Skip to content

Latest commit

 

History

History
33 lines (23 loc) · 1.53 KB

rules.md

File metadata and controls

33 lines (23 loc) · 1.53 KB

Working areas: How to define them

  • During the data ingestion phase xltoy start collecting cells from known working areas.

  • Working areas are named ranges set in Name Manager, they are of 3 types:

    • data: defines the area that delimits the data sources, it enable some features like diff positional or relative
    • model: define the model area, it must be of one row or one column.
    • names: sed to give a mnemonic name to each equation in the model area. It must be coupled to a model range of the same shape, usually in the same sheet. Type is distinguished by a regular expression defined in settings file. By default string with suffix data_ model_ names_ case insensitive identify .
  • One sheet - One model.
    Only one type for working area can be set for each sheet. The main idea here is to promote some kind of discipline and order models one for sheet.

  • Model must be a range of one column or row, optionally aligned with names range. Aligned means that ranges must start from same position and must have the same shape. With this technique, each names is coupled to his equation. If names range is not given equation will be named as sequence of anon_* names.

Parser: implemented syntax

  • basic operators: + - * /
  • basic functions: sqrt len sum min max ave len log
  • logic operators: and or
  • function: rand

Parser: unhandled syntax

Not all excel syntax in formulas are recognized by the parser , here some unhandled examples:

*+ [eg: 1*+len(A3)]