adrien/tchap-beta
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

plusieurs sources, plusieurs phrases

Browse Source
This commit is contained in:
Adrien MALINGREY
2026-02-05 16:02:30 +01:00
parent b4182cb71f
commit ccba64ed35
1 changed files with 11 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
markov.py
View File

@@ -2,17 +2,19 @@ from collections import defaultdict
from random import choice, randrange from random import choice, randrange
suivants = defaultdict(list) suivants = defaultdict(list)
for path in (
with open("fra_wikipedia_2021_10K-sentences.txt", "r", encoding="utf-8") as fichier: "fra_wikipedia_2021_10K-sentences.txt",
):
with open(path, "r", encoding="utf-8") as fichier:
for phrase in fichier: for phrase in fichier:
antepenultieme, penultieme = "", "" antepenultieme, penultieme = "", ""
for word in phrase.split(): for word in phrase.split():
suivants[(antepenultieme, penultieme)].append(word) suivants[(antepenultieme, penultieme)].append(word)
antepenultieme, penultieme = penultieme, word antepenultieme, penultieme = penultieme, word
def parle(): def parle(nb_phrases=1):
phrases = [] phrases = []
for _ in range(randrange(1, 4)): for _ in range(nb_phrases):
antepenultieme, penultieme = "", "" antepenultieme, penultieme = "", ""
phrase = [] phrase = []
while mots_possibles := suivants[(antepenultieme, penultieme)]: while mots_possibles := suivants[(antepenultieme, penultieme)]: