Decompose Emacs Org documents into hierarchical objects for NLP processing
Example MyDocument.org:
* my heading
** my subheading
** my other subheading
*** a smaller subheading
Translates to an object with the properties:
Document:
-- title: MyDocument.org
-- nodes:
--> Node 1
--- text: "my heading"
--- parents: MyDocument.org
--- children: Node 2, Node 3
--> Node 2
--- text: "my subheading"
--- parents: Node 1
--- children: None
--> Node 3
--- text: "my other subheading"
--- parents: Node 1
--- children: Node 4
--> Node 4
--- text: "a smaller subheading"
--- parents: Node 3
--- children: Node 4
Python example is easiest to run, just run python src/main/python/knowledgegraph.py /path/to/org/files/.
The script will parse the org files and return a prompt going through random lines from each file as a visual example.
Customization could be done by using the Collection object inside of if __name__=='__main__' to iterate through documents/nodes using Collection.documents: List[Document] and Collection.documents[*].nodes: List[Node].
All other languages were kinda just an exercise for myself but the associated build files are here:
Scala: sbt clean package
Java: mvn clean package
C++: make clean && make release