From eb9bada544b87d96ad81e927d8f02c140c217bca Mon Sep 17 00:00:00 2001
From: DarkCat09 <gh@dc09.ru>
Date: Sat, 27 Apr 2024 22:08:20 +0400
Subject: [PATCH] Genius lyrics parser (it works!)

---
 genius.py      | 56 ++++++++++++++++++++++++++++++++++++++++++++++++++
 test_genius.py | 42 +++++++++++++++++++++++++++++++++++++
 2 files changed, 98 insertions(+)
 create mode 100644 genius.py
 create mode 100644 test_genius.py

diff --git a/genius.py b/genius.py
new file mode 100644
index 0000000..a839d6a
--- /dev/null
+++ b/genius.py
@@ -0,0 +1,56 @@
+import re
+
+from lxml import html, etree  # type: ignore
+import urllib3
+
+
+# Matches not only latin symbols,
+# but also cyrillic alphabet, Chinese hyeroglyphics...
+word_regex = re.compile(r'\w+')
+
+# Get all <div>s with lyrics
+lyrics_xpath = etree.XPath('//div[@id="lyrics-root"][1]/div[@data-lyrics-container="true"]')
+
+# Get <br>s to add newlines after them for correct text parsing
+br_xpath = etree.XPath('.//br')
+
+
+def search(http: urllib3.PoolManager, title: str, artist: str) -> str:
+    '''Searches for Genius lyrics using SearXNG + Yahoo and returns the first URL.
+    Irrelevant texts should be picked manually'''
+
+    resp = http.request(
+        'GET',
+        'https://searx.dc09.ru/search',
+        fields={
+            'q': artist + ' ' + title + ' site:genius.com',
+            'engines': 'yahoo',
+            'safesearch': '0',
+            'format': 'json',
+        },
+    )
+
+    result: dict[str, str] = resp.json()['results'][0]
+    del resp
+
+    return result['url']
+
+
+def parse(http: urllib3.PoolManager, url: str) -> str:
+    '''Requests a lyrics page and parses with libxml2,
+    leaving only text and line breaks'''
+
+    resp = http.request('GET', url)
+    tree = html.document_fromstring(resp.data)
+    divs = lyrics_xpath(tree)
+    del resp, tree
+
+    return '\n'.join(_parse_text(div) for div in divs).strip()
+
+
+def _parse_text(elem: html.HtmlElement) -> str:
+
+    for br in br_xpath(elem):
+        br.tail = '\n' + (br.tail or '')
+
+    return elem.text_content()
diff --git a/test_genius.py b/test_genius.py
new file mode 100644
index 0000000..34ee5d2
--- /dev/null
+++ b/test_genius.py
@@ -0,0 +1,42 @@
+from unittest import TestCase
+import urllib3
+
+import genius
+
+
+TITLE = 'A Line In The Sand'
+ARTIST = 'Linkin Park'
+URL = 'https://genius.com/Linkin-park-a-line-in-the-sand-lyrics'
+
+LYR1 = '''[Intro: Mike Shinoda]
+Today, we stood on the wall'''
+
+LYR2 = '''little did we know
+
+[Instrumental Break]
+
+[Verse 1:'''
+
+LYR3 = '''you are gonna get yours
+
+[Chorus: Chester Bennington]
+Another day'''
+
+
+class TestGenius(TestCase):
+
+    def setUp(self) -> None:
+        self.http = urllib3.PoolManager()
+
+    def test_search_success(self) -> None:
+        url = genius.search(self.http, TITLE, ARTIST)
+        self.assertEqual(url, URL)
+
+    def test_lyrics_parsing(self) -> None:
+        lyrics = genius.parse(self.http, URL)
+        self.assertTrue(lyrics.startswith(LYR1))
+        self.assertTrue(LYR2 in lyrics)
+        self.assertTrue(LYR3 in lyrics)
+
+    def tearDown(self) -> None:
+        self.http.clear()