75 lines
3.4 KiB
Markdown
75 lines
3.4 KiB
Markdown
# Belatuk Body Parser
|
|
|
|
|
|
[](https://dart.dev/null-safety)
|
|
[](https://github.com/dart-backend/belatuk-common-utilities/blob/main/packages/body_parser/LICENSE)
|
|
|
|
**Replacement of `package:body_parser` with breaking changes to support NNBD.**
|
|
|
|
Parse request bodies and query strings in Dart, as well multipart/form-data uploads. No external dependencies required.
|
|
|
|
This is the request body parser powering the [Angel3 framework](https://pub.dev/packages/angel3_framework). If you are looking for a server-side solution with dependency injection, WebSockets, and more, then I highly recommend it as your first choice. Bam!
|
|
|
|
## Contents
|
|
|
|
- [Belatuk Body Parser](#belatuk-body-parser)
|
|
- [Contents](#contents)
|
|
- [About](#about)
|
|
- [Installation](#installation)
|
|
- [Usage](#usage)
|
|
- [Custom Body Parsing](#custom-body-parsing)
|
|
|
|
### About
|
|
|
|
I needed something like Express.js's `body-parser` module, so I made it here. It fully supports JSON requests. x-www-form-urlencoded fully supported, as well as query strings. You can also include arrays in your query, in the same way you would for a PHP application. Full file upload support will also be present by the production 1.0.0 release.
|
|
|
|
A benefit of this is that primitive types are automatically deserialized correctly. As in, if you have a `hello=1.5` request, then `body['hello']` will equal `1.5` and not `'1.5'`. A very semantic difference, yes, but it relieves stress in my head.
|
|
|
|
### Installation
|
|
|
|
To install Body Parser for your Dart project, simply add body_parser to your pub dependencies.
|
|
|
|
dependencies:
|
|
belatuk_body_parser: ^4.0.0
|
|
|
|
### Usage
|
|
|
|
Body Parser exposes a simple class called `BodyParseResult`. You can easily parse the query string and request body for a request by calling `Future<BodyParseResult> parseBody`.
|
|
|
|
```dart
|
|
import 'dart:convert';
|
|
import 'package:belatuk_body_parser/belatuk_body_parser.dart';
|
|
|
|
main() async {
|
|
// ...
|
|
await for (HttpRequest request in server) {
|
|
request.response.write(JSON.encode(await parseBody(request).body));
|
|
await request.response.close();
|
|
}
|
|
}
|
|
```
|
|
|
|
You can also use `buildMapFromUri(Map, String)` to populate a map from a URL encoded string.
|
|
|
|
This can easily be used with a library like [Angel3 JSON God](https://pub.dev/packages/angel3_json_god) to build structured JSON/REST APIs. Add validation and you've got an instant backend.
|
|
|
|
```dart
|
|
MyClass create(HttpRequest request) async {
|
|
return god.deserialize(await parseBody(request).body, MyClass);
|
|
}
|
|
```
|
|
|
|
### Custom Body Parsing
|
|
|
|
In cases where you need to parse unrecognized content types, `body_parser` won't be of any help to you on its own. However, you can use the `originalBuffer` property of a `BodyParseResult` to see the original request buffer. To get this functionality, pass `storeOriginalBuffer` as `true` when calling `parseBody`.
|
|
|
|
For example, if you wanted to [parse GraphQL queries within your server](https://github.com/dukefirehawk/graphql_dart)...
|
|
|
|
```dart
|
|
app.get('/graphql', (req, res) async {
|
|
if (req.headers.contentType.mimeType == 'application/graphql') {
|
|
var graphQlString = String.fromCharCodes(req.originalBuffer);
|
|
// ...
|
|
}
|
|
});
|
|
```
|