First commit.

.gitignore

@@ -0,0 +1,4 @@
+.DS_Store
+*.pyc
+*.swp
+.ropeproject

Dockerfile

@@ -0,0 +1,13 @@
+FROM ubuntu:14.04
+
+MAINTAINER John Beieler <[email protected]>
+
+RUN apt-get update && apt-get install -y git python-dev python-pip 
+
+ADD . /src
+
+RUN cd /src; pip install -r requirements.txt
+
+EXPOSE 5002
+
+CMD ["python", "/src/app.py"]

README.md

@@ -0,0 +1,38 @@
+hypnos
+======
+
+A RESTful API around the [PETRARCH](https://github.com/openeventdata/petrarch)
+event data coder. Using `docker compose`, this setup also integrates the 
+Stanford [CoreNLP](http://nlp.stanford.edu/software/corenlp.shtml) parser
+using Casey Hilland's [docker container](https://github.com/chilland/ccNLP).
+This setup allows the user to stream texts into the API, rather than the 
+batch mode seen in applications such as the [Phoenix pipeline](https://github.com/openeventdata/phoenix_pipeline).
+
+Running
+-------
+
+Running the system is as simple as using
+
+`docker-compose up`
+
+or 
+
+`docker-compose up -d`
+
+to run in the background.
+
+This assumes that you have `docker-compose` and `docker` installed.
+
+Usage
+-----
+
+```
+headers = {'Content-Type': 'application/json'}
+data = {'text': "At least 37 people are dead after Islamist radical group Boko
+Haram assaulted a town in northeastern Nigeria.", 'id': 'abc123', 'date':
+'20010101'}
+data = json.dumps(data)
+r = requests.get('http://localhost:5002/siesta/extract', data=data,
+                 headers=headers)
+r.json()
+```

app.py

@@ -0,0 +1,78 @@
+import os
+import json
+import requests
+from tornado.ioloop import IOLoop
+from tornado.wsgi import WSGIContainer
+from tornado.httpserver import HTTPServer
+from flask import Flask, jsonify, make_response
+from flask.ext.restful import Api, Resource, reqparse
+
+app = Flask(__name__)
+api = Api(app)
+
+cwd = os.path.abspath(os.path.dirname(__file__))
+
+
+@app.errorhandler(400)
+def bad_request(error):
+    return make_response(jsonify({'error': 'Bad request'}), 400)
+
+
+@app.errorhandler(404)
+def not_found(error):
+    return make_response(jsonify({'error': 'Not found'}), 404)
+
+
+class ExtractAPI(Resource):
+    def __init__(self):
+        self.reqparse = reqparse.RequestParser()
+        self.reqparse.add_argument('text', type=unicode, location='json')
+        self.reqparse.add_argument('id', type=unicode, location='json')
+        self.reqparse.add_argument('date', type=unicode, location='json')
+        super(ExtractAPI, self).__init__()
+
+    def get(self):
+        args = self.reqparse.parse_args()
+        text = args['text']
+        text = text.encode('utf-8')
+        storyid = args['id']
+        date = args['date']
+
+        headers = {'Content-Type': 'application/json'}
+        core_data = json.dumps({'text': text})
+        ccnlp = os.environ['CCNLP_PORT_5000_TCP_ADDR']
+        ccnlp_url = 'http://{}:5000/process'.format(ccnlp)
+        r = requests.post(ccnlp_url, data=core_data, headers=headers)
+        out = r.json()
+
+        event_dict = process_corenlp(out, date, storyid)
+
+        events_data = json.dumps({'events': event_dict})
+        print(events_data)
+        petr = os.environ['PETRARCH_PORT_5001_TCP_ADDR']
+        petr_url = 'http://{}:5001/petrarch/code'.format(petr)
+        events_r = requests.post(petr_url, data=events_data, headers=headers)
+        event_updated = events_r.json()
+
+        return event_updated
+
+
+def process_corenlp(output, date, STORYID):
+    event_dict = {STORYID: {}}
+    event_dict[STORYID]['sents'] = {}
+    event_dict[STORYID]['meta'] = {}
+    event_dict[STORYID]['meta']['date'] = date
+    for i, sent in enumerate(output['sentences']):
+        sents = output['sentences']
+        event_dict[STORYID]['sents'][i] = {}
+        event_dict[STORYID]['sents'][i]['content'] = ' '.join(sents[i]['tokens'])
+        event_dict[STORYID]['sents'][i]['parsed'] = sents[i]['parse'].upper().replace(')', ' )')
+
+    return event_dict
+
+api.add_resource(ExtractAPI, '/siesta/extract')
+
+if __name__ == '__main__':
+    http_server = HTTPServer(WSGIContainer(app))
+    http_server.listen(5002)
+    IOLoop.instance().start()

docker-compose.yml

@@ -0,0 +1,11 @@
+siesta:
+    build: .
+    links:
+        - ccnlp
+        - petrarch
+    ports:
+        - "5002:5002"
+ccnlp:
+    image: caerusassociates/ccnlp:1.0.0
+petrarch:
+    build: petrarch/.

petrarch/Dockerfile

@@ -0,0 +1,15 @@
+FROM ubuntu:14.04
+
+MAINTAINER John Beieler <[email protected]>
+
+RUN apt-get update && apt-get install -y git python-dev python-pip 
+
+RUN pip install git+https://github.com/openeventdata/petrarch.git 
+
+ADD . /src
+
+RUN cd /src; pip install -r requirements.txt
+
+EXPOSE 5001
+
+CMD ["python", "/src/petrarch_app.py"]

petrarch/petrarch_app.py

@@ -0,0 +1,53 @@
+import os
+from petrarch import petrarch
+from tornado.ioloop import IOLoop
+from tornado.wsgi import WSGIContainer
+from tornado.httpserver import HTTPServer
+from flask import Flask, jsonify, make_response
+from flask.ext.restful import Api, Resource, reqparse
+
+app = Flask(__name__)
+api = Api(app)
+
+cwd = os.path.abspath(os.path.dirname(__file__))
+
+
+@app.errorhandler(400)
+def bad_request(error):
+    return make_response(jsonify({'error': 'Bad request'}), 400)
+
+
+@app.errorhandler(404)
+def not_found(error):
+    return make_response(jsonify({'error': 'Not found'}), 404)
+
+
+
+class CodeAPI(Resource):
+    def __init__(self):
+        self.reqparse = reqparse.RequestParser()
+        self.reqparse.add_argument('events', type=dict)
+        super(CodeAPI, self).__init__()
+
+    def post(self):
+        args = self.reqparse.parse_args()
+        event_dict = args['events']
+
+        print(event_dict)
+        event_dict_updated = petrarch.do_coding(event_dict, None)
+
+        return event_dict_updated
+
+
+api.add_resource(CodeAPI, '/petrarch/code')
+
+if __name__ == '__main__':
+    config = petrarch.utilities._get_data('data/config/', 'PETR_config.ini')
+    print("reading config")
+    petrarch.PETRreader.parse_Config(config)
+    print("reading dicts")
+    petrarch.read_dictionaries()
+
+    http_server = HTTPServer(WSGIContainer(app))
+    http_server.listen(5001)
+    IOLoop.instance().start()

petrarch/requirements.txt

@@ -0,0 +1,11 @@
+pytest==2.6.3
+requests==2.4.3
+Flask==0.10.1
+Flask-RESTful==0.3.3
+Flask-HTTPAuth==2.5.0
+itsdangerous==0.24
+Jinja2==2.7.3
+MarkupSafe==0.23
+tornado==4.2
+simplejson==3.6.5
+Werkzeug==0.10.4

requirements.txt

@@ -0,0 +1,11 @@
+pytest==2.6.3
+requests==2.4.3
+Flask==0.10.1
+Flask-RESTful==0.3.3
+Flask-HTTPAuth==2.5.0
+itsdangerous==0.24
+Jinja2==2.7.3
+MarkupSafe==0.23
+tornado==4.2
+simplejson==3.6.5
+Werkzeug==0.10.4