Code about rhymes
Some useful resources:
Basis for syllabizing words is the CMU pronouncing Dictionary https://en.wikipedia.org/wiki/CMU_Pronouncing_Dictionary "Official version" (cited by cmu) is the svn repo: http://svn.code.sf.net/p/cmusphinx/code/trunk/cmudict/ Also at github: https://github.com/cmusphinx/cmudict (this might be better maintained, but the format has changed a bit, so the code below would need to be tweaked) Can also import this data via nltk, but installing nltk is annoying and maybe not super helpful. Note: github and nltk versions use lowercase entries and uppercase syllables, svn uses uppercase for both. Arbitrarily normalizing to lowercase here.
CMU dict uses an enriched form of ARPAbet: see https://en.wikipedia.org/wiki/ARPABET
CMU uses the 2-letter variant, with stress indicators. Stress is indicated by
digits appended to vowels, and is a little confusing: 1 is primary stress,
2 is secondary stress, and 0 is no stress. (so 1 > 2 > 0).
humanize_stress converts to a a sensible format
Readings: Zwicky, Linguistics and Folk Poetry
To do:
- html output generation fails due to titles, etc, preceding poem