Wq-io

API Improvements
- Cache index in `BaseIO` to speed up lookups when used as a `dict`
- Use an array instead of a generator in `CsvParser`

Breaking Changes
- Moved domain-specific `contrib` modules (`cocorahs` and `hydromet`) to the [climata](https://github.com/heigeo/climata) library

API Improvements
- Full read-write support for GIS data via the [Fiona](https://github.com/Toblerity/Fiona) and [Shapely](https://github.com/Toblerity/Shapely) libraries. (1)
- NetLoaders now send a custom user-agent string.
- Improved `guess_type` which is now utilized by [wq.db.contrib.files](http://wq.io/docs/files).
- Switch to built-in Python XML library to speed up pip install (4)

New Classes & Modules
- **[TimeSeriesMapper](http://wq.io/docs/mappers)**: Automatically maps text values containing dates or numbers to appropriate Python types. TimeSeriesMapper also includes a notion of "key fields" (e.g. date, time, location) as opposed to "parameter fields", or the actual data being recorded in the series.
- **[make_date_mapper(fmt)](http://wq.io/docs/mappers)**: Returns a function that parses dates with the provided [format string](http://docs.python.org/2/library/datetime.htmlstrftime-strptime-behavior).
- **[SkipPreludeReader](http://wq.io/docs/parsers)**: Subclass of [csv.DictReader](http://docs.python.org/2/library/csv.htmlcsv.DictReader) that allows ingestion of CSV files with non-CSV header preludes.
- **[contrib.hydromet](https://github.com/wq/wq.io/blob/master/contrib/hydromet.py)**: IOs for retrieving Hydromet and Agrimet data from http://www.usbr.gov

API improvements
- **[contrib.cocorahs](https://github.com/wq/wq.io/blob/master/contrib/cocorahs.py)**: `CocorahsIO` Now utilizes new `TimeSeriesMapper`

You may also be interested in [climata](https://github.com/heigeo/climata), a new wq.io-based library for loading Applied Climate Information System (ACIS) data.

API Improvements
- Better handling of Excel files in which data doesn't necessarily start on the second row.
- New `make_io()` function which can be used to mix the provided `loader`, `parser`, and optionally `mapper` classes together into a usable `IO` class. The default mapper is `TupleMapper`.
- New `load_file()` function will attempt to automatically determine which loaders and parsers to use and create an `IO` class instance on-the-fly.

Initial version