This is a branch that I abandoned over a year ago, going to revisit as a draft PR.

Will follow up where discussion left off to close #305

TODO:

• Consider more efficient implementation using a lookup table (e.g. texcl_to_ssc() option to derive low/high values from input texture class #305 (comment))
• Consider more efficient implementation with vectorized functions rather than iteration (applying to list of "texture groups" would be a special case that may require some form of iteration)
• Add support for ranges in sand fractions (e.g. "very fine sandy loam" vs. "sandy loam")
• Add support for informal subclasses of soil texture (e.g. "light silty clay loam" vs. "heavy silty clay loam"); this would be using some additional argument as a modifier on the base textural class, or parsing "light" and "heavy" from a class label
  • a simple assumption would be to divide the clay range for the whole class in half to determine range for light and heavy (e.g. sicl clay content range is from 27 to 40%, so light sicl would be from 27 to 33.5, with midpoint of 30.25; and so forth)