[DAS] 1.0 spec questions

Brian Gilman gilmanb@genome.wi.mit.edu
Sat, 2 Feb 2002 10:49:34 -0500 (EST)


Hello Andrew,

	If you look at our source tree in omnigene:

	omnigene.sourceforge.net
	
	We have a full working python implementation of das with "jaxb" in
there. Can we contribute this back to you??

				-B

-----------------------
Brian Gilman <gilmanb@genome.wi.mit.edu>
Sr. Software Engineer MIT/Whitehead Inst. Center for Genome Research
One Kendall Square, Bldg. 300 / Cambridge, MA 02139-1561 USA
phone +1 617  252 1069 / fax +1 617 252 1902


On Sat, 2 Feb 2002, Andrew Dalke wrote:

> Hi,
> 
> Let me start by saying that I haven't looked at DAS in a while
> nor have I followed the list.  I searched the list for my
> questions but didn't find anything appropriate.
> 
> So I'm starting to get a feel for DAS at the data interchange
> level.  I'm using a DTD to Python class generator called 'PAXB'
> (it's based on JAXB but is 0.1 code and it shows).  It looks
> like there are differences between
> 
> http://www.biodas.org/dtd/dasstyle.dtd
>   and
> http://www.biodas.org/documents/sample_stylesheet.xml
>   and the 1.0 spec at
> http://biodas.org/documents/spec.html
> 
> These are still present in the CVS files.
> 
> Here's the differences I found so far.
> 
> That XML file is not well-formed -
> 
> * End tag 'BOXS' does not match the start tag 'BOX'. Line 166, Position 13
> 
> Here's the relevant XML
>           <BOX>
>             <HEIGHT>3</HEIGHT>
>             <FGCOLOR>black</FGCOLOR>
>             <LABEL>yes</LABEL>
>             <BGCOLOR>red</BGCOLOR>
>             <BUMP>yes</BUMP>
>           </BOXS>
>           ^^^^^^^ should be 'BOX'
> 
> The DTD needs to have "WIDTH" changed to "HEIGHT" (according to
> Version 0.99 changelog info)
> 
> The DTD doesn't include the "FGCOLOR" nor "BGCOLOR" definitions.
> 
> In the DTD, LINE needs to include a "STYLE" and a "LINEWIDTH",
> which are used in the XML file.
> 
> The DTD is missing ANCHORED_ARROW, which is in the XML file
> and in the spec.  (I see that Brian's XMLSPY output at
> 
> http://www-genome.wi.mit.edu/personal/gilmanb/i3c/xmlspy_docs/das/style.html
> also doesn't have ANCHORED_ARROW nor does any of the code in the
> CVS release.)
> 
> The spec has a "HIDDEN" glyph which is not in the DTD nor is
> in the XML.  (Added in Version 1.0)
> 
> The XML file has LINEWIDTH for ANCHORED_ARROW, which probably
> implies ARROW needs it as well in the DTD.
> 
> The XML file has "POINT"s for TRIANGLE and BOX, but there is
> no POINT in the XML.  Nor is there a "POINT" in the spec.
> 
> The DTD defines a "CONNECTOR" which is not listed in the spec.
> 
> What's the interpretation of "PRIMERS"?
> 
>         <GLYPH>
>           <PRIMERS>
>             <HEIGHT>50</HEIGHT>
>             <FGCOLOR>#33ff33</FGCOLOR>
>             <LABEL>yes</LABEL>
>           </PRIMERS>
>         </GLYPH>
> 
> What's the interpretation of "SEGMENTS"?
>         <GLYPH>
>           <SEGMENTS>
>             <HEIGHT>50</HEIGHT>
>             <FGCOLOR>#666666</FGCOLOR>
>             <LABEL>yes</LABEL>
>             <LINEWIDTH>1</LINEWIDTH>
>           </SEGMENTS>
>         </GLYPH>
> 
> And 'TRANSCRIPT'?  Shouldn't all three of these be stored
> as a TYPE id?
> 
> There's still a 'FILLCOLOR' in sample_stylesheets.xml .
> I believe this should be 'BGCOLOR'
> 
> Line 571 of that file needs a /CATEGORY as otherwise the
> document is not well formed.  Eg, see the comment in the
> following grep
> 
> cvs:documents {410} grep CATEGORY sample_stylesheet.xml
>   <CATEGORY id="group">
>   </CATEGORY>
>   <CATEGORY id="structural">
>     </CATEGORY>
>     <CATEGORY id="similarity">
>     </CATEGORY>
>     <CATEGORY id="variation">
>     </CATEGORY>
>     <CATEGORY id="experimental">
>     </CATEGORY>
>     <CATEGORY id="translation">  *--- Where's the close?
>     <CATEGORY id="transcription">
>     </CATEGORY>
>     <CATEGORY id="repeat">
>     </CATEGORY>
>     <CATEGORY id="default">
>     </CATEGORY>
> cvs:documents {411}
> 
> Am I working off of outdated documents?  Or is the DTD really
> incomplete?
> 
> 
> I also did a quick scan of the geodesic code to see what it
> handles.  I see that:
>  - "CONNECTOR" is not on the spec page
>  - it doesn't support "ANCHORED_ARROW"
>  - it doesn't support "HIDDEN"
>  - it has "OUTLINECOLOR" instead of "BGCOLOR"
> I did not to a full search so may have missed things.
> 
> Also, where are the regression tests?  I wanted to see how
> easy it is to use them from Python.
> 
>                     Andrew Dalke
>                     dalke@dalkescientific.com
> 
> 
> _______________________________________________
> DAS mailing list
> DAS@biodas.org
> http://biodas.org/mailman/listinfo/das
>