1. Tigra Astronomy
  2. Open Source
  3. Horizon Data Interchange




Horizon Data Interchange is a command line utility for importing and exporting horizon data from various planetarium software and file formats.

This version supports AstroPlanner, ObserverPro, TheSkyX and ACP, we hope to augment this over time.

Community involvement is welcome and encouraged. Please feel free to fork the repository, work on it yourself and send us pull requests. We've tried to make the code easy to extend, but if you can see a better way, have at it!

This project is covered by the MIT License. This is a very permissive license which basically allows anyone to do anything at all with the software without obligation. Commercial use is expressly allowed. We would of course appreciate attribution and a link to this repository and/or the Tigra Astronomy web site.

What is this repository for?

  • Horizon data. Interchange. A quick and simple command line utility for moving horizon data between various astronomical software and file formats.
  • Current version is: 0.0 pre-release; API not frozen, expect breaking changes, anything goes.

Command Line Syntax

The command line requires that the user specifies exactly one importer (-i or --importer) and exactly one exporter (-e or --exporter). The remainder of the command line is specific to the importer and exporter chosen.

Supplying the --help option (or no options) will result in a help message being displayed.

10 Jan 2016 -- New feature, --UseLightDome will take the lightdome column of Astroplanner and create an ACP horizon in the ACP registry. Quit ACP before running the program. 24 Jul 2016 -- New feature, --UseLightDome will also work with ObserverPro 05 Aug 2016 -- New feature, --UseLightDome will also work with TheSkyX

Usage Example

The initial release has only one importer, which imports from an AstroPlanner horizon file in CSV format; and an ACP exporter, which writes horizon data directly to ACP's registry. The following two commands are equivalent.

horizon.exe --importer AstroPlanner --exporter Acp --sourcefile file.csv
horizon.exe -iAstroPlanner -eAcp -sfile.csv

Importers and Exporters may extend the command line with their own options.

How do I get set up?

  • You'll need Visual Studio 2010 or later, we recommend Visual Studio 2015. The Community Edition is free and now allows the use of plug-ins so we also highly recommend that you get yourself a ReSharper license.
  • The code targets .Net Framework 4.5 although we haven't used any fancy features and it should work fine with any version if you need to change it.
  • Dependencies are brought in automatically as part of the build process using NuGet.
  • If you are getting errors about "CodeContracts" missing, install (https://visualstudiogallery.msdn.microsoft.com/1ec7db13-3363-46c9-851f-1ce455f66970 "CodeContracts")
  • We use MSpec (Machine.Specifications) for our unit tests and FakeItEasy as our mocking framework. We are happy to accept any testing/mocking framework if you have a strong preference, as long as we can support it on our TeamCity build server and it doesn't exclude other developers (e.g. by having an expensive license fee).
  • This is a simple utility, we will use XCOPY deployment. No need for an installer, just copy the compiled files and run. Because of this, everything needed to run the utility must ship with it to maintain 'XCOPY deployment'. Please don't reference anything outside the project, or users will have a hard time ensuring they have the right dependencies to use the utility. It must be possible to simply XCopy the bin directory.
  • If you are new to Git, may we suggest Atlassian's SourceTree utility? It's free, works on Windows and Mac and can use both Git and Mercurial. It has both a graphical user interface and a command prompt if you prefer that, and it has its own bundled versions of Git and Mercurial, so it is absolutely everything you need.

Contribution guidelines

  • We try hard to work test-first and we encourage you to do the same. If you haven't done this before, it can take some getting used to. Why not use this project as an excuse to learn?
  • We're very informal, but we will review all pull requests before merging them. We used to be afraid of code reviews, but we realise now that good constructive criticism is a win for everyone, so please try to be open to suggestions. We would love it if other developers would review our code, but getting people to do code reviews seems to be as hard as pulling teeth! Anyone can participate in code reviews but please keep it constructive and forget about cosmetic details. We don't care if you don't like the layout of braces (we use Whitesmith's style, deal with it). We do care if our code is confusing, violates some guideline or best practice, or is overly complicated. We are interested in ways to make the code cleaner, more loosely coupled, more SOLID and more Agile.

Some Suggestions for Contributors

  • Please try to write Clean Code and stick to the SOLID principles of object oriented design.
  • We use GitFlow as our branching strategy, it has worked very well for us across a number of projects. The bare minimum you must know about GitFlow to get started is:
    • The master branch is reserved for releases. Commits to master are forbidden, except to merge a release/* branch which will normally be done by Tigra Astronomy.
    • Development happens on the develop branch, or preferably a feature branch off develop. Feature branches are conventionally named feature/*.
    • As far as possible, do one self-contained feature or bugfix per branch and then create a pull request to merge it back into develop, allowing time for code review.
    • Find out all the gory details at http://nvie.com/posts/a-successful-git-branching-model/
    • We are not a fan of rebasing. We prefer to push everything, so everyone can see what happened.
  • If you commit binaries or build artefacts to the source code repository, or create any folders containing the words 'copy', 'backup', 'old' or a version number, then we will hunt you down and give you the wedgie you deserve! We've seen people do all of the above and it reveals a singular lack of understanding about what a version control system does! Please think about what you commit and only commit source code, not build artefacts nor anything that is specific to you or your computer.

Who do I talk to?

Writing Your Own Importer or Exporter

Importers and exporters are discovered dynamically by searching for classes that implement IHorizonImporter or IHorizonExporter. Instances are created using the parameterless constructor based on command line arguments.

So for example, if the user specifies the following command:

<pre>Horizon.exe --Importer Wibbly --Exporter Wobbly</pre>

then an attempt will be made to load WibblyImporter and WobblyExporter. Your class name should therefore be meaningful to the potential user and must end in either Importer or Exporter and must have a parameterless constructor.

Importers must implement IHorizonImporter and exporters must implement IHorizonExporter. You can implement both in the same class, but two separate instances will be created at runtime so you'll need to consider the implications of that.

At run time, your ProcessCommandLineArguments() method will be called, passing in the raw command line argument list as an array of strings exactly as it was passed to Main() by the operating system and a parser object that you can use to parse your command line options. It is your responsibility to parse the command line and act on any options that are relevant to your class; you should ignore any options you don't recognize. You can do this any way you are comfortable with, but keep in mind that you get all of the command line options, not just the ones relevant to you. We suggest handling command line parsing like this:

Create a data transfer object (a class with nothing but properties) to hold the parsed results of your options. For example:

internal sealed class MyOptions
    [Option('s',"SourceFile", Required=true, HelpText = "The input file name.")]
    public string SourceFile { get; set; }

Next, use the command line parser to parse the arguments into your options:

    var options = parser.ParseArguments<MyOptions>(args);
    if (options.Errors.Any())
        Environment.ExitCode=-1;    // Set a -ve exit code to have help printed out on the console automatically
        throw new ArgumentException();  // Throw an appropriate exception.

Once everyone has had a chance to handle their command line options, your ImportHorizon() or ExportHorizon() method will be called.

  • Importers are expected to construct and return a HorizonData object populated with the imported data. Only add data that you've actually got, don't try to 'fill in the gaps'.
  • Exporters will be handed a HorizonData object and can use it to create exported data at whatever resolution is needed, up to a maximum of 1 degree steps (many apps only need measurements every 5 or 10 degrees).

The HorizonData object

The HorizonData object is implemented as a specialized derivative of Dictionary<int, HorizonDatum>, where the integer key is the azimuth of the reading and the value is an instance of HorizonDatum, which contains altitude measurements for both the solid horizon and the light dome.

Where HorizonData differs from an ordinary dictionary is that it is possible to read from keys that don't exist, and the values returned in the HorizonDatum will be interpolated values. Note that the altitude values are double, because they may have been interpolated. You may truncate or round these values as your needs dictate.

This automatic interpolation makes the import/export process rather straightforward.

  1. To import the horizon, add each imported datum to the appropriate azimuth in your HorizonData collection. If your imported data has measurements every 10 degrees, then you will create entries for 0, 10, 20, 30 and so on up to 340, 350. You only need to add as many items as you obtain from the imported data set, there is no need to 'fill in the gaps'
  2. To export, simply read off whatever azimuths you need to create your exported data. If your application expects a reading every 2 degrees, then simply read off values at 2 degree increments: 0, 2, 4, 6, ... 356, 358. There is no need to check whether each value exists - you will automatically get interpolated values if there were gaps in the data.

Important: Due to the fact that different programs and file formats use different azimuth resolutions and some applications use integer values for altitude, data is not necessarily round-trippable. Whenever data is moved from a higher resolution source to a lower resolution source, information loss occurs; when data is moved from a lower resolution source to a higher resolution source, interpolation occurs which introduces quantisation noise into the data. Therefore, it is best to pick one data source as being 'authoritative' and then only export from that source.

ObserverPro from http://www.observer.pro (Joshua Bury), allows an iOS device to create a Horizon file from the device camera. The file can be exported via email and has the text format "angle,deg.rees" Angle ranges from 0 to 359 by 1, deg.rees are precise to 2 decimal places. This is non-round trippable data.

You will need to move the new horizon file for TheSkyX into the expected location for your version of the file. The file needs to be named 'Custom Horizon.hrz'

Copyright © 2015 Tigra Astronomy, all rights reserved.