Always treat nullable/optional fields and parsed payloads as untrusted at boundaries: normalize None/missing values to the expected type before iterating/indexing, validate the parsed shape immediately (fail fast with a clear error), and explicitly handle sentinel/unknown states instead of “silently succeeding” with wrong assumptions.
Apply this as a local rule:
for x in (value or []) (or equivalent) when a field is “list-ish” but may be null..lower() strings).json.loads/spec parsing, verify the root/object shape (e.g., dict/mapping) before using keys.unknown/unresolvable, return an explicit structured error or platform: 'unknown'+hint; don’t proceed with the “web/mobile” logic.Example (list field that may be absent, null, or contain bad entries):
from collections import defaultdict
from typing import Any
def index_tags(dataset: dict[str, Any]) -> dict[str, list[int]]:
tag_index: dict[str, list[int]] = defaultdict(list)
tags = dataset.get('Tags') or [] # handles missing or null
if not isinstance(tags, list):
raise ValueError("Tags must be a list when present")
for tag in tags:
if not isinstance(tag, str):
continue # or raise, depending on policy
tag_index[tag.lower()].append(0)
return tag_index
Example (parsed payload root validation):
def require_mapping(obj: object) -> dict:
if not isinstance(obj, dict):
raise ValueError(f"Expected a mapping at the root, got {type(obj).__name__}")
return obj
spec = require_mapping(parsed_spec) # only then do spec['openapi'] checks
Outcome: fewer crashes on null, fewer opaque downstream TypeErrors, and safer behavior when inputs are missing or unresolved.