Usage
Basic usage
The library’s main interface is the Mixcloud
class.
Import it directly from aiomixcloud
and use it as an asynchronous
context manager. Pass a key (URL part corresponding to a unique API resource)
to its get()
method to fetch information
about that resource
:
from aiomixcloud import Mixcloud
async with Mixcloud() as mixcloud:
user = await mixcloud.get('bob')
Result data is available both as attributes and dictionary items:
user.city # 'London'
user['favorite_count'] # 38
user.pictures['medium'] # 'https://thumbnailer.mixcloud.com/unsafe/10...'
Datetime data gets automatically converted to
datetime
objects:
user.updated_time # datetime.datetime(2018, 3, 10, 7, 32, tzinfo=tzutc())
The discover()
shortcut returns information
about a tag:
tag = await mixcloud.discover('jazz')
tag['name'] # 'Jazz shows'
Resource
objects are dict-like and can be
iterated over:
# Using `tag` from previous snippet
for key, value in tag.items():
print(key, value)
Original dict
is stored in their data attribute:
tag.data # {'url': 'https://www.mixcloud.com/discover/jazz/', 'type': ...
Listing
The listing methods consist of:
and are responsible for downloading a list of resources:
popular = await mixcloud.popular()
popular # <ResourceList 'Popular Cloudcasts'>
# Index the list
popular[0] # <Resource: Cloudcast '/alice/pop-hits-episode-42/'>
# Items are resources
popular[3]['audio_length'] # 1676
# Iterate
for p in popular:
p.url # 'https://www.mixcloud.com/...'
Items of a resource list can also be accessed by their key:
popular['chris/funky-mix'] # <Resource: Cloudcast '/chris/funky-mix/'>
The search()
method, which can accept a
type argument, among 'cloudcast'
(default), 'user'
and 'tag'
,
returns resources matching the given term:
rock_cloudcasts = await mixcloud.search('rock')
johns = await mixcloud.search('john', type='user')
Listing methods can accept pagination arguments the API itself defines:
offset, limit, since and until. The former two concern net numbers
(counts) and the latter two can be UNIX timestamps, human-readable strings or
datetime
objects. Alternatively, instead any of those,
a page argument can be specified (zero-indexed), giving 20 results per page
(unless the per_page argument indicates otherwise):
hot = await mixcloud.hot(offset=40, limit=80)
new = await mixcloud.new(since='2018 Feb 12 13:00:00',
until='2019 March 28 21:15:04')
some_jazz = await mixcloud.search('jazz', page=2)
metal_music = await mixcloud.search('metal', page=4, per_page=30)
Note: Timezone-naive datetime values (either human-readable strings
or datetime
objects) will be treated as being in the
current local timezone.
Resource lists
have a
previous()
and a
next()
method which return the previous
and the next page of the current resource list, respecitvely. If there is no
such page available, these methods return None
:
older_metal = await metal_music.previous()
older_metal # <ResourceList>
newer_metal = await metal_music.next()
newer_metal # <ResourceList>
await newer_metal.next() # This returns None
When responding with a resource list
, the API sends most of the information
for each resource, but not all of it. That is an example of dealing with
non-full resources. Again, in a resource list, some of the data included,
represent resources related to each list item, for example each item in a
cloudcast list contains information about the user who uploaded the cloudcast.
The information about that user is also incomplete, making it another case of
a non-full resource. The load()
method of
Resource
objects can be used to load the full
information of a non-full resource:
# Using `hot` from previous snippet
some_hot_cloudcast = hot[5]
some_hot_cloudcast.description # raises AttributeError
await some_hot_cloudcast.load()
some_hot_cloudcast.description # 'The greatest set of all time...'
load()
also returns the freshly-loaded
object so it can be used in chained calls, something that can find elegant
application in synchronous library usage.
Connections
API resources can have sub-resources, or, connections, that is other API resources associated with (or, “owned” by) them. For example, a user can have followers, i.e a user resource has followers as a connection, which are other user resources themselves. The connections of a resource become available through methods of it, named after the respective connection names:
peter = await mixcloud.get('peter')
his_followers = await peter.followers()
his_followers # <ResourceList "Peter's followers">
nice_cloudcast = await mixcloud.get('luke/a-nice-mix')
comments = await nice_cloudcast.comments()
for comment in comments:
comment # <Resource: Comment '/comments/cr/.../'>
comment.comment # 'Nice set, keep up the good work!'
Embedding
Embedding information and HTML code for a cloudcast can be retrieved through
the embed_json()
and
embed_html()
methods, being able to take
width, height and color as arguments:
json_embed_info = await mixcloud.embed_json('someuser/the-best-mix')
html_embed_code = await mixcloud.embed_html('someuser/the-best-mix',
width=300, height=150)
oEmbed information for a resource (previous arguments applicable here as well) is available through:
oembed_info = await mixcloud.oembed(resource_key)
Actions
Authorized usage also enables actions, a group of methods about doing and undoing certain actions on specific resources:
Each of them takes a resource key as an argument (the two methods on the first row target a user, the rest of them target a cloudcast):
data = await mixcloud.follow('bob')
data['result']['message'] # 'Now following bob'
data = await mixcloud.unrepost('alice/fun-times-ep-25')
data.result.success # True
Uploading
Making authorized use of the API allows uploading cloudcasts and
editing existing uploads. Both upload()
and
edit()
share the following optional
arguments: picture (filename), description (text), tags (sequence of
strings), sections (sequence of mappings) and some fields available only to
pro accounts: publish_date (UNIX timestamp, human-readable string or
datetime
object), disable_comments (boolean),
hide_stats (boolean) and unlisted (boolean).
The upload()
method takes two positional
arguments, mp3 (filename) and name (string):
data = await mixcloud.upload('perfectmix.mp3', 'Perfect Mix',
picture='perfectpic.jpg',
description='The perfect house mix',
tags=['house', 'deep'],
sections=[{'chapter': 'Intro',
'start_time': 0},
{'artist': 'Somebody',
'song': 'Some song',
'start_time': 60},
{'artist': 'Cool DJ',
'song': 'Cool track',
'start_time': 240}])
data.result['success'] # True
edit()
takes a key positional argument and
a name optional argument:
data = await mixcloud.edit('amikrop/perfect-mix', name='The Perfect Mix',
description='The best house mix, right for summer',
tags=['house', 'deep', 'summer'])
data['result'].success # True
Targeting
Methods of Mixcloud
that target a specific resource
(and thus, take a key as first argument) are also available as methods of
Resource
objects:
someone = await mixcloud.get('certainuser')
await someone.unfollow() # {'result': ...
mix = await mixcloud.get('auser/acloudcast')
await mix.favorite() # {'result': ...
await mix.embed_html() # '<iframe width="100%" height=...'
my_mix = await mixcloud.get('amikrop/perfect-mix')
await my_mix.edit(description='The best house mix, perfect for summer!',
tags=['house', 'deep',
'summer', 'smooth']) # {'result': ...
Those methods include the actions, the embedding methods and
edit()
.
Synchronous mode
All the functionality of the package is also available for synchronous
(i.e blocking) usage. MixcloudSync
and
MixcloudOAuthSync
provide the same interface as
their asynchronous versions, with all the coroutine methods now being classic
methods. Context management becomes synchronous and methods of returned
objects are synchronous as well:
from aiomixcloud.sync import MixcloudOAuthSync, MixcloudSync
with MixcloudSync() as mixcloud:
oauth = MixcloudOAuthSync(client_id=CLIENT_ID,
client_secret=CLIENT_SECRET,
redirect_uri=REDIRECT_URI,
mixcloud=mixcloud)
# ... After getting user's permission and storing `code` ...
mixcloud.access_token = oauth.access_token(code)
some_cloudcast = mixcloud.get('someuser/somemix')
some_cloudcast.repost() # {'result': ...
# Chained calls
some_cloudcast.similar()[0].load().picture_primary_color # '02f102'