lyricsgen

gen.py

import requests
from random import choice
from bs4 import BeautifulSoup
import json
import peewee

db = peewee.SqliteDatabase("lyricsgen.sqlite3")
parser = "html.parser"


class Artist(peewee.Model):
    name = peewee.CharField(unique=True, index=True)
    num_songs = peewee.IntegerField()
    text = peewee.TextField()
    model = peewee.TextField(null=True)

    class Meta:
        database = db


def getLyrics(artist_url):
    try:
        artist = Artist.get(Artist.name == artist_name)
    except Artist.DoesNotExist:
        print("> Fetching songs…")
        url = f"http://www.karaoketexty.cz/texty-pisni/{artist_name}"
        artist_page = BeautifulSoup(requests.get(url).content, parser)
        songs = artist_page.select("td.left a")
        trainingText = ""
        songsProcessed = 0
        for song in songs:
            url = f"http://www.karaoketexty.cz/{song['href']}"
            song_page = BeautifulSoup(requests.get(url).content, parser)
            song_lyrics = song_page.select("p.text")[0].text
            trainingText += song_lyrics + "\n"
            songsProcessed += 1
        Artist.create(name=artist_name,
                      num_songs=songsProcessed,
                      text=trainingText,
                      model=None)
        print("> Songs saved to DB.")
    else:
        print("> Using songs from DB.")
        trainingText = artist.text
    return(trainingText)


def getModel(artist_name, order):
    def generate_model():
        print("> Model not found, generating one…")
        text = getLyrics(artist_name)
        model = {}
        for i in range(0, len(text) - order):
            fragment = text[i:i + order]  # Range is exclusive at upper bound
            nextLetter = text[i + order]  # So this is the next letter
            if fragment not in model:
                model[fragment] = {}
            if nextLetter not in model[fragment]:
                model[fragment][nextLetter] = 1
            else:
                model[fragment][nextLetter] += 1
        artist = Artist.get(Artist.name == artist_name)
        artist.model = json.dumps(model)
        artist.save()
        print("> Model saved to DB.")
        return(model)

    try:
        artist = Artist.get(Artist.name == artist_name)
    except Artist.DoesNotExist:
        model = generate_model()
    else:
        if artist.model is None:
            model = generate_model()
        else:
            print("> Using model from DB.")
            model = json.loads(artist.model)
    return(model)


def getNextCharacter(model, fragment):
    letters = []
    for letter in model[fragment].keys():
        for occurences in range(0, model[fragment][letter]):
            # So random.choice has a greater weighted chance of
            # selecting this one
            letters.append(letter)
    return(choice(letters))


def generateLyrics(artist_name, order, length):
    model = getModel(artist_name, order)
    text = getLyrics(artist_name)
    currentFragment = text[0:order]
    output = ""
    print("> Generating lyrics…\n\n")
    for i in range(0, length - order):
        newCharacter = getNextCharacter(model, currentFragment)
        output += newCharacter
        currentFragment = currentFragment[1:] + newCharacter
    return(output)


db.connect()

try:
    db.create_tables([Artist])
except peewee.OperationalError:
    pass

artist_name = input("Enter artist:\n")

newLyrics = generateLyrics(artist_name, 8, 600)
newLyrics = newLyrics[newLyrics.find("\n") + 1:newLyrics.rfind("\n")]
print(newLyrics)

kompletne.cz.py

import requests
from bs4 import BeautifulSoup
import sys
parser = "html.parser"


def eprint(*args, **kwargs):
    print(*args, file=sys.stderr, **kwargs)


def getLyrics(artist_name):
    url = f"http://texty.kompletne.cz/{artist_name}"
    artist_page = BeautifulSoup(requests.get(url).content, parser)
    songs = artist_page.select(".album_songy a")
    trainingText = ""
    songsProcessed = 0
    for song in songs:
        eprint(f"processing {song['href']}")
        url = f"http://texty.kompletne.cz/{song['href']}"
        song_page = BeautifulSoup(requests.get(url).content, parser)
        song_lyrics = song_page.select("#text_center")[0].text
        trainingText += song_lyrics + "\n"
        songsProcessed += 1
    eprint(songsProcessed)
    return trainingText


artist_name = str(sys.argv[1])
print(getLyrics(artist_name))

requirements.txt

beautifulsoup4
flake8
requests
peewee