This is a description of the formal PEG morphology algorithm. If there are or appear to be discrepancies between this description and the formal grammar, the formal grammar wins.
Given any string of characters, this algorithm will split it uniquely into words, including possibly some non-lojban words.
We examine the string from the left, and words are extracted one by one from the left.
1- Extract all pauses, if there are any.
2- Extract a word and then any following pauses, if there are any.
3- Repeat 2 until there is nothing left.
----------
See also Morphology: Rules of formation
----------
Morphology: words
Morphology: cmene
Morphology: cmavo
Morphology: brivla
Morphology: fu'ivla
Morphology: extended rafsi
Morphology: gismu
Morphology: rafsi
Morphology: syllables
Morphology: vowels
Morphology: consonants
Morphology: special characters