docs(v2): document optional dependencies and local dev (aws-powertools#1574)

Author: heitorlessa

README.md

@@ -1,11 +1,12 @@
+<!-- markdownlint-disable MD013 MD041 MD043  -->
 # AWS Lambda Powertools for Python
 
 [![Build](https://github.com/awslabs/aws-lambda-powertools-python/actions/workflows/python_build.yml/badge.svg)](https://github.com/awslabs/aws-lambda-powertools-python/actions/workflows/python_build.yml)
 [![codecov.io](https://codecov.io/github/awslabs/aws-lambda-powertools-python/branch/develop/graphs/badge.svg)](https://app.codecov.io/gh/awslabs/aws-lambda-powertools-python)
 ![PythonSupport](https://img.shields.io/static/v1?label=python&message=3.6%20|%203.7|%203.8|%203.9&color=blue?style=flat-square&logo=python) ![PyPI version](https://badge.fury.io/py/aws-lambda-powertools.svg) ![PyPi monthly downloads](https://img.shields.io/pypi/dm/aws-lambda-powertools)
 [![Join our Discord](https://dcbadge.vercel.app/api/server/B8zZKbbyET)](https://discord.gg/B8zZKbbyET)
 
-A suite of Python utilities for AWS Lambda functions to ease adopting best practices such as tracing, structured logging, custom metrics, and more. (AWS Lambda Powertools [Java](https://github.com/awslabs/aws-lambda-powertools-java) and [Typescript](https://github.com/awslabs/aws-lambda-powertools-typescript) is also available).
+A suite of Python utilities for AWS Lambda functions to ease adopting best practices such as tracing, structured logging, custom metrics, and more. (AWS Lambda Powertools for [Java](https://github.com/awslabs/aws-lambda-powertools-java), [Typescript](https://github.com/awslabs/aws-lambda-powertools-typescript), and [.NET](https://awslabs.github.io/aws-lambda-powertools-dotnet/){target="_blank"} are also available).
 
 **[📜Documentation](https://awslabs.github.io/aws-lambda-powertools-python/)** | **[🐍PyPi](https://pypi.org/project/aws-lambda-powertools/)** | **[Roadmap](https://awslabs.github.io/aws-lambda-powertools-python/latest/roadmap/)** | **[Detailed blog post](https://aws.amazon.com/blogs/opensource/simplifying-serverless-best-practices-with-lambda-powertools/)**
 

docs/core/tracer.md

@@ -19,6 +19,14 @@ Tracer is an opinionated thin wrapper for [AWS X-Ray Python SDK](https://github.
 ???+ tip
     All examples shared in this documentation are available within the [project repository](https://github.com/awslabs/aws-lambda-powertools-python/tree/develop/examples){target="_blank"}.
 
+### Install
+
+!!! info "This is not necessary if you're installing Powertools via [Lambda Layer](../index.md#lambda-layer){target="_blank"}"
+
+Add `aws-lambda-powertools[tracer]` as a dependency in your preferred tool: _e.g._, _requirements.txt_, _pyproject.toml_.
+
+This will ensure you have the required dependencies before using Tracer.
+
 ### Permissions
 
 Before your use this utility, your AWS Lambda function [must have permissions](https://docs.aws.amazon.com/lambda/latest/dg/services-xray.html#services-xray-permissions) to send traces to AWS X-Ray.

docs/index.md

@@ -11,7 +11,7 @@ description: AWS Lambda Powertools for Python
 A suite of utilities for AWS Lambda functions to ease adopting best practices such as tracing, structured logging, custom metrics, idempotency, batching, and more.
 
 ???+ note
-    Lambda Powertools is also available for [Java](https://awslabs.github.io/aws-lambda-powertools-java/){target="_blank"} and [TypeScript](https://awslabs.github.io/aws-lambda-powertools-typescript/latest/){target="_blank"}.
+    Powertools is also available for [Java](https://awslabs.github.io/aws-lambda-powertools-java/){target="_blank"}, [TypeScript](https://awslabs.github.io/aws-lambda-powertools-typescript/latest/){target="_blank"}, and [.NET](https://awslabs.github.io/aws-lambda-powertools-dotnet/){target="_blank"}
 
 ## Install
 
@@ -20,11 +20,29 @@ Powertools is available in the following formats:
 * **Lambda Layer**: [**arn:aws:lambda:{region}:017000801446:layer:AWSLambdaPowertoolsPython:33**](#){: .copyMe}:clipboard:
 * **PyPi**: **`pip install aws-lambda-powertools`**
 
+???+ info "Some utilities require additional dependencies"
+    You can stop reading if you're using Lambda Layer.
+
+    [Tracer](./core/tracer.md){target="_blank"}, [Validation](./utilities/validation.md){target="_blank"} and [Parser](./utilities/parser.md){target="_blank"} require additional dependencies. If you prefer to install all of them, use `pip install aws-lambda-powertools[all]`.
+
 ???+ hint "Support this project by using Lambda Layers :heart:"
     Lambda Layers allow us to understand who uses this library in a non-intrusive way. This helps us justify and gain future investments for other Lambda Powertools languages.
 
     When using Layers, you can add Lambda Powertools as a dev dependency (or as part of your virtual env) to not impact the development process.
 
+### Local development
+
+Powertools relies on the AWS SDK bundled in the Lambda runtime. This helps us achieve an optimal package size and initialization.
+
+This means you need to add AWS SDK as a development dependency (not as a production dependency).
+
+* **Pip**: `pip install aws-lambda-powertools[aws-sdk]`
+* **Poetry**: `poetry add aws-lambda-powertools[aws-sdk] --dev`
+* **Pipenv**: `pipenv install --dev "aws-lambda-powertools[aws-sdk]"`
+
+???+ note "Local emulation"
+    If you're running your code locally with [AWS SAM CLI](https://github.com/aws/aws-sam-cli){target="_blank"}, and not with your Python/IDE interpreter directly, this is not necessary. SAM CLI already brings the AWS SDK in its emulation image.
+
 ### Lambda Layer
 
 [Lambda Layer](https://docs.aws.amazon.com/lambda/latest/dg/configuration-layers.html){target="_blank"} is a .zip file archive that can contain additional code, pre-packaged dependencies, data,  or configuration files. Layers promote code sharing and separation of responsibilities so that you can iterate faster on writing business logic.

docs/utilities/parser.md

@@ -1,5 +1,5 @@
 ---
-title: Parser
+title: Parser (Pydantic)
 description: Utility
 ---
 <!-- markdownlint-disable MD043 -->
@@ -12,20 +12,25 @@ This utility provides data parsing and deep validation using [Pydantic](https://
 * Built-in envelopes to unwrap, extend, and validate popular event sources payloads
 * Enforces type hints at runtime with user-friendly errors
 
-**Extra dependency**
+## Getting started
 
-???+ warning
+### Install
+
+!!! info "This is not necessary if you're installing Powertools via [Lambda Layer](../index.md#lambda-layer){target="_blank"}"
+
+Add `aws-lambda-powertools[parser]` as a dependency in your preferred tool: _e.g._, _requirements.txt_, _pyproject.toml_.
 
+This will ensure you have the required dependencies before using Parser.
+
+???+ warning
     This will increase the compressed package size by >10MB due to the Pydantic dependency.
 
     To reduce the impact on the package size at the expense of 30%-50% of its performance [Pydantic can also be
     installed without binary files](https://pydantic-docs.helpmanual.io/install/#performance-vs-package-size-trade-off):
 
-    `SKIP_CYTHON=1 pip install --no-binary pydantic aws-lambda-powertools[pydantic]`
+	Pip example: `SKIP_CYTHON=1 pip install --no-binary pydantic aws-lambda-powertools[parser]`
 
-Install parser's extra dependencies using **`pip install aws-lambda-powertools[pydantic]`**.
-
-## Defining models
+### Defining models
 
 You can define models to parse incoming events by inheriting from `BaseModel`.
 
@@ -47,11 +52,11 @@ class Order(BaseModel):
 
 These are simply Python classes that inherit from BaseModel. **Parser** enforces type hints declared in your model at runtime.
 
-## Parsing events
+### Parsing events
 
 You can parse inbound events using **event_parser** decorator, or the standalone `parse` function. Both are also able to parse either dictionary or JSON string as an input.
 
-### event_parser decorator
+#### event_parser decorator
 
 Use the decorator for fail fast scenarios where you want your Lambda function to raise an exception in the event of a malformed payload.
 
@@ -104,7 +109,7 @@ handler(event=payload, context=LambdaContext())
 handler(event=json.dumps(payload), context=LambdaContext()) # also works if event is a JSON string
 ```
 
-### parse function
+#### parse function
 
 Use this standalone function when you want more control over the data validation process, for example returning a 400 error for malformed payloads.
 
@@ -149,7 +154,7 @@ def my_function():
 		}
 ```
 
-## Built-in models
+### Built-in models
 
 Parser comes with the following built-in models:
 
@@ -171,7 +176,7 @@ Parser comes with the following built-in models:
 | **KafkaSelfManagedEventModel**  | Lambda Event Source payload for self managed Kafka payload         |
 | **KafkaMskEventModel**          | Lambda Event Source payload for AWS MSK payload                    |
 
-### extending built-in models
+#### extending built-in models
 
 You can extend them to include your own models, and yet have all other known fields parsed along the way.
 
@@ -236,7 +241,9 @@ for order_item in ret.detail.items:
 3. Defined how part of our EventBridge event should look like by overriding `detail` key within our `OrderEventModel`
 4. Parser parsed the original event against `OrderEventModel`
 
-## Envelopes
+## Advanced
+
+### Envelopes
 
 When trying to parse your payloads wrapped in a known structure, you might encounter the following situations:
 
@@ -294,7 +301,7 @@ def handler(event: UserModel, context: LambdaContext):
 3. Parser parsed the original event against the EventBridge model
 4. Parser then parsed the `detail` key using `UserModel`
 
-### Built-in envelopes
+#### Built-in envelopes
 
 Parser comes with the following built-in envelopes, where `Model` in the return section is your given model.
 
@@ -312,7 +319,7 @@ Parser comes with the following built-in envelopes, where `Model` in the return
 | **LambdaFunctionUrlEnvelope** | 1. Parses data using `LambdaFunctionUrlModel`. <br/> 2. Parses `body` key using your model and returns it.                                                                                                  | `Model`                            |
 | **KafkaEnvelope**             | 1. Parses data using `KafkaRecordModel`. <br/> 2. Parses `value` key using your model and returns it.                                                                                                       | `Model`                            |
 
-### Bringing your own envelope
+#### Bringing your own envelope
 
 You can create your own Envelope model and logic by inheriting from `BaseEnvelope`, and implementing the `parse` method.
 
@@ -377,7 +384,7 @@ Here's a snippet of how the EventBridge envelope we demonstrated previously is i
 3. Then, we parsed the incoming data with our envelope to confirm it matches EventBridge's structure defined in `EventBridgeModel`
 4. Lastly, we call `_parse` from `BaseEnvelope` to parse the data in our envelope (.detail) using the customer model
 
-## Data model validation
+### Data model validation
 
 ???+ warning
     This is radically different from the **Validator utility** which validates events against JSON Schema.
@@ -394,7 +401,7 @@ Keep the following in mind regardless of which decorator you end up using it:
 * You must raise either `ValueError`, `TypeError`, or `AssertionError` when value is not compliant
 * You must return the value(s) itself if compliant
 
-### validating fields
+#### validating fields
 
 Quick validation to verify whether the field `message` has the value of `hello world`.
 
@@ -439,7 +446,7 @@ class HelloWorldModel(BaseModel):
 parse(model=HelloWorldModel, event={"message": "hello universe", "sender": "universe"})
 ```
 
-### validating entire model
+#### validating entire model
 
 `root_validator` can help when you have a complex validation mechanism. For example finding whether data has been omitted, comparing field values, etc.
 
@@ -470,7 +477,7 @@ parse(model=UserModel, event=payload)
 ???+ info
     You can read more about validating list items, reusing validators, validating raw inputs, and a lot more in <a href="https://pydantic-docs.helpmanual.io/usage/validators/">Pydantic's documentation</a>.
 
-## Advanced use cases
+### Advanced use cases
 
 ???+ tip "Tip: Looking to auto-generate models from JSON, YAML, JSON Schemas, OpenApi, etc?"
     Use Koudai Aono's [data model code generation tool for Pydantic](https://github.com/koxudaxi/datamodel-code-generator)
@@ -535,55 +542,3 @@ If what you're trying to use isn't available as part of the high level import sy
 ```python title="Pydantic import escape hatch"
 from aws_lambda_powertools.utilities.parser.pydantic import <what you'd like to import'>
 ```
-
-**What is the cold start impact in bringing this additional dependency?**
-
-No significant cold start impact. It does increase the final uncompressed package by **71M**, when you bring the additional dependency that parser requires.
-
-Artillery load test sample against a [hello world sample](https://github.com/aws-samples/cookiecutter-aws-sam-python) using Tracer, Metrics, and Logger with and without parser.
-
-**No parser**
-
-???+ info
-    **Uncompressed package size**: 55M, **p99**: 180.3ms
-
-```javascript
-Summary report @ 14:36:07(+0200) 2020-10-23
-Scenarios launched:  10
-Scenarios completed: 10
-Requests completed:  2000
-Mean response/sec: 114.81
-Response time (msec):
-    min: 54.9
-    max: 1684.9
-    median: 68
-    p95: 109.1
-    p99: 180.3
-Scenario counts:
-    0: 10 (100%)
-Codes:
-    200: 2000
-```
-
-**With parser**
-
-???+ info
-    **Uncompressed package size**: 128M, **p99**: 193.1ms
-
-```javascript
-Summary report @ 14:29:23(+0200) 2020-10-23
-Scenarios launched:  10
-Scenarios completed: 10
-Requests completed:  2000
-Mean response/sec: 111.67
-Response time (msec):
-    min: 54.3
-    max: 1887.2
-    median: 66.1
-    p95: 113.3
-    p99: 193.1
-Scenario counts:
-    0: 10 (100%)
-Codes:
-    200: 2000
-```

docs/utilities/validation.md

@@ -18,6 +18,14 @@ This utility provides JSON Schema validation for events and responses, including
 ???+ tip
     All examples shared in this documentation are available within the [project repository](https://github.com/awslabs/aws-lambda-powertools-python/tree/develop/examples){target="_blank"}.
 
+### Install
+
+!!! info "This is not necessary if you're installing Powertools via [Lambda Layer](../index.md#lambda-layer){target="_blank"}"
+
+Add `aws-lambda-powertools[validation]` as a dependency in your preferred tool: _e.g._, _requirements.txt_, _pyproject.toml_.
+
+This will ensure you have the required dependencies before using Validation.
+
 You can validate inbound and outbound events using [`validator` decorator](#validator-decorator).
 
 You can also use the standalone `validate` function, if you want more control over the validation process such as handling a validation error.