A well working technical strategy can get completely screwed on fundamental news releases, like the NFP report yesterday which caused a straight 200 Pips ralley. How would I implement a fundamental news filter which can be backtested as well? In this case I would have allowed only trades in the direction of the USD bearish sentiment.

I could provide a text file which contains all release dates, corresponding currencies and allowed directions

....
2016-05-02 23:00 NZD neutral
2016-05-03 8:00 EUR bearish
2016-05-04 14:30 USD

So x hours before 2016-05-04 14:30 all trades should be closed and no new trades should be opened. When the release number is known (some milliseconds after 14:30) an external program will add bearish at the end of the line and the script will know that until y hours after this release, USD may only be sold.

How can I call the simulated date and time during backtesting? In the manual I just found a timer, which I think is useless in this case.

Having fundamental indicators support in Zorro would be a tremendous advantage. One could program powerful news trading strategies like News Straddling with order/trade canceling if the release number is in line with the projected analysts number, fading the news (when data which is normaly not that important like the german Ifo business climate index) doesn't justify the spike or wait for a correction after a spike and then trade in the direction of the spike, riding the second wave.

Read the file into an array of structs containing time and text. At every bar, identify the array element close to the current bar time, and act according to its text.

Thanks for the hint jcl, I found what I need in the manual now.

how/where did you get such a file? It would be interesting to use it in a backtest for many things

Nanitek, you can download economic news release data from myfxbook in CSV format via the economic calendar on that site.