open-cmis-tck/src/open_cmis_tck/preflight.py

"""CMIS Browser Binding preflight runner."""

from __future__ import annotations

import json
from pathlib import Path
from typing import Any
from urllib.error import HTTPError, URLError
from urllib.request import Request, urlopen


def run(context: dict[str, Any]) -> dict[str, Any]:
    target = context["target_profile"]
    endpoint = _browser_endpoint(target)
    if endpoint is None:
        return {
            "result": "fail",
            "observations": [
                "Target profile does not declare a CMIS Browser Binding endpoint."
            ],
            "facts": {
                "endpoint_found": False,
            },
            "artifact_refs": [],
        }

    timeout = _timeout_seconds(context)
    artifact_refs: list[str] = []
    request = Request(
        endpoint["url"],
        headers={
            "Accept": "application/json, */*;q=0.1",
            "User-Agent": "guide-board-open-cmis-tck-preflight/0.1.0",
        },
    )
    try:
        with urlopen(request, timeout=timeout) as response:
            status_code = response.status
            content_type = response.headers.get("Content-Type", "")
            headers = dict(response.headers.items())
            body = response.read(1024 * 1024)
            artifact_refs = _write_response_artifacts(
                context,
                status_code,
                content_type,
                headers,
                body,
            )
    except HTTPError as exc:
        body = exc.read(1024 * 1024)
        content_type = exc.headers.get("Content-Type", "")
        artifact_refs = _write_response_artifacts(
            context,
            exc.code,
            content_type,
            dict(exc.headers.items()),
            body,
        )
        return {
            "result": "infrastructure_error",
            "observations": [
                f"CMIS Browser Binding endpoint returned HTTP {exc.code}."
            ],
            "facts": {
                "endpoint_found": True,
                "url": endpoint["url"],
                "http_status": exc.code,
                "content_type": content_type,
            },
            "artifact_refs": artifact_refs,
        }
    except URLError as exc:
        return {
            "result": "infrastructure_error",
            "observations": [
                f"CMIS Browser Binding endpoint is not reachable: {exc.reason}."
            ],
            "facts": {
                "endpoint_found": True,
                "url": endpoint["url"],
                "error": str(exc.reason),
            },
            "artifact_refs": artifact_refs,
        }
    except TimeoutError:
        return {
            "result": "infrastructure_error",
            "observations": [
                f"CMIS Browser Binding endpoint did not respond within {timeout} seconds."
            ],
            "facts": {
                "endpoint_found": True,
                "url": endpoint["url"],
                "timeout_seconds": timeout,
            },
            "artifact_refs": artifact_refs,
        }

    facts: dict[str, Any] = {
        "endpoint_found": True,
        "url": endpoint["url"],
        "binding": endpoint["binding"],
        "http_status": status_code,
        "content_type": content_type,
    }

    parsed = _parse_json(body)
    if parsed is None:
        facts["json_detected"] = False
        return {
            "result": "warning",
            "observations": [
                "CMIS Browser Binding endpoint is reachable but did not return parseable JSON."
            ],
            "facts": facts,
            "artifact_refs": artifact_refs,
        }

    facts["json_detected"] = True
    facts.update(_repository_facts(parsed))
    return {
        "result": "pass",
        "observations": [
            "CMIS Browser Binding endpoint is reachable and returned parseable JSON."
        ],
        "facts": facts,
        "artifact_refs": artifact_refs,
    }


def _browser_endpoint(target: dict[str, Any]) -> dict[str, Any] | None:
    for endpoint in target.get("endpoints", []):
        if endpoint.get("binding") == "cmis-browser":
            return endpoint
    return None


def _timeout_seconds(context: dict[str, Any]) -> float:
    runtime_policy = context["assessment_profile"].get("runtime_policy", {})
    configured = runtime_policy.get("timeout_seconds", 5)
    if not isinstance(configured, (int, float)):
        return 5.0
    return max(1.0, min(float(configured), 10.0))


def _parse_json(body: bytes) -> Any:
    try:
        return json.loads(body.decode("utf-8"))
    except (UnicodeDecodeError, json.JSONDecodeError):
        return None


def _repository_facts(value: Any) -> dict[str, Any]:
    if not isinstance(value, dict):
        return {"repository_shape": "unknown"}

    if "repositoryId" in value:
        return {
            "repository_shape": "single-repository-info",
            "repository_ids": [value["repositoryId"]],
            "cmis_version_supported": value.get("cmisVersionSupported"),
            "capabilities_present": isinstance(value.get("capabilities"), dict),
        }

    repository_ids = []
    for key, child in value.items():
        if isinstance(child, dict) and (
            "repositoryId" in child or "repositoryName" in child
        ):
            repository_ids.append(str(child.get("repositoryId", key)))

    if repository_ids:
        return {
            "repository_shape": "repository-map",
            "repository_ids": repository_ids,
        }

    return {
        "repository_shape": "object",
        "top_level_keys": sorted(str(key) for key in value.keys())[:20],
    }


def _write_response_artifacts(
    context: dict[str, Any],
    status_code: int,
    content_type: str,
    headers: dict[str, str],
    body: bytes,
) -> list[str]:
    run_dir = Path(context["run_dir"])
    artifact_dir = run_dir / "artifacts" / "open-cmis-tck" / "preflight"
    artifact_dir.mkdir(parents=True, exist_ok=True)

    response_ref = "artifacts/open-cmis-tck/preflight/response-body.bin"
    metadata_ref = "artifacts/open-cmis-tck/preflight/response-metadata.json"

    (run_dir / response_ref).write_bytes(body)
    (run_dir / metadata_ref).write_text(
        json.dumps(
            {
                "status_code": status_code,
                "content_type": content_type,
                "headers": headers,
                "byte_count": len(body),
            },
            indent=2,
            sort_keys=True,
        )
        + "\n",
        encoding="utf-8",
    )
    return [metadata_ref, response_ref]