Implement a restful web server in python

[ovidiupw]

Available routes should include:

• GET /api/textprocess/v1/annotate?sentence=$s
• GET /api/textprocess/v1/proofread?sentence=$s

Disambiguation:

• /api/textprocess/ is the base path of the textprocess api
• /v1 is the version of the textprocess api
• $s is a value that will be substituted with an sentence (for example what%20is%20your%20name)

Points to consider:

• any other urls should return a not-found/unauthorized error
• any other verbs should return unauthorized error

[ovidiupw]

Dec 9 2016 - Used Django to setup a local development server and implemented basic route handling [2 hours]

[ovidiupw]

Dec 10 2016 - Setup an EC2 instance and deployed the project to it. Live version can be found at http://ec2-52-212-128-197.eu-west-1.compute.amazonaws.com/api/v1 [4 hours]

[ovidiupw]

Dec 14 2016 - Fixed some request paths and added /api/v1/nltk/part-of-speech route

[ovidiupw]

Dec 15 2016 - Integrated lemma resolution and sentence annotation into the API (api/v1/nltk/sentence-annotation).

[ovidiupw]

Integrated language detection module into the API (api/v1/nltk/detect-language).

[ovidiupw]

Integrated response error message in the given language into the API (api/v1/nltk/error-message)

[ovidiupw]

Integrated anaphora resolution and autocorrect modules into the API

[ovidiupw]

Implementation details:

Framework: Django web framework for python https://www.djangoproject.com/

Hosting service: Amazon EC2 Instance https://aws.amazon.com/ec2/

Integration inside TextProcessing module: Python module communication

Communication outside TextProcessing module: JSON based messages http://www.json.org/

Testing framework: Local and prod environment tests using Postman https://www.getpostman.com/

[ovidiupw]

Integrated sentence keyword resolution into the API. New data available at api/v1/nltk/sentence-annotation in the sentenceKeywords tag