From 5b9076eeb50183979fdcab1d4d87ce98e284d590 Mon Sep 17 00:00:00 2001
From: julius <julius@0124816.xyz>
Date: Wed, 12 Apr 2023 06:47:38 +0000
Subject: [PATCH] New implementation with SQLModel

---
 mw_scraper.py | 134 ++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 134 insertions(+)
 create mode 100644 mw_scraper.py

diff --git a/mw_scraper.py b/mw_scraper.py
new file mode 100644
index 00000000..11b45a7a
--- /dev/null
+++ b/mw_scraper.py
@@ -0,0 +1,134 @@
+import random
+import re
+from pathlib import Path
+from time import sleep
+from urllib.parse import urlparse
+
+from requests_html import HTMLSession
+from sqlmodel import Field, Relationship, Session, SQLModel, create_engine, select
+
+with open("db.secrets", "r") as f:
+    db_pass = f.readline().strip()
+
+
+class Word(SQLModel, table=True):
+    id: int | None = Field(default=None, primary_key=True)
+    word: str = Field(index=True, unique=True)
+
+
+class Sense(SQLModel, table=True):
+    id: int | None = Field(default=None, primary_key=True)
+    word: str = Field(default=None, foreign_key="word.word")
+
+    word_class: str | None
+
+
+class Description(SQLModel, table=True):
+    id: int | None = Field(default=None, primary_key=True)
+    word: str | None = Field(default=None, foreign_key="word.word")
+    sense_id: int | None = Field(default=None, foreign_key="sense.id")
+    description: str
+    examples: list[str] | None
+
+
+class Pronunciation(SQLModel, table=True):
+    id: int | None = Field(default=None, primary_key=True)
+    word: str | None = Field(default=None, foreign_key="word.word")
+    sense: int | None = Field(default=None, foreign_key="sense.id")
+    pronunciation: str
+
+
+engine = create_engine(db_pass)
+SQLModel.metadata.create_all(engine)
+
+html_session = HTMLSession()
+
+QUEUE = {line.strip() for line in open("queue.db", "rt")}
+
+
+def add_word(word):
+    url = f"https://www.merriam-webster.com/dictionary/{word}"
+    r = html_session.get(url)
+    # r.html.render()
+    links = set()
+
+    with Session(engine) as session:
+        for c in r.html.find("div.entry-word-section-container"):
+            _word = c.find("h1.hword,p.hword", first=True).text
+            try:
+                _class = c.find("h2.parts-of-speech", first=True).text
+            except AttributeError:
+                print(f"fail: {_word}")
+                continue
+            _class = re.sub("\(.*\)", "", _class).strip()
+            links |= {
+                Path(urlparse(link.attrs["href"]).path).name
+                for link in c.find('a[href*="/dictionary/"]')
+            }
+
+            results = session.exec(select(Word).where(Word.word == _word)).one_or_none()
+            if results:
+                word = results
+            else:
+                word = Word(word=_word)
+                session.add(word)
+
+            results = session.exec(
+                select(Sense)
+                .where(Sense.word == _word)
+                .where(Sense.word_class == _class)
+            ).one_or_none()
+            if results:
+                sense = results
+            else:
+                sense = Sense(word=word.word, word_class=_class)
+                session.add(sense)
+
+            session.commit()
+
+            for sn in c.find("div.sense"):
+                _desc = []
+                _examples = []
+                for dt in sn.find("span.dt"):
+                    for dttext in dt.find("span.dtText"):
+                        _desc.append(dttext.text)
+                for dt in sn.find("div.sub-content-thread"):
+                    _example = []
+                    for sents in dt.find("span.sents"):
+                        _example.append(sents.text)
+                    _examples.append("; ".join(_example))
+                session.add(
+                    Description(
+                        word=word.word,
+                        sense_id=sense.id,
+                        description="; ".join(_desc),
+                        examples=_examples,
+                    )
+                )
+            for pron in c.find(
+                "span.prons-entries-list-inline div.prons-entry-list-item,a.prons-entry-list-item"
+            ):
+                session.add(
+                    Pronunciation(
+                        word=word.word, sense=sense.id, pronunciation=pron.text
+                    )
+                )
+            session.commit()
+    return links
+
+
+def present():
+    with Session(engine) as session:
+        return session.exec(select(Word.word)).unique()
+
+
+while True:
+    try:
+        QUEUE |= add_word(random.choice(list(QUEUE)))
+        QUEUE -= set(present())
+        print(len(QUEUE))
+        sleep(random.random() * 5)
+    except KeyboardInterrupt:
+        with open("queue.db", "wt") as f:
+            f.write("\n".join(list(QUEUE)))
+        exit(0)