pyodata

Common errors

• pyodata.exceptions.PyODataParserError: Unsupported Schema namespace - http://schemas.microsoft.com/ado/2007/05/edm

  cause The OData service's metadata document uses an XML schema namespace that `pyodata`'s default parser configuration does not recognize or support, often indicating a service that deviates from standard OData metadata specifications.
  fix Configure the `pyodata.Client` with custom XML namespaces if the service uses non-standard ones, or apply a tolerant parser policy (e.g., `PolicyWarning` or `PolicyIgnore`) via the `config` parameter to allow parsing despite minor schema issues.

• HTTP status code 403 Forbidden on POST/PATCH requests (e.g., when creating or updating an entity).

  cause This is typically caused by missing or invalid CSRF (Cross-Site Request Forgery) tokens, which are a security measure on OData services, especially for modifying operations.
  fix Before sending a POST or PATCH request, perform a HEAD request to the service URL (e.g., `session.head(SERVICE_URL, headers={'x-csrf-token': 'fetch'})`). Extract the `x-csrf-token` and `set-cookie` headers from the response and include them in the subsequent modifying request's headers or ensure the session object correctly manages them.

• Incomplete data retrieved for large collections; only a partial list of entities is returned.

  cause The OData service is implementing server-side pagination, returning only a subset of entities with a link to the next page, typically indicated by a `__next` field in the response.
  fix After executing a `get_entities()` request, check the response for a `__next` attribute. If present, it contains the URI for the next page of results. You'll need to implement logic to follow these `__next` links to retrieve all pages of data.

Warnings

• breaking Python 3.7 and 3.8 are no longer supported. The library now requires Python >= 3.9.
  Fix: Upgrade your Python environment to 3.9 or newer.
• gotcha When creating or updating entities (POST/PATCH requests), OData services often require CSRF tokens. You must correctly initialize your `requests.Session` to handle these, typically by fetching a token with a HEAD request first.
  Fix: Before making a POST/PATCH request, execute a HEAD request to the service URL to obtain the 'x-csrf-token' and 'set-cookie' headers, then include them in subsequent requests. Example: `session.head(SERVICE_URL, headers={'x-csrf-token': 'fetch'})`.
• gotcha The metadata parser can encounter errors with non-fully valid metadata documents. This can lead to issues if the service is not under your control. By default, missing properties might be filled with type-specific default values.
  Fix: Configure the metadata parser to recover from problems using `pyodata.Client`'s `config` parameter. You can specify parser policies (e.g., `PolicyWarning`, `PolicyIgnore`) and set `retain_null=False` to prevent substitution by default values.
• deprecated Directly setting custom XML namespaces via the client initialization is deprecated.
  Fix: Use the `config` object (e.g., `pyodata.client.config`) to manage XML namespaces and other parser settings instead.
• gotcha OData services may require non-standard URL query parameters (e.g., 'sap-client') or specific path encoding. The default behavior might not always match specific service implementations.
  Fix: Use the `.custom(name, value)` method on entity requests to add custom query parameters. To disable URL path encoding, set `encode_path=False` during client initialization or on the request.

Install

• pip install -U pyodata Install latest version

Imports

• Client

  import pyodata
  # ...
  client = pyodata.Client(...)

  The primary entry point for interacting with an OData service.
• GetEntitySetFilter

  from pyodata.v2.service import GetEntitySetFilter as esf

  Used for advanced filtering of entity sets in a more Pythonic/ORM-like style.
• HttpError

  from pyodata.exceptions import HttpError

  Base exception for HTTP errors (status codes >= 400) returned by the OData service.

Quickstart

Initializes an OData client using `requests.Session` and queries the 'Products' entity set from the Northwind OData V2 sample service. This demonstrates basic client setup and data retrieval.

import requests
import pyodata
import os

# Replace with your OData service URL
SERVICE_URL = os.environ.get('PYODATA_SERVICE_URL', 'http://services.odata.org/V2/Northwind/Northwind.svc/')

session = requests.Session()

# For services requiring authentication
# username = os.environ.get('PYODATA_USERNAME', '')
# password = os.environ.get('PYODATA_PASSWORD', '')
# if username and password:
#     session.auth = (username, password)

# Create instance of OData client
client = pyodata.Client(SERVICE_URL, session)

# Example: Query all products
products_request = client.entity_sets.Products.get_entities()
products = products_request.execute()

print(f"Found {len(products)} products:")
for product in products[:3]: # Print first 3 products
    print(f"- {product.ProductName} (ID: {product.ProductID})")