Flask-S3 ******** .. module:: flask_s3 Flask-S3 allows you to easily serve all your `Flask`_ application's static assets from `Amazon S3`_, without having to modify your templates. .. _Amazon S3: http://aws.amazon.com/s3 .. _Flask: http://flask.pocoo.org/ How it works ============ Flask-S3 has two main functions: 1. Walk through your application's static folders, gather all your static assets together, and upload them to a bucket of your choice on S3; 2. Replace the URLs that Flask's :func:`flask.url_for` function would insert into your templates, with URLs that point to the static assets in your S3 bucket. The process of gathering and uploading your static assets to S3 need only be done once, and your application does not need to be running for it to work. The location of the S3 bucket can be inferred from Flask-S3 `settings`_ specified in your Flask application, therefore when your application is running there need not be any communication between the Flask application and Amazon S3. Internally, every time ``url_for`` is called in one of your application's templates, `flask_s3.url_for` is instead invoked. If the endpoint provided is deemed to refer to static assets, then the S3 URL for the asset specified in the `filename` argument is instead returned. Otherwise, `flask_s3.url_for` passes the call on to `flask.url_for`. Installation ============ If you use pip then installation is simply:: $ pip install flask-s3 or, if you want the latest github version:: $ pip install git+git://github.com/e-dard/flask-s3.git You can also install Flask-S3 via Easy Install:: $ easy_install flask-s3 Dependencies ------------ Aside from the obvious dependency of Flask itself, Flask-S3 makes use of the `boto`_ library for uploading assets to Amazon S3. **Note**: Flask-S3 currently only supports applications that use the `jinja2`_ templating system. .. _boto: http://docs.pythonboto.org/en/latest/ .. _jinja2: http://jinja.pocoo.org/docs/ Using Flask-S3 ============== Flask-S3 is incredibly simple to use. In order to start serving your Flask application's assets from Amazon S3, the first thing to do is let Flask-S3 know about your :class:`flask.Flask` application object. .. code-block:: python from flask import Flask from flask_s3 import FlaskS3 app = Flask(__name__) app.config['FLASKS3_BUCKET_NAME'] = 'mybucketname' s3 = FlaskS3(app) In many cases, however, one cannot expect a Flask instance to be ready at import time, and a common pattern is to return a Flask instance from within a function only after other configuration details have been taken care of. In these cases, Flask-S3 provides a simple function, ``init_app``, which takes your application as an argument. .. code-block:: python from flask import Flask from flask_s3 import FlaskS3 s3 = FlaskS3() def start_app(): app = Flask(__name__) s3.init_app(app) return app In terms of getting your application to use external Amazon S3 URLs when referring to your application's static assets, passing your ``Flask`` object to the ``FlaskS3`` object is all that needs to be done. Once your app is running, any templates that contained relative static asset locations, will instead contain hosted counterparts on Amazon S3. Uploading your Static Assets ---------------------------- You only need to upload your static assets to Amazon S3 once. Of course, if you add or modify your existing assets then you will need to repeat the uploading process. Uploading your static assets from a Python console is as simple as follows. .. code-block:: python >>> import flask_s3 >>> from my_application import app >>> flask_s3.create_all(app) >>> Flask-S3 will proceed to walk through your application's static assets, including those belonging to *registered* `blueprints`_, and upload them to your Amazon S3 bucket. .. _blueprints: http://flask.pocoo.org/docs/blueprints/ Static Asset URLs ~~~~~~~~~~~~~~~~~ Within your bucket on S3, Flask-S3 replicates the static file hierarchy defined in your application object and any registered blueprints. URLs generated by Flask-S3 will look like the following: ``/static/foo/style.css`` becomes ``https://mybucketname.s3.amazonaws.com/static/foo/style.css``, assuming that ``mybucketname`` is the name of your S3 bucket, and you have chosen to have assets served over HTTPS. Setting Custom HTTP Headers ~~~~~~~~~~~~~~~~~ To set custom HTTP headers on the files served from S3 specify what headers you want to use with the `FLASKS3_HEADERS` option. .. code-block:: python FLASKS3_HEADERS = { 'Expires': 'Thu, 15 Apr 2010 20:00:00 GMT', 'Cache-Control': 'max-age=86400', } See `Yahoo!`_ more information on how to set good values for your headers. .. _Yahoo!: http://developer.yahoo.com/performance/rules.html#expires .. _settings: .. _configuration: Flask-S3 Options ---------------- Within your Flask application's settings you can provide the following settings to control the behaviour of Flask-S3. None of the settings are required, but if not present, some will need to be provided when uploading assets to S3. =========================== =================================================== `AWS_ACCESS_KEY_ID` Your AWS access key. This does not need to be stored in your configuration if you choose to pass it directly when uploading your assets. `AWS_SECRET_ACCESS_KEY` Your AWS secret key. As with the access key, this need not be stored in your configuration if passed in to `create_all`. `FLASKS3_BUCKET_DOMAIN` The domain part of the URI for your S3 bucket. You probably won't need to change this. **Default:** ``u's3.amazonaws.com'`` `FLASKS3_CDN_DOMAIN` AWS makes it easy to attach CloudFront to an S3 bucket. If you want to use this or another CDN, set the base domain here. This is distinct from the `FLASKS3_BUCKET_DOMAIN` since it will not include the bucket name in the base url. `FLASKS3_BUCKET_NAME` The desired name for your Amazon S3 bucket. Note: the name will be visible in all your assets' URLs. `FLASKS3_REGION` The AWS region to host the bucket in; an empty string indicates the default region should be used, which is the US Standard region. Possible location values include: `'DEFAULT'`, `'EU'`, `'USWest'`, `'APSoutheast'` `FLASKS3_URL_STYLE` Set to `'host'` to use virtual-host-style URLs, e.g. ``bucketname.s3.amazonaws.com``. Set to `'path'` to use path-style URLs, e.g. ``s3.amazonaws.com/bucketname``. **Default:** `'host'` `FLASKS3_USE_HTTPS` Specifies whether or not to serve your assets stored in S3 over HTTPS. Can be overriden per url, by using the `_scheme` argument as per usual Flask `url_for`. **Default:** `True` `FLASKS3_ACTIVE` This setting allows you to toggle whether Flask-S3 is active or not. When set to `False` your application's templates will revert to including static asset locations determined by `flask.url_for`. **Default:** `True` **Note**: if you run your application in `debug`_ mode (and `FLASKS3_DEBUG` is `False` - see next item), `FLASKS3_ACTIVE` will be changed to `False`. This allows the `FLASKS3_ACTIVE` config variable to be the definitive check as to whether `flask_s3.url_for` is overriding `flask.url_for`. `FLASKS3_DEBUG` By default, Flask-S3 will be switched off when running your application in `debug`_ mode, so that your templates include static asset locations specified by `flask.url_for`. If you wish to enable Flask-S3 in debug mode, set this value to `True`. **Note**: if `FLASKS3_ACTIVE` is set to `False` then templates will always include asset locations specified by `flask.url_for`. `FLASKS3_HEADERS` Sets custom headers to be sent with each file to S3. **Default:** `{}` `FLASKS3_FILEPATH_HEADERS` Sets custom headers for files whose filepath matches certain regular expressions. (Note that this cannot be used for CORS, that must be set per S3 bucket using an XML config string.) E.g. to add custom metadata when serving text files, set this to: `{r'\.txt$':` ` {'Texted-Up-By': 'Mister Foo'}` `}` **Default:** `{}` `FLASKS3_ONLY_MODIFIED` Only upload files that have been modified since last upload to S3. SHA-1 file hashes are used to compute file changes. You can delete `.file-hashes` from your S3 bucket to force all files to upload again. Defaults to `False`. `FLASKS3_GZIP` Compress all assets using GZIP and set the corresponding Content-Type and Content-Encoding headers on the S3 files. Defaults to `False`. `FLASKS3_GZIP_ONLY_EXTS` A list of file extensions that should be gzipped. ``FLASKS3_GZIP`` should be ``True`` for this to take effect. If mentioned and non-empty, then only files with the specified extensions are gzipped. Defaults to empty list, meaning all files will be gzipped. Eg:- ``['.js', '.css']`` will gzip only js and css files. `FLASKS3_FORCE_MIMETYPE` Always set the Content-Type header on the S3 files irrespective of gzipping. Defaults to `False`. =========================== =================================================== .. _debug: http://flask.pocoo.org/docs/config/#configuration-basics API Documentation ================= Flask-S3 is a very simple extension. The few exposed objects, methods and functions are as follows. The FlaskS3 Object ------------------ .. autoclass:: FlaskS3 .. automethod:: init_app S3 Interaction -------------- .. autofunction:: create_all .. autofunction:: url_for