Discussion:
Need teilite.dtd and conversion tool
(too old to reply)
elel
2007-10-24 17:18:30 UTC
Permalink
I am trying to index what I suspect are some very old tei files.
Lucene is not liking them. Does anyone out there know of a tei version
1 to current version + sgml to xml converter. Also, where can I find
teilite.dtd? The one I have is producing errors.

Thanks for the help.
Lynn
William F Hammond
2007-10-24 18:25:41 UTC
Permalink
Post by elel
I am trying to index what I suspect are some very old tei files.
Lucene is not liking them. Does anyone out there know of a tei version
1 to current version + sgml to xml converter. Also, where can I find
teilite.dtd? The one I have is producing errors.
You didn't say how old, and you didn't say what errors.

This might be a place for you to start:

http://www.tei-c.org/Lite/DTD/


-- Bill
elel
2007-10-24 20:25:33 UTC
Permalink
Thanks for the link. Even with the new dtd, I am trying to use SP to
convert sgml to xml and nearly every element is giving the following
error:

Error: undeclared element 'teiheader'

Starting with this:

Error: the root element 'tei.2' must match the name 'TEI.2' from the
document type declaration
Line 4: <tei.2>
Col 8: -------^

Make any sense to anyone?

Thanks again.
Peter Flynn
2007-10-24 23:52:27 UTC
Permalink
Post by elel
Thanks for the link. Even with the new dtd, I am trying to use SP to
convert sgml to xml and nearly every element is giving the following
Error: undeclared element 'teiheader'
The XML version of the TEI uses camelCased names, so it has to be
teiHeader. SP cannot know or guess this, so it can't convert the names to
the correct case automagically. By default it just lowercases everything.

The script I wrote to convert the CELT documents from SGML to XML fixed
this by extracting all the element type names used by a document, doing a
caseless grep in the DTD files themselves for the element declaration,
and capturing the cased version of the names to make a sed script, which
is then applied to the document after normalization. I can dig out a copy
if you're interested.

///Peter

Loading...