Python API Client Double Encoding excludePartOfSpeech Parameter

[Elahd Bar-Shai]

Hey all,

The Wordnik Python API fails when including a comma separated list in the WordsApi.getRandomWord(excludePartOfSpeech) parameter. The failure symptom is that the server returns a 500 error.

This issue has been reported before, but only implicitly, and I can't find a discussion of root cause and remedies. The issue is that the API client library double encodes commas to %252C instead of %2C (comma encodes to %2C, %2C encodes to %252C). It looks like the server is not double decoding.

The below is my code (in Python):

apiUrl = 'http://api.wordnik.com/v4'
apiKey = '[redacted]'
client = swagger.ApiClient(apiKey, apiUrl)

wordsApi = WordsApi.WordsApi(client)
randomWord = wordsApi.getRandomWord(
 hasDictionaryDef="true",
 includePartOfSpeech="noun",
 excludePartOfSpeech="proper-noun,proper-noun-plural,proper-noun-posessive,given-name,family-name",
 minCorpusCount=10000,
 maxCorpusCount=-1,
 minDictionaryCount=3,
 maxDictionaryCount=-1,
 minLength=5,
 maxLength=5)

The Wordnik Python API library generates this malformed request:

GET http://api.wordnik.com/v4/words.json/randomWord?maxDictionaryCount=-1&minLength=5&minCorpusCount=10000&hasDictionaryDef=true&maxCorpusCount=-1&minDictionaryCount=3&maxLength=5&includePartOfSpeech=noun&excludePartOfSpeech=proper-noun%252Cproper-noun-plural%252Cproper-noun-posessive%252Cgiven-name%252Cfamily-name

The request should look like this:

GET http://api.wordnik.com/v4/words.json/randomWord?maxDictionaryCount=-1&minLength=5&minCorpusCount=10000&hasDictionaryDef=true&maxCorpusCount=-1&minDictionaryCount=3&maxLength=5&includePartOfSpeech=noun&excludePartOfSpeech=proper-noun%2Cproper-noun-plural%2Cproper-noun-posessive%2Cgiven-name%2Cfamily-name

I could submit a pull request through GitHub to fix the issue, but this seems like a hack-y solution considering that the Wordnik Python API is automatically generated by Swagger. Can anyone on the Wordnik team address this issue?

Thanks,

Elahd