[Bug] Documentation for XMLAdapter missing

[stargazer33]

What happened?

Current documentation under https://dspy.ai/api/ does not mention XMLAdapter.
Please describe it.

Steps to reproduce

https://dspy.ai/api/

DSPy version

latest

[stargazer33]

Screenshot of documentation:

XMLAdapter: https://github.com/stanfordnlp/dspy/blob/main/dspy/adapters/xml_adapter.py

[chenmoneygithub]

@stargazer33 Thanks for reporting the issue! XMLAdapter doesn't make much difference from JSONAdapter now, so we are not exposing it with full visibility. We will find bandwidth to look into #8482, which makes real use of XML format.

[stargazer33]

Thanks @chenmoneygithub ! So the current implementation of XMLAdapter: it should work with something like this

  <out>
      <field1>...</<field1>
      <field2>...</<field2>
      ...
      <fieldN>...</<fieldN>
  </out>

Right?

The <out> tag is there just because XML must have ONE root element... just to satisfy it...
(more details about this XML schema and my use case: https://discord.com/channels/1161519468141355160/1423709109936455691 )

[chenmoneygithub]

@stargazer33 Sorry about the late reply. What you are describing is what we want to ship with #8482, and the current XMLAdapter doesn't include the root tag.

[stargazer33]

... the current XMLAdapter doesn't include the root tag.

aaa, this is important, thanks for the info!
NOW I understand why current XMLAdapter is not documented ))))
Well then, #8482 is a really important thing!

Well, in case I decide to use the current XMLAdapter: I should expect just this output, without the root tag:

      <field1>...</<field1>
      <field2>...</<field2>
      ...
      <fieldN>...</<fieldN>

Is this correct?
(I do not need nested fields, and do not need lists in XML... at the moment... and can adapt my code for "XML without root tag" output)