Poller Metadata Toolkit

Helpers for launching document analysis jobs with adjustable polling cadence, exposing operation identifiers from service metadata, and resuming in-flight work from continuation tokens.

Capabilities

Start analysis with custom polling

• Launches a document analysis run using the caller-provided polling interval (seconds) and returns the completed result plus a continuation token for future resume. @test

Surface operation metadata

• Provides the operation identifier emitted by the long-running operation metadata alongside a summary of the completed analysis (e.g., page count). @test

Resume from saved token

• Rehydrates an existing analysis using only a continuation token without re-supplying the input file, yielding the same operation identifier and finished content. @test

Implementation

@generates

API

from dataclasses import dataclass
from typing import Any

@dataclass
class AnalysisJob:
    operation_id: str
    continuation_token: str
    page_count: int
    content: Any

def analyze_document_with_metadata(
    client: Any,
    model_id: str,
    document_path: str,
    poll_interval: int = 1
) -> AnalysisJob:
    """
    Starts a document analysis with a configurable polling interval, then returns
    the completed content plus operation metadata and a continuation token.
    """

def resume_analysis_with_metadata(
    client: Any,
    continuation_token: str,
    poll_interval: int = 1
) -> AnalysisJob:
    """
    Resumes a prior analysis using a continuation token and returns the same
    operation metadata and final content.
    """

Dependencies { .dependencies }

azure-ai-documentintelligence { .dependency }

Provides document analysis client, enhanced poller metadata, and continuation token support.