 2001/01/03 17:38:09 jes Exp $"
>         date="2001/01/03"
>         version="1.1">
> 
> <CSF cvs_id="$Id: csf.dtd,v 1.1 2001/01/03 17:38:09 jes Exp $"
>         date="2001/01/03"
>         version="1.1">
> 
> <ALS cvs_id="$Id: als.dtd,v 1.1 2001/01/03 17:38:09 jes Exp $"
>         date="2001/01/03"
>         version="1.1">
> 
> Todd Peterson
> NCGR
> mailto:tf...@nc...
> (505) 995-4417



"Jiaye Zhou" <ze...@in...> writes:

> Hi michael,
> 
> Quoting Michael Pear <mic...@ho...>:
> 
> > Found the following errors so far with xml2db.pl
> > 
> > 1) Stopped on a contact "genex" that was in the initial database (not
> > added by
> > me) when looking up the record. 
> > DBD::Pg::st execute failed: ERROR:  parser: parse error at or near ""
> > 7Contact: SELECT con_pk from Contact where contact_person = 'genex' AND 
> >  at /var/genex/local/bin/xml2db.pl line 1656.
> 
> Hmmmm, why is genex in the contact info? -U option is only used for
> initializing the connection string. The above error occured when the
> program is checking for existing data in the database.

I knew this would happen, but I couldn't think of a better name.

There are two uses of the name 'genex':

1) A postgres user, entered into the pg_shadow table. This user is
   given SELECT, INSERT, and DELETE privelege to all tables in the
   genex DB dump

2) An entry in the UserSec table. This entry is listed as owning all
   experiment sets, arraymeasurements and USF entries in the public DB
   dump. It also has a corresponding entry in the Contact table.

They have nothing to do with one another, except (confusingly) they
have the same name.

The error your seeing is with the UserSec and Contact entries.

I hope it's a bit clearer, and I would *love* suggestions as to how to
make this clearer in the future.

jas.



Michael,

  The fixed version is in CVS now. Thanks again for point this out.

jiaye

Quoting Michael Pear <mic...@ho...>:

> On the key removal for putative_csf_fk and nonunique_flag,
> I consider what I did as a temporary kludge. I hope there is
> a better permanent solution....Jiaye knows better than I how
> these keys are set, from XML, programmatically, etc. and what
> needs to be done to avoid their original creation.
> 
> Regards,



In the previous message I suggested removing line 900-910. Actually, you might 
want to keep the line that removes putative_csf_id attribute in case it shows 
up, unless the CT XML output no longer contains it.

jiaye



Hi michael,

Quoting Michael Pear <mic...@ho...>:

> Found the following errors so far with xml2db.pl
> 
> 1) Stopped on a contact "genex" that was in the initial database (not
> added by
> me) when looking up the record. 
> DBD::Pg::st execute failed: ERROR:  parser: parse error at or near ""
> 7Contact: SELECT con_pk from Contact where contact_person = 'genex' AND 
>  at /var/genex/local/bin/xml2db.pl line 1656.

Hmmmm, why is genex in the contact info? -U option is only used for 
initializing the connection string. The above error occured when the program is 
checking for existing data in the database.


> 2)  XML Parser error on a special character in the USF file.
> Known problem from before. I'm leaving it in the data I'm working with
> so that we can reproduce it in the future. 
> 
> Response: Edited non-ascii char out and went on.

I should add this to the realease note or a README file.


> 3) Problem due to missing field related to CSF removal and other schema
> changes
> 
> 
> DBD::Pg::db do failed: ERROR:  Relation 'usersequencefeature' does not
> have attribute 'putative_csf_fk'
> 7UserSequenceFeature: INSERT INTO UserSequenceFeature (usf_name,
> putative_csf_fk, nonunique_flag, spc_fk, short_description, owner_us_fk,
> usf_type) VALUES ('E06a', '1', '1', '50', 'bone morphogenetic protein 4 
> (BMP4) + bone morphogenetic protein 2B  (BMP2B)', '1', 'gene_name')  at
> /var/genex/local/bin/xml2db.pl line 1609.
> 
> Inserting UserSequenceFeature, this may take a few minutes...
> DBD::Pg::db do failed: ERROR:  Relation 'usersequencefeature' does not
> have attribute 'nonunique_flag'
> 7UserSequenceFeature: INSERT INTO UserSequenceFeature (usf_name,
> nonunique_flag, spc_fk, short_description, owner_us_fk, usf_type) VALUES
> ('E06a', '1', '50', 'bone morphogenetic protein 4  (BMP4) + bone
> morphogenetic protein 2B  (BMP2B)', '1', 'gene_name')  at
> /var/genex/local/bin/xml2db.pl line 1615.
> 
> response: Added 
> 
>   if ($table eq 'UserSequenceFeature'){
>     delete($$contentRef{putative_csf_fk});
>     delete($$contentRef{nonunique_flag});
>   }
> 

Oh, I see. Michael, the addition of above code can be avoided, here's why. If 
you look at line 900-910, you will see that the program is checking for the two 
attributes that don't exist anymore. Just remove these ten lines, we'll be ok. 
(this was there was because since the two attributes still existed in the 
database until early this week, with not null option. I used it as sort of a 
hack)

> at the top of pgInsert (after   my ($table, $contentRef) = @_; )
> to remove the keys.
> 
> 
> Then: Success! dataset added.


Thanks for pointing this out.

Jiaye



On the key removal for putative_csf_fk and nonunique_flag,
I consider what I did as a temporary kludge. I hope there is
a better permanent solution....Jiaye knows better than I how
these keys are set, from XML, programmatically, etc. and what
needs to be done to avoid their original creation.

Regards,

Michael Pear


> > 3) Problem due to missing field related to CSF removal and other
> >    schema changes
> >
> >
> > DBD::Pg::db do failed: ERROR:  Relation 'usersequencefeature' does not have attribute
'putative_csf_fk'
> > 7UserSequenceFeature: INSERT INTO UserSequenceFeature (usf_name, putative_csf_fk,
nonunique_flag, spc_fk, short_description, owner_us_fk, usf_type) VALUES ('E06a', '1', '1', '50',
'bone morphogenetic protein 4  (BMP4) + bone morphogenetic protein 2B  (BMP2B)', '1', 'gene_name')
at /var/genex/local/bin/xml2db.pl line 1609.
> >
> > Inserting UserSequenceFeature, this may take a few minutes...
> > DBD::Pg::db do failed: ERROR:  Relation 'usersequencefeature' does not have attribute
'nonunique_flag'
> > 7UserSequenceFeature: INSERT INTO UserSequenceFeature (usf_name, nonunique_flag, spc_fk,
short_description, owner_us_fk, usf_type) VALUES ('E06a', '1', '50', 'bone morphogenetic protein 4
(BMP4) + bone morphogenetic protein 2B  (BMP2B)', '1', 'gene_name')  at
/var/genex/local/bin/xml2db.pl line 1615.
> >
> > response: Added
> >
> >   if ($table eq 'UserSequenceFeature'){
> >     delete($$contentRef{putative_csf_fk});
> >     delete($$contentRef{nonunique_flag});
> >   }
> >
> > at the top of pgInsert (after   my ($table, $contentRef) = @_; )
> > to remove the keys.
> >
> >
> > Then: Success! dataset added.
>
> Great! I'm glad it was as simple as just removing the keys... Whew!!
> That's an enormous help Michael.
>
> jas.
>
> _______________________________________________
> Genex-dev mailing list
> Gen...@li...
> http://lists.sourceforge.net/lists/listinfo/genex-dev
>




"Michael Pear" <mic...@ho...> writes:

> Found the following errors so far with xml2db.pl
> 
> 1) Stopped on a contact "genex" that was in the initial database
>    (not added by me) when looking up the record. 
> 
> DBD::Pg::st execute failed: ERROR:  parser: parse error at or near ""
> 7Contact: SELECT con_pk from Contact where contact_person = 'genex' AND   at /var/genex/local/bin/xml2db.pl line 1656.
> 
> Response: I removed the offending <contact> </contact> from the submitted XML
> file with an editor and went on.

The 'genex' contact is used by the DB as the contact person for the
installation. It is the corresponding Contact entry for the 'genex'
UserSec entry. That UserSec entry is specified as the owner of all the
NCGR curated data in the DB. 

The problem seems to be that even though Contact:organization is
specified to be NOT NULL in the DB, the 'genex' Contact entry has a
blank organization... This will have to come out in the DB update.

> 3) Problem due to missing field related to CSF removal and other
>    schema changes
> 
> 
> DBD::Pg::db do failed: ERROR:  Relation 'usersequencefeature' does not have attribute 'putative_csf_fk'
> 7UserSequenceFeature: INSERT INTO UserSequenceFeature (usf_name, putative_csf_fk, nonunique_flag, spc_fk, short_description, owner_us_fk, usf_type) VALUES ('E06a', '1', '1', '50', 'bone morphogenetic protein 4  (BMP4) + bone morphogenetic protein 2B  (BMP2B)', '1', 'gene_name')  at /var/genex/local/bin/xml2db.pl line 1609.
> 
> Inserting UserSequenceFeature, this may take a few minutes...
> DBD::Pg::db do failed: ERROR:  Relation 'usersequencefeature' does not have attribute 'nonunique_flag'
> 7UserSequenceFeature: INSERT INTO UserSequenceFeature (usf_name, nonunique_flag, spc_fk, short_description, owner_us_fk, usf_type) VALUES ('E06a', '1', '50', 'bone morphogenetic protein 4  (BMP4) + bone morphogenetic protein 2B  (BMP2B)', '1', 'gene_name')  at /var/genex/local/bin/xml2db.pl line 1615.
> 
> response: Added 
> 
>   if ($table eq 'UserSequenceFeature'){
>     delete($$contentRef{putative_csf_fk});
>     delete($$contentRef{nonunique_flag});
>   }
> 
> at the top of pgInsert (after   my ($table, $contentRef) = @_; )
> to remove the keys.
> 
> 
> Then: Success! dataset added.

Great! I'm glad it was as simple as just removing the keys... Whew!!
That's an enormous help Michael.

jas.



Hey,

So it turns out you need to instruct MakeMaker what your prerequisites
are for it to auto fetch packages with CPAN. Luckily, it's easy, just
add a PREREQ_PM line:

   use ExtUtils::MakeMaker;
   
   WriteMakefile(
       NAME	   => 'Class::ObjectTemplate::DB',
       VERSION_FROM   => 'DB.pm',	      
       PREREQ_PM => { 'Class::ObjectTemplate' => '0.2'}
   );

These changes will wend their way into CPAN over the next few
days. I've also got them at:

http://download.sourceforge.net/openinformatics/Class-ObjectTemplate-0.4.tar.gz
http://download.sourceforge.net/openinformatics/Class-ObjectTemplate-DB-0.24.tar.gz

There are no new code changes, so *don't* upgrade (just an FYI) ..

jas.



Found the following errors so far with xml2db.pl

1) Stopped on a contact "genex" that was in the initial database (not =
added by
me) when looking up the record.=20

DBD::Pg::st execute failed: ERROR:  parser: parse error at or near ""
7Contact: SELECT con_pk from Contact where contact_person =3D 'genex' =
AND   at /var/genex/local/bin/xml2db.pl line 1656.

Response: I removed the offending <contact> </contact> from the =
submitted XML
file with an editor and went on.


2)  XML Parser error on a special character in the USF file.
Known problem from before. I'm leaving it in the data I'm working with =
so that we can reproduce it in the future.=20

Response: Edited non-ascii char out and went on.


3) Problem due to missing field related to CSF removal and other schema =
changes


DBD::Pg::db do failed: ERROR:  Relation 'usersequencefeature' does not =
have attribute 'putative_csf_fk'
7UserSequenceFeature: INSERT INTO UserSequenceFeature (usf_name, =
putative_csf_fk, nonunique_flag, spc_fk, short_description, owner_us_fk, =
usf_type) VALUES ('E06a', '1', '1', '50', 'bone morphogenetic protein 4  =
(BMP4) + bone morphogenetic protein 2B  (BMP2B)', '1', 'gene_name')  at =
/var/genex/local/bin/xml2db.pl line 1609.

Inserting UserSequenceFeature, this may take a few minutes...
DBD::Pg::db do failed: ERROR:  Relation 'usersequencefeature' does not =
have attribute 'nonunique_flag'
7UserSequenceFeature: INSERT INTO UserSequenceFeature (usf_name, =
nonunique_flag, spc_fk, short_description, owner_us_fk, usf_type) VALUES =
('E06a', '1', '50', 'bone morphogenetic protein 4  (BMP4) + bone =
morphogenetic protein 2B  (BMP2B)', '1', 'gene_name')  at =
/var/genex/local/bin/xml2db.pl line 1615.

response: Added=20

  if ($table eq 'UserSequenceFeature'){
    delete($$contentRef{putative_csf_fk});
    delete($$contentRef{nonunique_flag});
  }

at the top of pgInsert (after   my ($table, $contentRef) =3D @_; )
to remove the keys.


Then: Success! dataset added.





Hi All,

tho not extensive, I've added back a bit of functionality and lots of interface/typo/doc fixes.

-  stopped APD's Makefile from trying to remake the dendro.jar file (the one that works) and removed the requirement of jpythonc from the installation script

- removed the restrictions from the Rcluster permutation code to allow it to run any data (uploaded and from a database query) tho that will later have to be checked before runtime to check the data
size and estimate if it can be run on the system.

- rearranged the demo pages for rcluster and cybert to put them both on the same page

- added the help pages back to the rcluster demo page in addition to the nav bar.

- changed cybert forms to more generic statement about enabling vnc (which actually CAN run in a browser (!) - but needs more time to spiff it up.

- various spellos, typos


And yes, I forgot to add that I need to get the SF web page up so that there's someplace to release it FROM. sheesh.


-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...



"Harry Mangalam" <man...@ho...> writes:

> Jason, you god,
> 
> the install script is MUCH improved (at least for developers) as it
> now senses when a particular phase has been done and only rebuilds
> if it hasn't - very Makefile-like in operation - makes for a ~10X
> re-install speed when checking things!

Ah... you noticed that speed improvement, eh?? I thought I could sneak
that one in there without anyone noticing... ;-)

I changed two things:
* I stat all the .in files and there targets and only replace when the
  .in files are more recent (or if the Config.pm file is more
  recent). On a fresh install there are only the .in files, so
  everything is built, on an update only a few things..
* The old substitution code was quick to write, but slow to perform,
  loop over every .in file, loop over every line, loop over every
  subst param ==> !! n^3 !!. I changed it to be ~n^2 by only looping
  over the subst params *only if* there were any on the line. Since
  most files only have <1% of their lines substitutable, its an order
  of magnitude better.

> Thanks!  Just about finished my rcluster bashing and some interface edits.

de nada ...

jas.



Jason, you god,

the install script is MUCH improved (at least for developers) as it now senses when a particular phase has been done and only rebuilds if it hasn't - very Makefile-like in operation - makes for a ~10X
re-install speed when checking things!

Thanks!  Just about finished my rcluster bashing and some interface edits.
-- 
Cheers,
Harry

Harry J Mangalam -- (949) 856 2847 (v&f) -- hj...@nc... || man...@ho...



"Michael Pear" <mic...@ho...> writes:

>  It is working for me now, thanks for the late night effort!!!
> Sounds like you started running into problems around 2:00am...
> maybe a fresh look at things in the morning will cure some
> of them.

Yawn .... ;-) 

> The install of the Genex.pm seems to have gone as expected.
> One thing to note...db2xml.pl is not moved to the local/bin file
> through the make. It may require a separate note for the time
> being in the install README.

Yes, thanks. install-all.pl handles that in the primary install, and
it will soon leave home to go live in an xml_utils module together
with its brother, xml2db.pl

> Also, is tabs2xml.pl used anywhere?

No, not really. But I used it to convert a bunch of tab delimited text
files to XML once, and thought others might like it.

> Thanks very much for your efforts...we're on our way again.

Excellent, glad to hear it!

jas.



Curation tool won't barf if the files aren't referenced in the GeneXML
external file list.

----- Original Message -----
From: "Michael Pear" <mic...@ho...>
To: "Jiaye Zhou" <ze...@in...>
Cc: <gen...@li...>
Sent: Friday, January 26, 2001 7:03 AM
Subject: Re: [GeneX-dev] Re: CanonicalSequenceFeature is missing...


> You might want to look at the code with the idea
> of ignoring any CSF section that comes in an XML file...
> If I resubmit one of my existing experiment sets from last
> week, if nothing is there to handle it, will it fail if you remove all
> mention of it? Will it fail if I use the current version? In the
long-term,
> you'll need to consider compatibility from release to release, so this
> might be a sample preparation of how to handle the situation.
>
> My guess is as the most downstream module, you should only be
> changing this after everyone else, with the only immediate action
> to ignore any tags associated with CSF as they no longer have meaning
> to the db. This probably includes a possible reference from USF.
>
> Regards,
>
> Michael Pear
>
> ----- Original Message -----
> From: "Jiaye Zhou" <ze...@in...>
> To: <gen...@li...>; "Michael Pear"
<mic...@ho...>
> Sent: Thursday, January 25, 2001 8:01 PM
> Subject: Re: [GeneX-dev] Re: CanonicalSequenceFeature is missing...
>
>
> >
> > Hi Michael,
> >
> >   Actually, I don't think this will break the xml2db code. It will just
ignore
> > the missing top level xml elements which are CannonicalSequenceFeature.
In fact,
> > I believe that I have tested the code against files that are missing
CSF, but
> > that would be a while ago and I haven't done any development on the CSF
end.
> >   I have known of the decision on not supporting CSF, but since the
Curation
> > Tool has been outputing CSF, I never removed the code that handled it.
Perhaps
> > this is a good time for me to clean that part of the code up.
> >
> > Jiaye
> >
> > Quoting Michael Pear <mic...@ho...>:
> >
> > > I missed the Release Notes for the DB...I see them on sourceforge
> > > now.
> > >
> > > I did a quick check by doing an rgrep on the install tree and there
> > > are still a lot of references to CanonicalSequenceFeature in xml2db.pl
> > > as well, so even if the curation tool can handle it, having nothing
> > > there
> > > may break submission using xml2db.pl
> > >
> > > Regards,
> > >
> > > Michael Pear
> > >
> > > ----- Original Message -----
> > > From: "Jason E. Stewart" <ja...@op...>
> > > To: "Michael Pear" <mic...@ho...>
> > > Cc: "Jason E. Stewart" <ja...@op...>; "Jiaye Zhou"
> > > <ze...@in...>;
> > > <gen...@li...>
> > > Sent: Thursday, January 25, 2001 6:56 PM
> > > Subject: [GeneX-dev] Re: CanonicalSequenceFeature is missing...
> > >
> > >
> > > > "Michael Pear" <mic...@ho...> writes:
> > > >
> > > > > I tried to download a control bundle on my new installation, but
> > > > > db2xml.pl failed because it couldn't find
> > > > > CanonicalSequenceFeature.pm. I don't see it or anything in
> > > > > Genex/CanonicalSequenceFeature in the install tree. I no longer
find
> > > > > this table in the database either. What's up?
> > > >
> > > > This was in the Release Notes for DB v0.8.1, but for some reason, I
> > > > didn't include it in the Server install notes:
> > > >
> > > > From the release notes:
> > > >
> > > > * All tables related to Canonical Sequence Features were removed
> > > >   (CanonicalSequenceFeature, CSF_ExternalDBLink, UserCanonicalLink,
> > > >   RelatedCSF)
> > > > * the two CSF related columns in UserSequenceFeature were removed
> > > >   (putative_csf_fk, nonunique_flag)
> > > >
> > > > And it looks like *I* didn't debug things enough :-(
> > > >
> > > > CSF was supposed to go away a month ago, it has always been an
> > > > unsupported feature, and since it was never really broken, it never
> > > > made it high enough on my priority to remove. I did it last release.
> > > >
> > > > Todd and/or Lonny. Will the CT break without CSF in the control
> > > > bundle? Should I just output a blank CSF section for now?
> > > >
> > > > Mea culpa, mea culpa,
> > > > jas.
> > > >
> > > >
> > > >
> > > > _______________________________________________
> > > > Genex-dev mailing list
> > > > Gen...@li...
> > > > http://lists.sourceforge.net/lists/listinfo/genex-dev
> > >
> > >
> > > _______________________________________________
> > > Genex-dev mailing list
> > > Gen...@li...
> > > http://lists.sourceforge.net/lists/listinfo/genex-dev
> > >
>
>
> _______________________________________________
> Genex-dev mailing list
> Gen...@li...
> http://lists.sourceforge.net/lists/listinfo/genex-dev
>




Hi Jason,
 It is working for me now, thanks for the late night effort!!!
Sounds like you started running into problems around 2:00am...
maybe a fresh look at things in the morning will cure some
of them.

The install of the Genex.pm seems to have gone as expected.
One thing to note...db2xml.pl is not moved to the local/bin file
through the make. It may require a separate note for the time
being in the install README.

Also, is tabs2xml.pl used anywhere?

Thanks very much for your efforts...we're on our way again.

Regards,

Michael Pear

----- Original Message ----- 
From: "Jason E. Stewart" <ja...@op...>
To: <gen...@li...>
Sent: Friday, January 26, 2001 2:37 AM
Subject: [GeneX-dev] Bug in cb download, jar not running


> Hey Michael,
> 
> The new db2xml in Genex.pm 2.5.0 should fix all the problems you were
> seeing. When run manually it works. 
> 
> When run via the CGI scripts it runs, but the final download step
> fails (for me anyway) because the jar file is not created. Let me know
> what happens for you. I'm too tired to debug it further...
> 
> jas.
> 
> _______________________________________________
> Genex-dev mailing list
> Gen...@li...
> http://lists.sourceforge.net/lists/listinfo/genex-dev




You might want to look at the code with the idea
of ignoring any CSF section that comes in an XML file...
If I resubmit one of my existing experiment sets from last
week, if nothing is there to handle it, will it fail if you remove all 
mention of it? Will it fail if I use the current version? In the long-term,
you'll need to consider compatibility from release to release, so this
might be a sample preparation of how to handle the situation.

My guess is as the most downstream module, you should only be
changing this after everyone else, with the only immediate action
to ignore any tags associated with CSF as they no longer have meaning
to the db. This probably includes a possible reference from USF.

Regards,

Michael Pear

----- Original Message ----- 
From: "Jiaye Zhou" <ze...@in...>
To: <gen...@li...>; "Michael Pear" <mic...@ho...>
Sent: Thursday, January 25, 2001 8:01 PM
Subject: Re: [GeneX-dev] Re: CanonicalSequenceFeature is missing...


> 
> Hi Michael,
> 
>   Actually, I don't think this will break the xml2db code. It will just ignore 
> the missing top level xml elements which are CannonicalSequenceFeature. In fact, 
> I believe that I have tested the code against files that are missing CSF, but 
> that would be a while ago and I haven't done any development on the CSF end. 
>   I have known of the decision on not supporting CSF, but since the Curation 
> Tool has been outputing CSF, I never removed the code that handled it. Perhaps 
> this is a good time for me to clean that part of the code up.
> 
> Jiaye
> 
> Quoting Michael Pear <mic...@ho...>:
> 
> > I missed the Release Notes for the DB...I see them on sourceforge
> > now.
> > 
> > I did a quick check by doing an rgrep on the install tree and there
> > are still a lot of references to CanonicalSequenceFeature in xml2db.pl
> > as well, so even if the curation tool can handle it, having nothing
> > there
> > may break submission using xml2db.pl
> > 
> > Regards,
> > 
> > Michael Pear
> > 
> > ----- Original Message -----
> > From: "Jason E. Stewart" <ja...@op...>
> > To: "Michael Pear" <mic...@ho...>
> > Cc: "Jason E. Stewart" <ja...@op...>; "Jiaye Zhou"
> > <ze...@in...>;
> > <gen...@li...>
> > Sent: Thursday, January 25, 2001 6:56 PM
> > Subject: [GeneX-dev] Re: CanonicalSequenceFeature is missing...
> > 
> > 
> > > "Michael Pear" <mic...@ho...> writes:
> > >
> > > > I tried to download a control bundle on my new installation, but
> > > > db2xml.pl failed because it couldn't find
> > > > CanonicalSequenceFeature.pm. I don't see it or anything in
> > > > Genex/CanonicalSequenceFeature in the install tree. I no longer find
> > > > this table in the database either. What's up?
> > >
> > > This was in the Release Notes for DB v0.8.1, but for some reason, I
> > > didn't include it in the Server install notes:
> > >
> > > From the release notes:
> > >
> > > * All tables related to Canonical Sequence Features were removed
> > >   (CanonicalSequenceFeature, CSF_ExternalDBLink, UserCanonicalLink,
> > >   RelatedCSF)
> > > * the two CSF related columns in UserSequenceFeature were removed
> > >   (putative_csf_fk, nonunique_flag)
> > >
> > > And it looks like *I* didn't debug things enough :-(
> > >
> > > CSF was supposed to go away a month ago, it has always been an
> > > unsupported feature, and since it was never really broken, it never
> > > made it high enough on my priority to remove. I did it last release.
> > >
> > > Todd and/or Lonny. Will the CT break without CSF in the control
> > > bundle? Should I just output a blank CSF section for now?
> > >
> > > Mea culpa, mea culpa,
> > > jas.
> > >
> > >
> > >
> > > _______________________________________________
> > > Genex-dev mailing list
> > > Gen...@li...
> > > http://lists.sourceforge.net/lists/listinfo/genex-dev
> > 
> > 
> > _______________________________________________
> > Genex-dev mailing list
> > Gen...@li...
> > http://lists.sourceforge.net/lists/listinfo/genex-dev
> >




Hey Michael,

The new db2xml in Genex.pm 2.5.0 should fix all the problems you were
seeing. When run manually it works. 

When run via the CGI scripts it runs, but the final download step
fails (for me anyway) because the jar file is not created. Let me know
what happens for you. I'm too tired to debug it further...

jas.



Hey,

My previous two attempts to list a valid patch for LoginUtils.pm.in
failed. So here are both the correct display-error.pl.in
LoginUtils.pm.in 

jas.


Hey,

View the release notes:

http://sourceforge.net/project/shownotes.php?release_id=21940

Download the tarball:

http://download.sourceforge.net/genex/Bio-Genex-2.5.0.tar.gz

But whatever you do, read the README:

   Configuring
   -----------
   For Genex.pm to be useful, you must have already installed a working
   GeneX server on your system. In order for Genex.pm to properly
   configure itself, it must find the Genex/Config.pm that was installed
   with the rest of the GeneX perl modules. Since the Genex/Config.pm was
   likely installed in the private Genex lib directory, you will need to
   be sure that it is in your Perl lib path. To ensure this, include the
   GeneX lib dir in your PERL5LIB:
   
      $ export PERL5LIB=/var/lib/genex/perl5:$PERL5LIB
   
   Substitute in the proper directory for /var/lib/genex/perl5

Changes since 2.4.0:
      - db2xml.pl:
        caching of experiment sets and control-bunldes implemented
        now has --no_cache option to avoid using or setting cache
	--user/--password now optional unless --authenticate is specified
	Removed citationlink()
	Fixed aweful citation bug that never showed up, because
	   citationlink table was never populated
      - Fixed API for get_objects() by creating new method
	   get_all_objects().
      - fixed bug in check_password()
      - XMLUtils.pm.in
	Added new framework for implementing class-specific XML output. 
	Implemented output for Chromosome, and partial output for Species
	   (still needs USF for completeness).
      - DBUtils.pm.in (Repository): 
	Added new assert_* methods: 
	   assert_species
	   assert_protocol
	   assert_scanner
	   assert_software
	   assert_spotter
	   assert_array
	   assert_experiment
	check_password() no longer dies() with a bad password, instead it
	   returns undef
      - UserSequenceFeature, ExperimentSet and ArrayMeasurement now have
           provider_con_fk() and  provider_con_obj() methods
	post_process() no longer automatically derefences the con_obj()
	   from a us_obj, it first tests the definedness
      - ExternalDatabase: Fixed full name
      - list_tables.pl.in: Fixed name of HTML title


Enjoy,
jas.



Hey,

Sorry for the previous bogus patches... They worked, but the produced
bogus email addresses. Here are the corrected ones.

No need to substitute on the '@', assigning the text to a variable is
enough. Otherwise the '\' comes through and turns the email adress
into 'jason\@localhost', which is pretty useless.

Sorry,
jas.

Index: curation-tool/LoginUtil.pm.in
===================================================================
RCS file: /home/opengenex/analysis_scripts_install/curation-tool/LoginUtil.pm.in,v
retrieving revision 1.3
diff -b -B -u -r1.3 LoginUtil.pm.in

--- curation-tool/LoginUtil.pm.in	2001/01/22 22:36:53	1.3
+++ curation-tool/LoginUtil.pm.in	2001/01/26 06:53:36
@@ -23,6 +23,7 @@
 my $base_job_url = "%%CGITMPURL%%/$base_job";
 my $base_job_dir = "%%CGITMPDIR%%/$base_job";
 my $basesubmitdir = "%%GENEX_SUBMISSION_HOME%%";
+my $contact_email = "%%CONTACT_EMAIL%%";
 
 # -----------------------------------Subroutine create_checksum()
 #
@@ -392,7 +393,7 @@
 <q>$error</q><br>
 Sorry for the trouble. Please try again.<br>
 <hr>
-You can notify <a href=mailto:%%CONTACT_EMAIL%%>%%CONTACT_EMAIL%%</a> about this problem.<br></div>
+You can notify <a href=mailto:$contact_email>$contact_email</a> about this problem.<br></div>
 </body></html>
 EofHTML
 

Index: curation-tool/display-error.pl.in
===================================================================
RCS file: /home/opengenex/analysis_scripts_install/curation-tool/display-error.pl.in,v
retrieving revision 1.3
diff -b -B -u -r1.3 display-error.pl.in
--- curation-tool/display-error.pl.in	2001/01/22 22:37:47	1.3
+++ curation-tool/display-error.pl.in	2001/01/26 06:49:50
@@ -22,15 +22,15 @@
 $ENV{PATH} = "/bin:/usr/bin";
 delete @ENV{ 'IFS','CDPATH','ENV','BASH_ENV' };
 
-#print "Content-type: text/html\n\n";
-#print "\n<br>************Starting Auth-Login.pl!!!!!!!!<br>\n";
+# we need to escape any '@' for string interpolation
+my $contact_email = '%%CONTACT_EMAIL%%';
 
 # Get the CGI parameters
 # The error message is passed as the first command line argument
 @ErrorMsg = @ARGV;
 
 # And generate the error HTML page
-print <<EofHTML;
+print <<"EofHTML";
 Content-type: text/html
 
 <html><head><title>An error has occured!</title></head>
@@ -49,7 +49,7 @@
 <p>Please include this information in any bug reports.</p>
 Sorry for the trouble. Please try again.<br>
 <hr>
-You can notify <a href=mailto:%%CONTACT_EMAIL%%>%%CONTACT_EMAIL%%</a> about this problem.<br></div>
+You can notify <a href=mailto:$contact_email>$contact_email</a> about this problem.<br></div>
 </body></html>
 EofHTML
 



Hi Michael,

  Actually, I don't think this will break the xml2db code. It will just ignore 
the missing top level xml elements which are CannonicalSequenceFeature. In fact, 
I believe that I have tested the code against files that are missing CSF, but 
that would be a while ago and I haven't done any development on the CSF end. 
  I have known of the decision on not supporting CSF, but since the Curation 
Tool has been outputing CSF, I never removed the code that handled it. Perhaps 
this is a good time for me to clean that part of the code up.

Jiaye

Quoting Michael Pear <mic...@ho...>:

> I missed the Release Notes for the DB...I see them on sourceforge
> now.
> 
> I did a quick check by doing an rgrep on the install tree and there
> are still a lot of references to CanonicalSequenceFeature in xml2db.pl
> as well, so even if the curation tool can handle it, having nothing
> there
> may break submission using xml2db.pl
> 
> Regards,
> 
> Michael Pear
> 
> ----- Original Message -----
> From: "Jason E. Stewart" <ja...@op...>
> To: "Michael Pear" <mic...@ho...>
> Cc: "Jason E. Stewart" <ja...@op...>; "Jiaye Zhou"
> <ze...@in...>;
> <gen...@li...>
> Sent: Thursday, January 25, 2001 6:56 PM
> Subject: [GeneX-dev] Re: CanonicalSequenceFeature is missing...
> 
> 
> > "Michael Pear" <mic...@ho...> writes:
> >
> > > I tried to download a control bundle on my new installation, but
> > > db2xml.pl failed because it couldn't find
> > > CanonicalSequenceFeature.pm. I don't see it or anything in
> > > Genex/CanonicalSequenceFeature in the install tree. I no longer find
> > > this table in the database either. What's up?
> >
> > This was in the Release Notes for DB v0.8.1, but for some reason, I
> > didn't include it in the Server install notes:
> >
> > From the release notes:
> >
> > * All tables related to Canonical Sequence Features were removed
> >   (CanonicalSequenceFeature, CSF_ExternalDBLink, UserCanonicalLink,
> >   RelatedCSF)
> > * the two CSF related columns in UserSequenceFeature were removed
> >   (putative_csf_fk, nonunique_flag)
> >
> > And it looks like *I* didn't debug things enough :-(
> >
> > CSF was supposed to go away a month ago, it has always been an
> > unsupported feature, and since it was never really broken, it never
> > made it high enough on my priority to remove. I did it last release.
> >
> > Todd and/or Lonny. Will the CT break without CSF in the control
> > bundle? Should I just output a blank CSF section for now?
> >
> > Mea culpa, mea culpa,
> > jas.
> >
> >
> >
> > _______________________________________________
> > Genex-dev mailing list
> > Gen...@li...
> > http://lists.sourceforge.net/lists/listinfo/genex-dev
> 
> 
> _______________________________________________
> Genex-dev mailing list
> Gen...@li...
> http://lists.sourceforge.net/lists/listinfo/genex-dev
> 




2001	Jan (135)	Feb (57)	Mar (84)	Apr (43)	May (77)	Jun (51)	Jul (21)	Aug (55)	Sep (37)	Oct (56)	Nov (75)	Dec (23)
2002	Jan (32)	Feb (174)	Mar (121)	Apr (70)	May (55)	Jun (20)	Jul (23)	Aug (15)	Sep (12)	Oct (58)	Nov (203)	Dec (90)
2003	Jan (37)	Feb (15)	Mar (14)	Apr (57)	May (7)	Jun (40)	Jul (36)	Aug (1)	Sep (56)	Oct (38)	Nov (105)	Dec (2)
2004	Jan	Feb (117)	Mar (69)	Apr (160)	May (165)	Jun (35)	Jul (7)	Aug (80)	Sep (47)	Oct (23)	Nov (8)	Dec (42)
2005	Jan (19)	Feb (2)	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
2007	Jan	Feb	Mar	Apr	May	Jun (2)	Jul	Aug	Sep	Oct	Nov	Dec
2010	Jan	Feb (2)	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec

genex-dev Mailing List for GeneX Gene Expression Database (Page 107)

genex-dev — GeneX developers