mirror of
https://github.com/nodejs/node.git
synced 2025-05-19 00:19:36 +00:00
2d441493a4
PR-URL: https://github.com/nodejs/io.js/pull/2286 Reviewed-By: Roman Reiss <me@silverwind.io>
57 lines
1.6 KiB
Markdown
57 lines
1.6 KiB
Markdown
# tryit
|
|
|
|
Tiny module wrapping try/catch in JavaScript.
|
|
|
|
It's *literally 11 lines of code*, [just read it](tryit.js) that's all the documentation you'll need.
|
|
|
|
|
|
## install
|
|
|
|
```
|
|
npm install tryit
|
|
```
|
|
|
|
## usage
|
|
|
|
What you'd normally do:
|
|
```js
|
|
try {
|
|
dangerousThing();
|
|
} catch (e) {
|
|
console.log('something');
|
|
}
|
|
```
|
|
|
|
With try-it (all it does is wrap try-catch)
|
|
```js
|
|
var tryit = require('tryit');
|
|
|
|
tryit(dangerousThing);
|
|
```
|
|
|
|
You can also handle the error by passing a second function
|
|
```js
|
|
tryit(dangerousThing, function (e) {
|
|
if (e) {
|
|
console.log('do something');
|
|
}
|
|
})
|
|
```
|
|
|
|
The second function follows error-first pattern common in node. So if you pass a callback it gets called in both cases. But will have an error as the first argument if it fails.
|
|
|
|
## WHAT? WHY DO THIS!?
|
|
|
|
Primary motivation is having a clean way to wrap things that might fail, where I don't care if it fails. I just want to try it.
|
|
|
|
This includes stuff like blindly reading/parsing stuff from localStorage in the browser. If it's not there or if parsing it fails, that's fine. But I don't want to leave a bunch of empty `catch (e) {}` blocks in the code.
|
|
|
|
Obviously, this is useful any time you're going to attempt to read some unknown data structure.
|
|
|
|
In addition, my understanding is that it's hard for JS engines to optimize code in try blocks. By actually passing the code to be executed into a re-used try block, we can avoid having to have more than a single try block in our app. Again, this is not a primary motivation, just a potential side benefit.
|
|
|
|
|
|
## license
|
|
|
|
[MIT](http://mit.joreteg.com/)
|