metro_service_extract's Introduction

Metro Data Extractor

This is a quick hack job to convert metro service reports into some form of structured data.

The entire output is in JSON, which I've dumped in the test_data/ directory.

Some issues:

Sometimes two messages get combined by accident.
Delays and Gaps are just whatever number happens to be in the message along with the word ('minute' and ('gap' or 'delay'))
I didn't check the timestamp too heavily either.
Python: My module/class/file/import structure is terrible. Can you help me fix it?

Future Stuff:

More accurate color/line reporting
What station did it happen at?
Depending on the station, which lines would it effect?
What direction was the train traveling?

Sample Data

[{
    "proc_text": "4:38 p.m. a shady grove-bound red line train at cleveland park was offloaded due to a door problem. passengers experienced a 10-minute delay.",
    "expressed": false,
    "id": "3292_16_38",
    "gap": null,
    "delay": {
      "minutes": 10
    },
    "colors": [
      "red"
    ],
    "event_dtg": "2015-08-06T16:38:00"
  },
{
    "proc_text": "3:37 p.m. a greenbelt-bound green line train at branch avenue did not operate, resulting in a 6-minute gap in service.",
    "expressed": false,
    "id": "3292_15_37",
    "gap": {
      "minutes": 6
    },
    "delay": null,
    "colors": [
      "green"
    ],
    "event_dtg": "2015-08-06T15:37:00"
  },
{
    "proc_text": "8:10 a.m. a branch avenue-bound green line train at suitland was expressed for schedule adherence/improved train spacing.",
    "expressed": true,
    "id": "3292_8_10",
    "gap": null,
    "delay": null,
    "colors": [
      "green"
    ],
    "event_dtg": "2015-08-06T08:10:00"
  }]