Vocabulary¶

Functions for use inside a claim, bound with let and called like any other: let nfkc = mathema_language.text.nfkc, for s in L[unicode], f(nfkc(s)) == f(s). Each carries a stable identifier (text.nfkc@1) that the record keeps, so a claim means the same thing when another tool reads it.

Text¶

In mathema_language.text.

Function	Returns
`nfc(s)`, `nfd(s)`	the canonical composed or decomposed form
`nfkc(s)`, `nfkd(s)`	the compatibility forms: ligatures, fullwidth letters, `℀` folded
`casefold(s)`, `lower(s)`, `upper(s)`	the full Unicode case mappings
`count(s, part)`	how many times `part` occurs
`startswith(s, prefix)`, `endswith(s, suffix)`	whether `s` begins or ends with it
`strip(s)`	leading and trailing whitespace removed
`splitlines(s)`	the lines, on every Unicode line boundary
`nfc_len(s)`	the length in NFC code points
`utf8_len(s)`	the length in UTF-8 bytes, what a byte-limited field counts

f = username_key
let nfkc = mathema_language.text.nfkc, for name in L[unicode], f(nfkc(name)) == f(name)
    falsified   name = '℀': 'a/c' vs '℀'

Trees¶

In mathema_language.tree, over any nested dict, list, tuple or record. On a tree of records they count records, as the refinements inside L[...] do; on plain dicts and lists they count containers and values.

Function	Returns
`depth(v)`	records along the deepest path, else containers; 0 for a scalar
`nodes(v)`	the records, else every value and container
`width(v)`	the most children any one node has
`leaves(v)`	the nodes with no children