putting together existing accentuation code to explain accentuation of each word in GNT

At the moment it's focused on describing the rules around proclitics and enclitics. analyse_morphgnt.py will output something like the following (for the entire SBLGNT, which I've committed as accent-analysis.txt):

011037 #::----:1A::##--::---#--:-----:--::-:--# οὐκ οὐ οὐ
011037 #::--#-:--::--##::----#-:-----:E#::-:--# ἔστιν ἐστί(ν) εἰμί
011037 -::----:E#::--#-::------:-----:--::-:--- μου μου ἐγώ

What this is basically saying (in compact form) is:

• οὐκ is not the same as its normalised form οὐ
• οὐκ is preceded by an oxytone
• οὐκ is a proclitic
• οὐκ is an exception to the way ἐστίν is accented
• οὐκ has a movable (kappa)
• οὐκ is followed by an enclitic
• ἔστιν is not the same as its normalised form ἐστί(ν)
• ἔστιν is preceded by a proclitic
• ἔστιν is preceded by an unaccented word
• ἔστιν is an enclitic
• ἔστιν is a dissyllabic enclitic
• ἔστιν is a variant of ἐστί
• ἔστιν is preceded by a word that causes it to be accented ἔστιν
• ἔστιν is followed by an enclitic
• μου is preceded by a the specially-accented ἔστι
• μου is enclitic

It makes it easy to both match certain conditions like "find all the enclitics with a preceding accented proclitic which are themselves also accented because what follows is another enclitic" and also verify that a text has been accented correctly according to the rules (or verify the rules adequately explain the accentuation of the text).

Lots more coming! I soon plan to incorporated parts of greek-accentuation to annotate why verbs are accented the way they are and eventually nominals.