用 2captcha 破解驗證碼，輕鬆抓取台股券商分點買賣日報

pip install 2captcha-python selenium pandas beautifulsoup4

import time, base64
from selenium import webdriver
from selenium.webdriver.common.by import By

browser = webdriver.Chrome()
browser.get('https://bsr.twse.com.tw/bshtm/')
time.sleep(1)

# 切換到主查詢 frame
browser.switch_to.frame('page1')

# 取得驗證碼圖片元素
img_element = browser.find_elements(By.TAG_NAME, 'img')[1]

# 把圖片轉為 base64
img_base64 = browser.execute_script("""
    var img = arguments[0];
    var canvas = document.createElement('canvas');
    var ctx = canvas.getContext('2d');
    canvas.width = img.width;
    canvas.height = img.height;
    ctx.drawImage(img, 0, 0);
    return canvas.toDataURL('image/jpeg').substring(22);
""", img_element)

# 存成 code.png
with open('code.png', 'wb') as f:
    f.write(base64.b64decode(img_base64))
print("已存出驗證碼圖片 code.png")

import requests
import time

api_key = '你的API_KEY'  # 請填入自己的key
file = {'file': open('code.png', 'rb')}
data = {'key': api_key, 'method': 'post'}

# 上傳圖片取得ID
resp = requests.post('http://2captcha.com/in.php', files=file, data=data)
if 'OK' in resp.text:
    captcha_id = resp.text.split('|')[1]
    # 等待辨識
    for i in range(10):
        time.sleep(3)
        check = requests.get(
            f'http://2captcha.com/res.php?key={api_key}&action=get&id={captcha_id}'
        )
        if 'OK' in check.text:
            code = check.text.split('|')[1].upper()
            print('驗證碼:', code)
            break
        elif 'CAPCHA_NOT_READY' in check.text:
            continue
        else:
            print('驗證失敗:', check.text)
else:
    print('上傳失敗:', resp.text)

from selenium.webdriver.common.by import By
import pandas as pd
from bs4 import BeautifulSoup
import datetime

StockNo = '2330'  # 查詢股票 2330
browser.find_element(By.NAME, 'TextBox_Stkno').send_keys(StockNo)
browser.find_element(By.NAME, 'CaptchaControl1').send_keys(code)
browser.find_element(By.NAME, 'btnOK').click()
time.sleep(2)

# 轉換 frame 取得表格
browser.switch_to.default_content()
browser.switch_to.frame('page2')

soup = BeautifulSoup(browser.page_source, "html.parser")
tbody_list = [
    t for t in soup.find('body').find_all('tbody')
    if '交易日期' not in str(t)
]

# 整理成 DataFrame
rows = []
for tbody in tbody_list:
    tds = tbody.find_all('td')
    if len(tds) == 5:
        rows.append([td.text.strip() for td in tds])
df = pd.DataFrame(rows, columns=['序', '證券商', '成交單價', '買進股數', '賣出股數'])
df = df[df['序'].str.isnumeric()]
df['成交單價'] = df['成交單價'].astype(float)
df['買進股數'] = df['買進股數'].astype(float)
df['賣出股數'] = df['賣出股數'].astype(float)

# 儲存結果
df.to_excel(f"{StockNo}_{datetime.datetime.now():%Y-%m-%d}.xlsx", index=False)
print("已儲存查詢結果。")

import re

# 清理券商欄位，僅保留代碼
def extract_code(s):
    m = re.search(r'[A-Za-z0-9]+', str(s))
    return m.group(0) if m else ''
df['券商代號'] = df['證券商'].apply(extract_code)

# 載入券商名冊
broker_url = 'https://www.twse.com.tw/rwd/zh/brokerService/outPutExcel'
broker_df = pd.read_excel(broker_url)
broker_df = broker_df[['證券商代號', '證券商名稱']].rename(
    columns={'證券商代號': '券商代號', '證券商名稱': '券商名稱'}
)
broker_df['券商代號'] = broker_df['券商代號'].astype(str).str.strip()

# 合併券商名稱
df_merged = pd.merge(df, broker_df, on='券商代號', how='left')

# 計算每家券商買進賣出總和
grouped = df_merged.groupby(['券商代號', '券商名稱'], as_index=False)[['買進股數', '賣出股數']].sum()
grouped['買賣超'] = grouped['買進股數'] - grouped['賣出股數']
grouped = grouped.sort_values('買賣超', ascending=False)

# 取前15名、後15名
top_15_buy = grouped.head(15)
top_15_sell = grouped.tail(15).sort_values('買賣超')

print("主力買超前15名：")
print(top_15_buy[['券商名稱', '券商代號', '買進股數', '賣出股數', '買賣超']])
print("\n主力賣超前15名：")
print(top_15_sell[['券商名稱', '券商代號', '買進股數', '賣出股數', '買賣超']])