python-leetcode 1.0.1

Leetcode API

Leetcode API implementation.

This implements methods that are available publicly for leetcode. There is already an implementation of them in form of CLI [1], but it has a list of disadvantages.

1. It is written in JS

2. Even for JS in order to reuse it you have to invoke it via CLI

3. It is not supported very well, so authentication doesn't work anymore

So I have decided to create my own implementation and here is it.

Using the swagger file you'll be able to generate the code for any language you like and start using leetcode API directly from your code.

Just keep in mind that swagger doesn't really support cookie auth, which is needed in order to use LC API. If you use python you can just use the generated code from this repo. Otherwise you'll have to implement something like fix_cookies.patch for your target language.

Minimal working example

First set up a virtualenv

virtualenv -p python3 leetcode
. leetcode/bin/activate
pip3 install python-leetcode

Then in python shell initialize the client

import leetcode

# Get the next two values from your browser cookies
csrf_token = "xxx"
leetcode_session = "yyy"

configuration = leetcode.Configuration()

configuration.api_key["x-csrftoken"] = csrf_token
configuration.api_key["csrftoken"] = csrf_token
configuration.api_key["LEETCODE_SESSION"] = leetcode_session
configuration.api_key["Referer"] = "https://leetcode.com"
configuration.debug = False

api_instance = leetcode.DefaultApi(leetcode.ApiClient(configuration))

Now once the client is initilized, you can start performing actual queries

graphql_request = leetcode.GraphqlQuery(
    query="""
      {
        user {
          username
          isCurrentUserPremium
        }
      }
    """,
    variables=leetcode.GraphqlQueryVariables(),
)

print(api_instance.graphql_post(body=graphql_request))

You should get something like that in the response

{'data': {'question': None,
          'user': {'is_current_user_premium': True, 'username': 'omgitspavel'}}}

This confirms you've set up auth correctly.

Advanced example

Now let's try to do something more complicated. For example calculate the percentage of the problems we've solved by topic.

For that we have to acquire the list of all the problems we solved.

api_response = api_instance.api_problems_topic_get(topic="algorithms")

slug_to_solved_status = {
    pair.stat.question__title_slug: True if pair.status == "ac" else False
    for pair in api_response.stat_status_pairs
}

Now for each problem we want to get its tags

import time

from collections import Counter


topic_to_accepted = Counter()
topic_to_total = Counter()


# Take only the first 10 for test purposes
for slug in list(slug_to_solved_status.keys())[:10]:
    time.sleep(1)  # Leetcode has a rate limiter
    
    graphql_request = leetcode.GraphqlQuery(
        query="""
            query getQuestionDetail($titleSlug: String!) {
              question(titleSlug: $titleSlug) {
                topicTags {
                  name
                  slug
                }
              }
            }
        """,
        variables=leetcode.GraphqlQueryVariables(title_slug=slug),
        operation_name="getQuestionDetail",
    )

    api_response = api_instance.graphql_post(body=graphql_request)
    
    for topic in (tag.slug for tag in api_response.data.question.topic_tags):
        topic_to_accepted[topic] += int(slug_to_solved_status[slug])
        topic_to_total[topic] += 1

print(
    list(
        sorted(
            ((topic, accepted / topic_to_total[topic]) for topic, accepted in topic_to_accepted.items()),
            key=lambda x: x[1]
        )
    )
)

The output will look like this:

[('memoization', 0.0),
 ('number-theory', 0.0),
 ('binary-search-tree', 0.0),
 ('quickselect', 0.0),
 ('recursion', 0.0),
 ('suffix-array', 0.0),
 ('topological-sort', 0.0),
 ('shortest-path', 0.0),
 ('trie', 0.0),
 ('geometry', 0.0),
 ('brainteaser', 0.0),
 ('combinatorics', 0.0),
 ('line-sweep', 0.0),
 
 ...
 
 ('union-find', 0.3076923076923077),
 ('linked-list', 0.3333333333333333),
 ('string-matching', 0.3333333333333333),
 ('segment-tree', 0.4),
 ('data-stream', 0.5),
 ('strongly-connected-component', 0.5),
 ('minimum-spanning-tree', 0.6666666666666666),
 ('merge-sort', 1.0),
 ('doubly-linked-list', 1.0)]

So it is clearly visible which topics we should focus on in our preparation. In this case memoization topic is one of the targets for improvement, so I can go to https://leetcode.com/tag/memoization/ and choose a new memoization problem. Or use python to automate the process.

You can find other examples of usage in example.py

Autogenerated by swagger documentation can be found here.

[1] https://github.com/skygragon/leetcode-cli