Using parser with base64 encoded query string

[kderck]

Hi,

I have a string that is a base64 encoded query string is there anyway I can use the parser function to nicely parse this into class without excessive boilerplate?

class Message:
  msg: string 
  user_id: string 
  ...

[leandrodamascena]

Hi @kderck! Could you please elaborate more on this? Are you trying to use our Parser utility for this?

Thank you.

[heitorlessa]

What @leandrodamascena was after was a reproducible info or snippet to offer a more prescriptive guidance :) e.g., how your function gets invoked, sample payload, whether you're using Powertools Parser (Pydantic) etc.

TL;DR: You can use JMESPath Utilities to use our powertools_base64 along with powertools_json to decode and deserialize query strings back.

Assuming you meant the encoded query string once decoded looks like a dictionary:

b64 JSON value: eyJtc2ciOiAiaGVsbG8iLCAidXNlcl9pZCI6ICJkY2M2ZDAzMS00MmJjLTQ0MmUtYmIwYS1lMWJhNzk2YmNkMWMifQ==

{
    "msg": "hello",
    "user_id": "dcc6d031-42bc-442e-bb0a-e1ba796bcd1c"
}

You can reduce the common boilerplate of decode/encode/json.loads) by using JMESPath Utility. There are built-in functions like powertools_base64andpowertools_json` that when combined leads to the result you want.

Reproducible example

import base64
import json
from aws_lambda_powertools.utilities.jmespath_utils import extract_data_from_envelope
from aws_lambda_powertools.utilities.parser import BaseModel

# Reusing your Message class as a Pydantic model... Python @dataclass also works
class Message(BaseModel):
    msg: str
    user_id: str

# Raw dict query string as a JSON string
raw_qs = json.dumps({"msg": "hello", "user_id": "dcc6d031-42bc-442e-bb0a-e1ba796bcd1c"})

# Base64 encoded JSON query string as string
b64_encoded_qs: str = base64.b64encode(raw_qs.encode()).decode()

# Mimicking the payload you shared
event = {
    "body": b64_encoded_qs
}

## What You Actually Need
#
# 1. Decode base64 string available in `body` key
# 2. Deserialize JSON string into a normal dict
decoded_qs = extract_data_from_envelope(data=event, envelope="powertools_json( powertools_base64(body) )")

# Use expansion to convert dict into key=value arguments to instantiate Message class
message = Message(**decoded_qs)

assert message.msg == "hello"

PS: The name extract_data_from_envelope is an old tech debt we haven't addressed yet, we should have a more friendly name.

[heitorlessa]

If you are however using API Gateway as an event source, you can use Event Handler feature to further reduce boilerplate for handling a specific webhook request... then use app.current_event.decoded_body to get a string representation.

That's why @leandrodamascena asked for more information, as there's a quick way to solve the problem (JMESPath) and there's the elegant way with Event Handler ;)

from aws_lambda_powertools import Logger
from aws_lambda_powertools.event_handler import APIGatewayRestResolver
from aws_lambda_powertools.logging import correlation_paths
from aws_lambda_powertools.utilities.typing import LambdaContext

logger = Logger()
app = APIGatewayRestResolver()


class Message(BaseModel):
    msg: str
    user_id: str


@app.get("/slack_webhook")
@tracer.capture_method
def get_todos():
    payload = json.loads(app.current_event.decoded_body)
    message = Message(**payload)

    return {"message": message.msg}


# You can continue to use other utilities just as before
@logger.inject_lambda_context(correlation_id_path=correlation_paths.API_GATEWAY_REST)
def lambda_handler(event: dict, context: LambdaContext) -> dict:
    return app.resolve(event, context)

[kderck]

Hi @heitorlessa thank for your incredibly well wrote answer. I appreciate it, and I apologise for the lack of in-depth context in the request. I'l follow up after I've given it ago!

[heitorlessa]

No worries at all ;) Looking forward to hearing from you

[kderck]

Hi @heitorlessa @leandrodamascena haven't forgotten just haven't got back to this particular lambda yet.

[leandrodamascena]

Closing this discussion as it is resolved.

Feel free to reopen if you have more questions @kderck.