Remove old train.parquet file and update new file with larger size; enhance convert.py to drop duplicates and reset index; add PTT article crawler script

crawler/ptt/main.py

@@ -0,0 +1,125 @@
+from bs4 import BeautifulSoup
+import requests
+import re
+import logging
+from typing import Dict, List, Tuple
+
+logging.basicConfig(
+    level=logging.INFO,
+    format='%(asctime)s - %(levelname)s - %(message)s'
+)
+logger = logging.getLogger(__name__)
+
+def parse_article(html_content: str) -> Tuple[str, List[Dict]]:
+    soup = BeautifulSoup(html_content, "html.parser")
+    main_content_div = soup.find("div", id="main-content")
+    if not main_content_div:
+        return "", []
+    main_content_copy = main_content_div.__copy__()
+    for tag in main_content_copy.find_all(["div", "span"], class_=["article-metaline", "article-metaline-right", "f2", "push"]):
+        tag.decompose()
+    # 移除網址
+    text = main_content_copy.get_text(strip=True, separator="\n")
+    text = re.sub(r'https?://\S+', '', text)
+    pushes = soup.find_all("div", class_="push")
+    push_list = []
+    for push in pushes:
+        try:
+            tag = push.find("span", class_="push-tag").get_text(strip=True)
+            user_id = push.find("span", class_="push-userid").get_text(strip=True)
+            content = push.find("span", class_="push-content").get_text(strip=True).lstrip(": ")
+            time = push.find("span", class_="push-ipdatetime").get_text(strip=True)
+            push_list.append({
+                "tag": tag,
+                "user_id": user_id,
+                "content": content,
+                "time": time
+            })
+        except AttributeError as e:
+            logger.warning(f"解析推文時發生錯誤: {e}")
+            continue
+    return text, push_list
+
+def main():
+    """從網址讀取多篇文章，並顯示內文"""
+    import re
+    import time
+    import urllib
+    from multiprocessing import Pool
+    from requests_html import HTMLSession
+
+    def parse_article_entries(elements):
+        results = []
+        for element in elements:
+            try:
+                push = element.find('.nrec', first=True).text
+                mark = element.find('.mark', first=True).text
+                title = element.find('.title', first=True).text
+                author = element.find('.meta > .author', first=True).text
+                date = element.find('.meta > .date', first=True).text
+                link = element.find('.title > a', first=True).attrs['href']
+
+                # 使用正規表示式過濾合法 PTT 網址
+                if not re.match(r"^/bbs/\w+/M\.\d+\.A\.\w+\.html$", link):
+                    continue
+
+            except AttributeError:
+                if '(本文已被刪除)' in title:
+                    match_author = re.search('\[(\w*)\]', title)
+                    if match_author:
+                        author = match_author.group(1)
+                elif re.search('已被\w*刪除', title):
+                    match_author = re.search('\<(\w*)\>', title)
+                    if match_author:
+                        author = match_author.group(1)
+            results.append({
+                'push': push,
+                'mark': mark,
+                'title': title,
+                'author': author,
+                'date': date,
+                'link': link
+            })
+        return results
+
+    def parse_next_link(controls):
+        link = controls[1].attrs['href']
+        return urllib.parse.urljoin('https://www.ptt.cc/', link)
+
+    def get_posts(post_links):
+        with Pool(processes=8) as pool:
+            responses = pool.map(session.get, post_links)
+            return responses
+
+    session = HTMLSession()
+    session.cookies.set('over18', '1')
+    url = 'https://www.ptt.cc/bbs/LoL/index.html'
+
+    num_page = 10
+    post_links = []
+    for _ in range(num_page):
+        response = session.get(url)
+        metadata = parse_article_entries(elements=response.html.find('div.r-ent'))
+        next_page_url = parse_next_link(controls=response.html.find('.action-bar a.btn.wide'))
+        post_links += [urllib.parse.urljoin(url, meta['link']) for meta in metadata]
+        url = next_page_url
+
+    start_time = time.time()
+    results = get_posts(post_links)
+    print(f'共 {len(results)} 篇文章，花費: {time.time() - start_time:.2f} 秒\n')
+
+    txt_file = "ptt.txt"
+    for i, res in enumerate(results):
+        main_text, _ = parse_article(res.text)
+        # print(f'--- 第 {i+1} 篇文章 ---')
+        # print(main_text)
+        # print('\n')
+        # write to text file
+        with open(txt_file, "a", encoding="utf-8") as f:
+            f.write(main_text + "\n\n")
+    print(f"文章已寫入 {txt_file}")
+
+        
+
+if __name__ == "__main__":
+    main()

data/train.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ec0d1e7c7b5bcf622a7b6248ec81f31a33abe5c4bcb8b0646524500dffd7697c
-size 101412
+oid sha256:61b17507086708ca1fce53157fdd6fa9bc36049750b1e776adf6cdc134bab144
+size 636201

prepare/convert.py

@@ -49,7 +49,9 @@ def main():
     
     # Create DataFrame
     df = pd.DataFrame(all_data, columns=['text', 'origin'])
-    
+    df.drop_duplicates(subset=['text'], inplace=True)
+    df.reset_index(drop=True, inplace=True)
+    print(f"Total unique lines: {len(df)}")
     # Save to CSV
     # csv_path = "train.csv"
     # df.to_csv(csv_path, index=False)

prepare/threads.txt

@@ -548,4 +548,5 @@ Academia Sinica 中央研究院 電波生物教師 LLM實習工程師
 不願成為麻木的人，所以盡力去共情。我堅持的正義是我不變成施暴者，也不變成烏合之眾聽他人讒言。
 我要保持我的判斷，我的野心是我的思想，和他人毫無關係。
 朋友擔心我到泰國不安全 讓我每天狂發日常給她看 順帶也發一下這裡
-朋友說有一種我屬於富士山的美。
+朋友說有一種我屬於富士山的美。
+記得下次實習生報到要先問用不用機械鍵盤，用的話就給警告不能拿來公司用

prepare/train.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f72440b5e35b6ba5178ac61b8d742759843cb226a25377fd93338a91357dabbe
-size 102537