feat(02-03): add MediaAttachment model, WhatsApp normalizer, and signature verification

- Add MediaType(StrEnum) and MediaAttachment(BaseModel) to shared/models/message.py
- Add media: list[MediaAttachment] field to MessageContent
- Add whatsapp_app_secret, whatsapp_verify_token, and MinIO settings to shared/config.py
- Add normalize_whatsapp_event() to gateway/normalize.py (text, image, document support)
- Create whatsapp.py adapter with verify_whatsapp_signature() and verify_hub_challenge()
- 30 new passing tests (signature verification + normalizer)
This commit is contained in:
2026-03-23 14:41:48 -06:00
parent b2e86f1046
commit 370a860622
7 changed files with 1987 additions and 3 deletions

View File

@@ -0,0 +1,365 @@
"""
Unit tests for WhatsApp event normalization to KonstructMessage.
Tests CHAN-03: Messages are normalized to the unified KonstructMessage format.
Covers:
- Text message normalizes correctly
- Image message normalizes with MediaAttachment
- Document message normalizes with MediaAttachment
- Correct field mapping: sender, channel, metadata
- ChannelType includes 'whatsapp'
- thread_id set to sender wa_id (WhatsApp has no threads)
- tenant_id is None after normalization
"""
from __future__ import annotations
from datetime import timezone
import pytest
from shared.models.message import ChannelType, MediaAttachment, MediaType
def make_text_webhook_payload(
phone_number_id: str = "12345678901",
wa_id: str = "5511987654321",
message_id: str = "wamid.abc123",
text: str = "Hello, I need help with my order",
timestamp: str = "1711234567",
) -> dict:
"""Minimal valid Meta Cloud API v20.0 text message webhook payload."""
return {
"object": "whatsapp_business_account",
"entry": [
{
"id": "WHATSAPP_BUSINESS_ACCOUNT_ID",
"changes": [
{
"value": {
"messaging_product": "whatsapp",
"metadata": {
"display_phone_number": "15550000000",
"phone_number_id": phone_number_id,
},
"contacts": [
{
"profile": {"name": "John Customer"},
"wa_id": wa_id,
}
],
"messages": [
{
"from": wa_id,
"id": message_id,
"timestamp": timestamp,
"text": {"body": text},
"type": "text",
}
],
},
"field": "messages",
}
],
}
],
}
def make_image_webhook_payload(
phone_number_id: str = "12345678901",
wa_id: str = "5511987654321",
message_id: str = "wamid.img456",
media_id: str = "media_id_xyz789",
mime_type: str = "image/jpeg",
) -> dict:
"""Minimal valid Meta Cloud API v20.0 image message webhook payload."""
return {
"object": "whatsapp_business_account",
"entry": [
{
"id": "WHATSAPP_BUSINESS_ACCOUNT_ID",
"changes": [
{
"value": {
"messaging_product": "whatsapp",
"metadata": {
"display_phone_number": "15550000000",
"phone_number_id": phone_number_id,
},
"contacts": [
{
"profile": {"name": "Jane Customer"},
"wa_id": wa_id,
}
],
"messages": [
{
"from": wa_id,
"id": message_id,
"timestamp": "1711234567",
"image": {
"id": media_id,
"mime_type": mime_type,
"sha256": "abc123hash",
},
"type": "image",
}
],
},
"field": "messages",
}
],
}
],
}
def make_document_webhook_payload(
phone_number_id: str = "12345678901",
wa_id: str = "5511987654321",
message_id: str = "wamid.doc789",
media_id: str = "media_id_doc001",
filename: str = "invoice.pdf",
) -> dict:
"""Minimal valid Meta Cloud API v20.0 document message webhook payload."""
return {
"object": "whatsapp_business_account",
"entry": [
{
"id": "WHATSAPP_BUSINESS_ACCOUNT_ID",
"changes": [
{
"value": {
"messaging_product": "whatsapp",
"metadata": {
"display_phone_number": "15550000000",
"phone_number_id": phone_number_id,
},
"contacts": [
{
"profile": {"name": "Bob Customer"},
"wa_id": wa_id,
}
],
"messages": [
{
"from": wa_id,
"id": message_id,
"timestamp": "1711234567",
"document": {
"id": media_id,
"mime_type": "application/pdf",
"filename": filename,
},
"type": "document",
}
],
},
"field": "messages",
}
],
}
],
}
class TestChannelTypeWhatsApp:
"""Tests for ChannelType enum including WhatsApp."""
def test_whatsapp_in_channel_type(self) -> None:
"""ChannelType enum must include 'whatsapp' value."""
assert ChannelType.WHATSAPP == "whatsapp"
assert "whatsapp" in [c.value for c in ChannelType]
def test_whatsapp_channel_type_str(self) -> None:
"""ChannelType.WHATSAPP must equal string 'whatsapp'."""
assert str(ChannelType.WHATSAPP) == "whatsapp"
class TestMediaAttachmentModel:
"""Tests for the MediaAttachment model."""
def test_media_attachment_image_fields(self) -> None:
"""MediaAttachment must accept image type with required fields."""
attachment = MediaAttachment(
media_type=MediaType.IMAGE,
mime_type="image/jpeg",
)
assert attachment.media_type == MediaType.IMAGE
assert attachment.mime_type == "image/jpeg"
assert attachment.url is None
assert attachment.storage_key is None
def test_media_attachment_document_with_filename(self) -> None:
"""MediaAttachment must accept document type with filename."""
attachment = MediaAttachment(
media_type=MediaType.DOCUMENT,
mime_type="application/pdf",
filename="report.pdf",
size_bytes=102400,
)
assert attachment.media_type == MediaType.DOCUMENT
assert attachment.filename == "report.pdf"
assert attachment.size_bytes == 102400
def test_media_attachment_all_fields_optional_except_type(self) -> None:
"""MediaAttachment requires only media_type; all other fields are optional."""
attachment = MediaAttachment(media_type=MediaType.AUDIO)
assert attachment.media_type == MediaType.AUDIO
assert attachment.url is None
assert attachment.storage_key is None
assert attachment.mime_type is None
assert attachment.filename is None
assert attachment.size_bytes is None
class TestNormalizeWhatsAppText:
"""Tests for normalizing WhatsApp text messages."""
def test_channel_type_is_whatsapp(self) -> None:
"""Normalized message must have channel=whatsapp."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload()
msg = normalize_whatsapp_event(payload)
assert msg.channel == ChannelType.WHATSAPP
def test_sender_user_id_is_wa_id(self) -> None:
"""sender.user_id must be the WhatsApp ID (wa_id) of the sender."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload(wa_id="5511987654321")
msg = normalize_whatsapp_event(payload)
assert msg.sender.user_id == "5511987654321"
def test_content_text_extracted(self) -> None:
"""content.text must contain the message body text."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload(text="Can you help me track my order?")
msg = normalize_whatsapp_event(payload)
assert msg.content.text == "Can you help me track my order?"
def test_channel_metadata_phone_number_id(self) -> None:
"""channel_metadata must contain phone_number_id."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload(phone_number_id="99988877766")
msg = normalize_whatsapp_event(payload)
assert msg.channel_metadata["phone_number_id"] == "99988877766"
def test_channel_metadata_message_id(self) -> None:
"""channel_metadata must contain the WhatsApp message ID (wamid)."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload(message_id="wamid.unique123")
msg = normalize_whatsapp_event(payload)
assert msg.channel_metadata["message_id"] == "wamid.unique123"
def test_thread_id_is_sender_wa_id(self) -> None:
"""thread_id must be the sender's wa_id (WhatsApp has no threads)."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload(wa_id="5511999998888")
msg = normalize_whatsapp_event(payload)
assert msg.thread_id == "5511999998888"
def test_tenant_id_none_after_normalization(self) -> None:
"""tenant_id must be None immediately after normalization."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload()
msg = normalize_whatsapp_event(payload)
assert msg.tenant_id is None
def test_timestamp_is_utc(self) -> None:
"""timestamp must be a timezone-aware datetime in UTC."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload(timestamp="1711234567")
msg = normalize_whatsapp_event(payload)
assert msg.timestamp.tzinfo is not None
assert msg.timestamp.tzinfo == timezone.utc
def test_no_media_attachments_for_text_message(self) -> None:
"""Text message must not produce any media attachments."""
from gateway.normalize import normalize_whatsapp_event
payload = make_text_webhook_payload()
msg = normalize_whatsapp_event(payload)
assert msg.content.media == []
class TestNormalizeWhatsAppImage:
"""Tests for normalizing WhatsApp image messages."""
def test_image_message_has_media_attachment(self) -> None:
"""Image message must produce a MediaAttachment in content.media."""
from gateway.normalize import normalize_whatsapp_event
payload = make_image_webhook_payload()
msg = normalize_whatsapp_event(payload)
assert len(msg.content.media) == 1
def test_image_attachment_type_is_image(self) -> None:
"""MediaAttachment media_type must be IMAGE for image messages."""
from gateway.normalize import normalize_whatsapp_event
payload = make_image_webhook_payload()
msg = normalize_whatsapp_event(payload)
assert msg.content.media[0].media_type == MediaType.IMAGE
def test_image_attachment_mime_type(self) -> None:
"""MediaAttachment mime_type must be preserved from webhook payload."""
from gateway.normalize import normalize_whatsapp_event
payload = make_image_webhook_payload(mime_type="image/png")
msg = normalize_whatsapp_event(payload)
assert msg.content.media[0].mime_type == "image/png"
def test_image_attachment_url_contains_media_id(self) -> None:
"""MediaAttachment url must contain the media_id as a placeholder for later download."""
from gateway.normalize import normalize_whatsapp_event
payload = make_image_webhook_payload(media_id="media_id_xyz789")
msg = normalize_whatsapp_event(payload)
assert msg.content.media[0].url is not None
assert "media_id_xyz789" in msg.content.media[0].url
def test_image_message_text_is_empty_string(self) -> None:
"""Image message with no caption must have empty string as text content."""
from gateway.normalize import normalize_whatsapp_event
payload = make_image_webhook_payload()
msg = normalize_whatsapp_event(payload)
assert msg.content.text == ""
class TestNormalizeWhatsAppDocument:
"""Tests for normalizing WhatsApp document messages."""
def test_document_message_has_media_attachment(self) -> None:
"""Document message must produce a MediaAttachment in content.media."""
from gateway.normalize import normalize_whatsapp_event
payload = make_document_webhook_payload()
msg = normalize_whatsapp_event(payload)
assert len(msg.content.media) == 1
def test_document_attachment_type_is_document(self) -> None:
"""MediaAttachment media_type must be DOCUMENT for document messages."""
from gateway.normalize import normalize_whatsapp_event
payload = make_document_webhook_payload()
msg = normalize_whatsapp_event(payload)
assert msg.content.media[0].media_type == MediaType.DOCUMENT
def test_document_attachment_filename(self) -> None:
"""MediaAttachment filename must be preserved from webhook payload."""
from gateway.normalize import normalize_whatsapp_event
payload = make_document_webhook_payload(filename="contract.pdf")
msg = normalize_whatsapp_event(payload)
assert msg.content.media[0].filename == "contract.pdf"