OrgModeParser parses Emacs OrgMode files and builds a C++ object tree that represents the Org files structure. This functionality is made available in a library that can be integrated into other applications. The motivation behind it was to be able to integrate existing applications written in C++ with OrgMode. For example, a calendar application could write an OrgMode file to make TODO items available in the OrgMode agenda. An email program could offer functionality to add TODO items from the user's inbox. Et cetera. The possibilities are endless.

OrgModeParser requires Qt 5.3 and CMake 3.2. Newer versions should work as well. On most Linux distributions, shipped packages should be fine. Installation is straight-forward:

> git clone https://github.com/mirkoboehm/OrgModeParser.git
> mkdir OrgModeParser/build
> cd OrgModeParser/build
> cmake ../
> make
> make install

This install the OrgModeParser dynamic library, as well as the tools it provides. The default installation prefix is /usr/local.

An OrgMode file is parsed by passing an open text stream to the parse method:

> OrgElement::Pointer orgfile = parser.parse(&stream, inputFile);

orgfile now holds a hierarchical data structure of the content of the file. The concept is similar to DOM processing of HTML files. OrgMode file sections are called headlines and start with an asterisk. Filtering can be used to create flat lists of headlines that match certain criteria. All headlines can be collected using the element type Headline like this:

> auto const headlines = findElements<Headline>(orgfile);

Headline is a subclass of OrgElement. Other element types are for example clock lines or drawers. By adding a predicate, additional criteria can be implemented:

> auto const todos = findElements<Headline>(orgfile, isTODO);

See the TODOCounter demo for more a compilable version of these examples. The Writer class can be used to write out OrgMode files.

As of now, OrgModeParser installs OrgModeParser_ClockTimeSummary, a tool that create a one-line summary of the OrgMode clock time entries created this week. Its output looks something like this:

02:10: TODO Write README for OrgModeParser   00:00/26:59

The left duration shows the time spend on the currently active task, followed by the item headline. On the right, it displays the totals of the time clocked today, and this week.

All functionality of the parser is contained in the OrgModeParser library. The content of an OrgMode file is represented as a tree of objects that represent the various elements of the file, like headlines, drawers, properties, tags and clocklines. Other classes like Clock or Property provide access to the values of the tree elements. A simple demo that counts the number of TODO items in an OrgMode file can be found in Demos/TODOCounter:

    OrgElement::Pointer orgfile = parser.parse(&stream, inputFile);
    auto const headlines = findElements<Headline>(orgfile);
    wcout << "Number of headlines: " << headlines.count() << endl;
    auto isTODO = [](const Headline::Pointer& element) {
        return element->caption().startsWith(QStringLiteral("TODO"));
    };
    auto const todos = findElements<Headline>(orgfile, isTODO);
    wcout << "Number of TODOs: " << todos.count() << endl;

OrgModeParser is developed in a test-driven development model. To run the tests, go to the build directory and run make test.

To fix or extend the functionality of OrgModeParser, provide examples of Org files and specify where OrgModeParser parses it wrong or does not implement features. When submitting pull requests on Github. remember every new piece of code needs to be auto-tested.