25 lines
1.2 KiB
Markdown
25 lines
1.2 KiB
Markdown
|
# image-proxy
|
||
|
|
|
||
|
|
An image proxy server that resizes images before serving them.
|
||
|
|
Intended to only be used with images that already have a consistent size (such as already processed avatars).
|
||
|
|
|
||
|
|
## Configuration
|
||
|
|
|
||
|
|
Configuration is done in a `config.json` file, with the following keys:
|
||
|
|
|
||
|
|
- `port`: the port the server will listen on
|
||
|
|
- `debug`: whether the server will log all request paths and durations
|
||
|
|
- `proxy_target`: the target for the reverse proxy, used when the requested size is the image's actual size.
|
||
|
|
Should be the full base URL of the target server, without a trailing slash.
|
||
|
|
- `patterns`: an array of patterns to accept
|
||
|
|
- `input`: (string) the input regex. This should start with a `/` and use unnamed capture groups.
|
||
|
|
- `output`: (string) the URL to rewrite the link to, using `$1` `$2` (etc) as placeholders for the capture groups.
|
||
|
|
- `cache_key`: (string) the file name used for cached files. Must be unique across all patterns.
|
||
|
|
- `content_type`: (string) the `Content-Type` header to return with requests.
|
||
|
|
- `sizes`: (array of numbers) the acceptable sizes for this pattern.
|
||
|
|
The first size in the list is the default, and will be passed through unmodified.
|
||
|
|
|
||
|
|
## To-do list
|
||
|
|
|
||
|
|
- Clean up old, unused cache entries
|