IMPORTANT: I've maintained UYTS for almost nine months now, but YouTube is tightening their scraper-prevention methods and I don't have the time to work around their frequent changes, so I've taken the decision to archive this project. Please take a look at youtube-search-python for all your YouTube scraping needs from now on!
Unlimited YouTube Search (UYTS) is a quick and easy way to search YouTube from your Python program without the need for the YouTube Data API. It's a Python port of youtube-scrape by Herman Fassett and has the majority of its features.
Simply run pip install unlimited-youtube-search
in the command prompt to install it from the Python Package Index. Alternatively, clone the repo to your PC, navigate to its folder, and run python setup.py install
. Both of these methods will install UYTS and its dependencies.
You'll need to import uyts
at the start of your project in order to use Unlimited YouTube Search. From then on, you can search YouTube using search = uyts.Search('search query')
. This returns a search object which has the attribute results
storing a list of Video, Playlist and Channel objects. For example, you could run print(search.results[0].title)
to see its title. For more information, read on.
Here's a simple program to show how easy it is to search YouTube with UYTS.
import uyts
query = input("Search query: ")
search = uyts.Search(query)
for result in search.results:
print(result)
That's already built in! To host a server, simply run the following code:
from uyts import Server
app = Server()
app.run()
This will host a Flask server on port 80. More information on how to customise this is in the documentation below.
Usage: uyts.Search(query,minResults=0)
Parameters:
query
: the string to search forlanguage
(optional): the language to return search results in, defaults to "en", accepts ISO codes (e.g. "en", "en-gb", "fr-fr"). For best results, use with thecountry
parametercountry
(optional): the region to make the request from, defaults to "GB", accepts ISO codes (e.g. "GB", "US", "FR"). For best results, use with thelanguage
parameterminResults
(optional): the minimum number of results to return. UYTS will continue making requests until it reaches this number or runs out of results. The default value of 0 will make one search request.timeout
(optional): the number of seconds to wait before giving up on the request, defaults to 5
Attributes:
results
: list of search resultsresultsJSON
: JSON object of search resultsquery
: the original search queryresultsCount
: the number of search results returnedmaxResultsCount
: YouTube's estimation of total possible search resultssuggestedSearches
: list of suggested searches related to your query
The following three classes are returned in the search results, and while they can be created yourself, there's pretty much no reason you would want to do that so I haven't included how to do so here. It's self explanatory in the code however.
Attributes:
id
: the ID of the YouTube videotitle
: the title of the YouTube videothumbnail_src
: the URL of the thumbnailviews
: the number of viewsauthor
: the name of the uploaderduration
: the duration of the videoresultType
: the type of result (in this case,video
)accountType
: the type of the author's account, either "verified", "music" or "regular"ToJSON()
: returns the video as a JSON objectToXML()
: returns the video as an XML string
Attributes:
id
: the ID of the playlisttitle
: the title of the playlistthumbnail_src
: the URL of the thumbnaillength
: the number of videos in the playlistauthor
: the name of the creatorresultType
: the type of result (in this case,playlist
)ToJSON()
: returns the playlist as a JSON objectToXML()
: returns the playlist as an XML string
Attributes:
id
: the ID of the channeltitle
: the name of the channelsubs
orsubscriber_count
: the number of subscribers the channel hasresultType
: the type of result (in this case,channel
)accountType
: the type of the author's account, either "verified", "music" or "regular"ToJSON()
: returns the channel as a JSON objectToXML()
: returns the channel as an XML string
The server must be initialised before you can call run()
.
Parameters:
serverName
(optional): the name for the Flask server, defaults touyts-api
serverMessage
(optional): the message that appears on the server homepage, defaults toServer online
rawHTML
(optional): ifTrue
, treatsserverMessage
as raw HTML rather than a string, defaults toFalse
Methods:
run(host="0.0.0.0",port=80)
: runs a Flask server on your local IP on port 80, unless specified otherwise
Attributes:
app
: this is a Flask object for the server. For most use-cases you won't need to directly interact with it. However, if you wanted to deploy this to Heroku you would need to do something likeserver = uyts.Server().app
and then use Gunicorn to run it withweb: gunicorn main:server
for the Procfile as you can't run something from inside a class (e.g.web: gunicorn main:server.app
would be invalid).
Server routes:
/
: main page, should show "Server online" if the server is online/api
: API page, either GET/api/<query>
or/api/<query>/<minResults>
depending on whether you want to specify the minimum results. The response should look like this (but with more results):
[
{
"id": "dQw4w9WgXcQ",
"title": "Rick Astley - Never Gonna Give You Up (Video)",
"thumbnail_src": "http://i.ytimg.com/vi/dQw4w9WgXcQ/maxresdefault.jpg",
"views": "746,623,786 views",
"author": "Official Rick Astley",
"duration": "3:32",
"resultType": "video",
"accountType": "music"
},
{
"id": "UC-lHJZR3Gqxm24_Vd_AJ5Yw",
"title": "PewDiePie",
"subscriber_count": "106M subscribers",
"resultType": "channel",
"accountType": "verified"
}
]