Refactor splitting functions to more manageable functions

[Sujanadh]

Updates:

There have been many changes to the clean code base.

• Heavy functions such as split_by_sql and splitBySQL are broken down into smaller functions to make them manageable
• Improve performance of splitting
• Allow custom features without osm_id and tags during splitting using fmtm-splitter algorithm
• Extract linear features when custom features are provided
• If no extracts is provided, we retrieve both polygons and linestrings (buildings and road networks/rivers/...)

Outcome:

These changes have uplifted the performance of the fmtm-splitter, I tested using the whole Tokha municipality; it was successful without a timeout. Previously, it wasn't possible to split the whole Tokha municipality.

To Do:

• Optimize osm-rawdata to fetch and process data extracts
• Allow users to select obstacles to use them during splitting (create config yaml based on user input ) Configurable data extracts via YAML config, so user can define which linear feature type to split on #39
• Run splitting algorithm in a modular way and visualize each output of different steps (in the near future) Modularisation of FMTM Splitter #62, Halt at each step of splitting algorithm for intermediate product inspection #79
• Create separate splitter workflow to specify a number of enumerators #59