New DelimitedTextTimeSeriesFormatter

[wkramer]

We should make a configurable DelimitedTextTimeSeriesFormatter, which should handle reading comma separated values (.csv) and fixed column width files (e.g. .noos). We can use this in combination with the TimeSeriesFormatterDataObject and TimeSeriesFormatterStochObserver. The configuration could be something like this:

<?xml version="1.0" encoding="UTF-8"?>
<timeSeriesFormatterConfig xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
  <formatter class="org.openda.exchange.timeseries.DelimitedTextTimeSeriesFormatter">
    <dateTimePattern>yyyyMMddHHmm</dateTimePattern>
    <timeZone>GMT</timeZone>
    <delimiter>\s+</delimiter>
  </formatter>
  <timeSeries id="waterlevel@denhelder" status="use" standardDeviation="0.05">
    den_helder_waterlevel_astro.noos
  </timeSeries>
  <timeSeries id="waterlevel@aberdeen">
    aberdeen_waterlevel_astro.noos
 </timeSeries>
 <timeSeries id="waterlevel@moon" status="ignore">
    does_not_exist
 </timeSeries>
</timeSeriesFormatterConfig>

TODO:

• [x] Support delimiter

• [x] Support dateTimePattern & timeZone

• [x] Support skipLines for skipping the first n lines

• [x] Support comment for skipping lines starting with the specified string.

• [x] Support dateTimeSelector, valueSelector and standardDeviationSelector

• [x] implement write method

• [ ] preserve whitespace when writting

• [ ] preserve comments and header when writing

• [ ] create xsd for configuration

[wkramer]

We could also try to parse TIM (D-Flow FM) files and add a unit test