Skip to main content

XML parsing for LLM outputs

Project description

llmxml - XML parsing for LLM outputs

XML outputs are a common format for LLM outputs, but they are not always easy to parse. This package provides a way to parse XML outputs into Pydantic models.

This supports nested models, lists, and unions as well as partial parsing for streamed responses. llmxml also supports generating a prompt template for the XML schema to pass to the LLM.

Parser Demo

Usage

Parsing XML

from llmxml import parse_xml, generate_prompt_template
from pydantic import BaseModel, Field

class Movie(BaseModel):
    title: str = Field(..., description="The title of the movie")
    director: str = Field(..., description="The director of the movie")

class Response(BaseModel):
    movies: list[Movie] = Field(
        ..., description="A list of movies that match the query"
    )

class ResponseObject(BaseModel):
    response: Response = Field(
        ..., description="The response object that contains the movies"
    )

xml: str = """
<response>
    <movies>
        <movie>
            <title>The Matrix</title>
            <director>The Wachowskis</director>
        </movie>
    </movies>
</response>
"""

result: ResponseObject = parse_xml(ResponseObject, xml)
print(result)

Output:

response=Response(movies=[Movie(title='The Matrix', director='The Wachowskis')])

Generating a prompt template

prompt: str = generate_prompt_template(
    model=Response, include_instructions=True  # Default is true
)
print(prompt)

Output:

<response_instructions>
You are to understand the content and provide the parsed objects in xml that match the following xml_schema:

Make sure to return an instance of the XML, not the schema itself

Each field in the field_schema has a description and a type.
Example:
<field_name>
[type]
[description]
</field_name>

Schema:
<movies>
[type: list]
[A list of movies]
</movies>
</response_instructions>

Development

Running tests

$ uv run -m pytest -n auto tests/test_parser.py

License

This project is licensed under the MIT License. See the LICENSE file for details.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

llmxml-0.4.0.tar.gz (16.9 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

llmxml-0.4.0-py3-none-any.whl (13.1 kB view details)

Uploaded Python 3

File details

Details for the file llmxml-0.4.0.tar.gz.

File metadata

  • Download URL: llmxml-0.4.0.tar.gz
  • Upload date:
  • Size: 16.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.5.0

File hashes

Hashes for llmxml-0.4.0.tar.gz
Algorithm Hash digest
SHA256 efddb078e5b6b31ec20ca59994bb01bc70059455d3155961fcc87a011df4926b
MD5 367f5a7c0f445134468abff808dfb625
BLAKE2b-256 dd4c1ebbcc1e9430dd3c63b5cb250d8d6b751a414fe905874996e0571c8c747b

See more details on using hashes here.

File details

Details for the file llmxml-0.4.0-py3-none-any.whl.

File metadata

  • Download URL: llmxml-0.4.0-py3-none-any.whl
  • Upload date:
  • Size: 13.1 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.5.0

File hashes

Hashes for llmxml-0.4.0-py3-none-any.whl
Algorithm Hash digest
SHA256 49ff1a0e983870909478b9c6cfb32e7ddf14740f6a1b10d9b7218ff17e2e2513
MD5 b42110ccdfe1aaf0f81b8a61616c2ccd
BLAKE2b-256 26d086d44ee437c6522ea43dc7eb22c8e0a4cebd184661af27416f2e5ea79643

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page