plint

French poetry validator
git clone https://a3nm.net/git/plint/
Log | Files | Refs | README

commit e3450cac6e652683a4917913e298d7d8178147e9
parent 39ba6402b2d72e508a4c33f43e8889c933b6858d
Author: Antoine Amarilli <a3nm@a3nm.net>
Date:   Thu, 17 Aug 2017 01:10:45 +0200

elide inside words also for unaspirated h

Diffstat:
TODO | 3---
verse.py | 15+++++++++++++++
versetest.py | 7+++++++
3 files changed, 22 insertions(+), 3 deletions(-)

diff --git a/TODO b/TODO @@ -1,6 +1,3 @@ -should be accepted (elision should also occur within compound words): -"Le navire, doublant le cap de Sainte-Hélène" - "pays" and "paysage": "payse", "abbaye" should be rejected (hemistiche is broken): diff --git a/verse.py b/verse.py @@ -177,6 +177,21 @@ class Verse: for i, w in enumerate(word[:-1]): if w['text'] == "e-": w['weights'] = [0] # force elision + if (w['text'] == "e" and i < len(word) - 1 and + word[i+1]['text'].startswith("-h")): + # collect what follows until the next hyphen or end + flw = word[i+1]['original'].split('-')[1] + ii = i+2 + while ii < len(word): + flw += word[ii]['original'].split('-')[0] + if '-' in word[ii]['original']: + break + ii += 1 + # TODO: not sure if this reconstruction of the original word is bulletproof... + if haspirater.lookup(normalize(flw)): + w['weights'] = [0] + else: + w['weights'] = [1] # remove leading and trailing crap for w in self.chunks: diff --git a/versetest.py b/versetest.py @@ -151,6 +151,13 @@ class ExceptionCounts(Counts): f = self.runCount("porte-avions porte-avions", limit=6) self.assertEqual(1, len(f)) self.assertEqual(self.getWeight(f[0]), 6) + + def testSainteHelene(self): + # from "Les Trophées", José-Maria de Heredia + text = "Le navire, doublant le cap de Sainte-Hélène," + f = self.runCount(text, limit=12) + self.assertEqual(1, len(f)) + self.assertEqual(self.getWeight(f[0]), 12) class AspiratedCounts(Counts): def testBaudelaire1half(self):