Is there an Oracle wrapper for Python that supports xmltype columns?

Question

It seems cx_Oracle doesn't.

Any other suggestion for handling xml with Oracle and Python is appreciated.

Thanks.

Answer 1

I managed to do this with cx_Oracle.

I used the sys.xmltype.createxml() function in the statement that inserts the rows in a table with XMLTYPE fields; then I used prepare() and setinputsizes() to specify that the bind variables I used for XMLTYPE fields were of cx_Oracle.CLOB type.

Answer 2

I managed to get this to work by wrapping the XMLElement call in a call to XMLType.GetClobVal():

For example:

select xmltype.getclobval(xmlelement("rowcount", count(1)))
from...

No idea of the limitations yet but it got me out of trouble. Found the relelvant info on Oracle site: Mastering Oracle+Python, Part 1: Querying Best Practices

Answer 3

(edited to remove mention of a non-Oracle Python DB-API module and add some more relevant and hopefully useful info).

Don't know of any alternative to cx_oracle (as the DCOracle2 author says, "DCOracle2 is currently unmaintained, and no support is available." so it's not really an alternative).

However, a recent article on Oracle's own site asserts that (at least with recent releases such as Oracle 10g XE -- and presumably recent cx_oracle releases) Python can work with Oracle's XML support -- I don't know if the examples in that article can help you address your issues, but I sure hope so!