Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -11,7 +11,7 @@ import gradio as gr
|
|
| 11 |
os.system("playwright install chromium")
|
| 12 |
|
| 13 |
# ==========================================
|
| 14 |
-
# 🛠️ 共用工具:翻譯與學名藥濾網
|
| 15 |
# ==========================================
|
| 16 |
def translate_lang(text, target_lang):
|
| 17 |
try:
|
|
@@ -22,14 +22,20 @@ def translate_lang(text, target_lang):
|
|
| 22 |
return text
|
| 23 |
|
| 24 |
def is_generic(brand_name, company_name, ingredient):
|
| 25 |
-
"""通用學名藥濾網:排除成分名開頭
|
| 26 |
b_lower = brand_name.lower()
|
| 27 |
c_lower = company_name.lower()
|
| 28 |
i_lower = ingredient.lower()
|
| 29 |
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
|
| 34 |
if b_lower.startswith(i_lower) or i_lower in b_lower: return True
|
| 35 |
if any(gk in b_lower or gk in c_lower for gk in generic_keywords): return True
|
|
@@ -37,7 +43,6 @@ def is_generic(brand_name, company_name, ingredient):
|
|
| 37 |
return False
|
| 38 |
|
| 39 |
def clean_brand_name(raw_name):
|
| 40 |
-
"""切除劑量、劑型等後綴,保留純商品名"""
|
| 41 |
return re.split(r'(皮下注|錠|カプセル|顆粒|シロップ|OD|細粒|液|\d+)', raw_name)[0].replace('®', '').strip()
|
| 42 |
|
| 43 |
# ==========================================
|
|
@@ -47,66 +52,113 @@ def get_usa_originator(ingredient):
|
|
| 47 |
log, brands, companies = [], set(), set()
|
| 48 |
try:
|
| 49 |
session = curl_req.Session(impersonate="chrome120")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
payload = {"drugname": ingredient, "discontinued": "RX,OTC,DISCN", "submit": "Search"}
|
| 51 |
res = session.post("https://www.accessdata.fda.gov/scripts/cder/ob/search_product.cfm", data=payload, timeout=30, verify=False)
|
|
|
|
|
|
|
| 52 |
soup = BeautifulSoup(res.text, 'html.parser')
|
| 53 |
table = soup.find('table', id='example')
|
| 54 |
if table:
|
|
|
|
|
|
|
| 55 |
headers = [th.get_text(strip=True).lower() for th in table.find_all('th')]
|
| 56 |
brand_idx = next((i for i, h in enumerate(headers) if 'proprietary name' in h), 2)
|
| 57 |
rld_idx = next((i for i, h in enumerate(headers) if 'rld' in h), 8)
|
| 58 |
mfg_idx = next((i for i, h in enumerate(headers) if 'applicant holder' in h), 10)
|
| 59 |
-
|
|
|
|
| 60 |
tds = tr.find_all('td')
|
| 61 |
if len(tds) > max(rld_idx, brand_idx) and "RLD" in tds[rld_idx].get_text(strip=True).upper():
|
| 62 |
brands.add(tds[brand_idx].get_text(strip=True))
|
| 63 |
if len(tds) > mfg_idx: companies.add(tds[mfg_idx].get_text(strip=True))
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
|
|
|
|
|
|
|
|
|
| 67 |
|
| 68 |
def get_belgium_originator(ingredient):
|
|
|
|
| 69 |
log, brands = [], set()
|
| 70 |
try:
|
| 71 |
session = curl_req.Session(impersonate="chrome120")
|
|
|
|
| 72 |
calc_id, ds = "5b55a713-dc8e-46bd-a82a-cea4641fb7b1", "c9852b22-1d58-4471-b8cc-a821763e821e"
|
| 73 |
url = f"https://www.test-achats.be/CalculatorsAPI/Medicine/GetSearchResults/{calc_id}?datasource={ds}&term={quote(ingredient)}"
|
| 74 |
res = session.get(url, headers={"X-Requested-With": "XMLHttpRequest"}, timeout=30, verify=False)
|
|
|
|
|
|
|
|
|
|
| 75 |
if res.status_code == 200:
|
| 76 |
-
|
|
|
|
|
|
|
| 77 |
label = item.get('label', '')
|
| 78 |
-
if
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
|
| 83 |
def get_france_originator(ingredient):
|
| 84 |
-
brands = set()
|
| 85 |
try:
|
| 86 |
session = curl_req.Session(impersonate="chrome120")
|
|
|
|
| 87 |
res = session.get(f"https://base-donnees-publique.medicaments.gouv.fr/medicament/recherche/resultat?contains={quote(ingredient)}", timeout=30, verify=False)
|
|
|
|
| 88 |
soup = BeautifulSoup(res.text, 'html.parser')
|
| 89 |
|
| 90 |
-
|
| 91 |
-
|
|
|
|
| 92 |
info_div = div.find('div', class_='infos')
|
| 93 |
if info_div and info_div.find('a'):
|
| 94 |
title = info_div.find('a').get_text(strip=True)
|
| 95 |
if not is_generic(title, "", ingredient) and '- gé' not in title.lower():
|
| 96 |
brands.add(clean_brand_name(title))
|
| 97 |
|
| 98 |
-
if brands: return ", ".join(brands), "法國
|
| 99 |
-
return "查無原廠", "-", "
|
| 100 |
-
except Exception as e: return "執行失敗", "-", str(e)
|
| 101 |
|
| 102 |
def get_australia_originator(ingredient):
|
| 103 |
-
brands = set()
|
| 104 |
try:
|
| 105 |
session = curl_req.Session(impersonate="chrome120")
|
|
|
|
| 106 |
res = session.get(f"https://www.tga.gov.au/resources/artg?keywords={ingredient}", timeout=30, verify=False)
|
|
|
|
| 107 |
soup = BeautifulSoup(res.text, 'html.parser')
|
|
|
|
|
|
|
|
|
|
| 108 |
cands = []
|
| 109 |
-
for article in
|
| 110 |
title_tag, time_tag = article.find('h3'), article.find('time')
|
| 111 |
if title_tag and time_tag:
|
| 112 |
full_t = title_tag.get_text(strip=True)
|
|
@@ -117,131 +169,176 @@ def get_australia_originator(ingredient):
|
|
| 117 |
cands.append({"brand": brand, "date": time_tag.get('datetime')})
|
| 118 |
if cands:
|
| 119 |
cands = sorted(cands, key=lambda x: x['date'])
|
| 120 |
-
|
| 121 |
-
|
| 122 |
-
|
|
|
|
| 123 |
|
| 124 |
def get_sweden_originator(ingredient):
|
| 125 |
-
brands, companies = set(), set()
|
| 126 |
try:
|
| 127 |
session = curl_req.Session(impersonate="chrome120")
|
|
|
|
| 128 |
res = session.get(f"https://www.fass.se/LIF/search?query={quote(ingredient)}", timeout=30, verify=False)
|
|
|
|
| 129 |
soup = BeautifulSoup(res.text, 'html.parser')
|
| 130 |
-
|
|
|
|
|
|
|
|
|
|
| 131 |
title_tag, comp_tag = item.find('a', class_='product-name'), item.find('span', class_='company-name')
|
| 132 |
if title_tag and comp_tag:
|
| 133 |
title, comp = title_tag.get_text(strip=True), comp_tag.get_text(strip=True)
|
| 134 |
if not is_generic(title, comp, ingredient):
|
| 135 |
brands.add(clean_brand_name(title))
|
| 136 |
companies.add(comp)
|
| 137 |
-
if brands: return ", ".join(brands), ", ".join(companies), "
|
| 138 |
-
return "查無原廠", "-", "
|
| 139 |
-
except Exception as e: return "執行失敗", "-", str(e)
|
| 140 |
|
| 141 |
# ==========================================
|
| 142 |
# 🚀 模組 B:使用 Playwright 抓取 (英、加、日、德、瑞)
|
| 143 |
# ==========================================
|
| 144 |
def get_uk_originator(ingredient, page):
|
| 145 |
-
brands, companies = set(), set()
|
| 146 |
try:
|
|
|
|
| 147 |
page.goto(f"https://www.medicines.org.uk/emc/search?q={ingredient}", timeout=30000)
|
| 148 |
page.wait_for_selector('.search-results-product-info-title-link', timeout=15000)
|
| 149 |
soup = BeautifulSoup(page.content(), 'html.parser')
|
| 150 |
-
|
|
|
|
|
|
|
| 151 |
title = link.get_text(strip=True)
|
| 152 |
if not title.lower().startswith(ingredient.lower()):
|
| 153 |
brands.add(clean_brand_name(title))
|
| 154 |
p_div = link.find_parent(class_='search-results-product-info')
|
| 155 |
if p_div and p_div.find(class_='search-results-product-info-company'):
|
| 156 |
companies.add(p_div.find(class_='search-results-product-info-company').get_text(strip=True))
|
| 157 |
-
if brands: return ", ".join(brands), ", ".join(companies), "
|
| 158 |
-
return "查無原廠", "-", "
|
| 159 |
-
except Exception as e: return "執行失敗", "-", str(e)
|
| 160 |
|
| 161 |
def get_canada_originator(ingredient, page):
|
|
|
|
| 162 |
try:
|
|
|
|
| 163 |
page.goto("https://health-products.canada.ca/dpd-bdpp/index-eng.jsp", timeout=45000, wait_until="domcontentloaded")
|
| 164 |
page.locator('input[id="activeIngredient"]').fill(ingredient)
|
| 165 |
page.keyboard.press("Enter")
|
| 166 |
-
page.wait_for_selector('table#results', timeout=15000)
|
| 167 |
|
| 168 |
-
|
|
|
|
|
|
|
|
|
|
| 169 |
table = soup.find('table', id='results')
|
| 170 |
-
if not table or not table.find('tbody'): return "查無資料", "-", "❌ 表格未生成"
|
| 171 |
|
| 172 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 173 |
tds = tr.find_all('td')
|
| 174 |
if len(tds) >= 4:
|
| 175 |
comp = tds[2].get_text(strip=True)
|
| 176 |
if is_generic(tds[3].get_text(strip=True), comp, ingredient): continue
|
| 177 |
m = re.search(r'\d+', tds[1].get_text(strip=True))
|
| 178 |
if m and tds[1].find('a'):
|
| 179 |
-
all_cands.append({"company": comp, "product": tds[3].get_text(strip=True), "din": int(m.group())
|
| 180 |
|
| 181 |
-
if not all_cands: return "查無原廠", "-", "❌ 剩餘皆為學名藥"
|
| 182 |
|
| 183 |
all_cands = sorted(all_cands, key=lambda x: x['din'])
|
| 184 |
orig_comp = all_cands[0]['company']
|
| 185 |
brands = set([c['product'] for c in all_cands if c['company'] == orig_comp])
|
| 186 |
-
|
| 187 |
-
|
|
|
|
| 188 |
|
| 189 |
def get_japan_originator(ing_ja, page):
|
| 190 |
-
brands, companies = set(), set()
|
| 191 |
try:
|
| 192 |
-
|
| 193 |
page.goto("https://www.pmda.go.jp/PmdaSearch/iyakuSearch/", timeout=30000, wait_until="domcontentloaded")
|
| 194 |
-
page.wait_for_selector('input#txtName', timeout=15000)
|
| 195 |
-
page.fill('input#txtName', ing_ja)
|
| 196 |
-
page.press('input#txtName', 'Enter')
|
| 197 |
|
| 198 |
-
|
| 199 |
-
|
| 200 |
-
|
| 201 |
-
|
| 202 |
-
|
| 203 |
-
|
| 204 |
-
|
| 205 |
-
|
| 206 |
-
|
| 207 |
-
|
| 208 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 209 |
|
| 210 |
def get_switzerland_originator(ing_de, page):
|
| 211 |
-
brands, companies = set(), set()
|
| 212 |
try:
|
|
|
|
| 213 |
page.goto(f"https://compendium.ch/search?q={ing_de}", timeout=30000)
|
| 214 |
-
|
|
|
|
| 215 |
page.wait_for_load_state('networkidle', timeout=20000)
|
| 216 |
|
| 217 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 218 |
h3, strong = card.find('h3'), card.find('strong', class_='info')
|
| 219 |
if h3 and strong:
|
| 220 |
title, comp = h3.get_text(strip=True), strong.get_text(strip=True)
|
| 221 |
if not is_generic(title, comp, ing_de):
|
| 222 |
brands.add(clean_brand_name(title))
|
| 223 |
companies.add(comp)
|
| 224 |
-
if brands: return ", ".join(brands), ", ".join(companies), "
|
| 225 |
-
return "查無原廠", "-", "❌ 查無資料"
|
| 226 |
-
except Exception as e: return "執行失敗", "-", str(e)
|
| 227 |
|
| 228 |
def get_germany_originator(ing_de, page):
|
| 229 |
-
brands, companies = set(), set()
|
| 230 |
try:
|
|
|
|
| 231 |
page.goto(f"https://www.gelbe-liste.de/profi-suche/results?substance={quote(ing_de)}", timeout=30000)
|
| 232 |
page.wait_for_selector('.product-list', timeout=15000)
|
| 233 |
-
|
|
|
|
|
|
|
| 234 |
if ul:
|
| 235 |
-
|
|
|
|
|
|
|
| 236 |
h5, p_tag = li.find('h5'), li.find('p', class_='small')
|
| 237 |
if h5:
|
| 238 |
title, comp = h5.get_text(strip=True), (p_tag.get_text(strip=True) if p_tag else "-")
|
| 239 |
if not is_generic(title, comp, ing_de):
|
| 240 |
brands.add(clean_brand_name(title))
|
| 241 |
companies.add(comp)
|
| 242 |
-
if brands: return ", ".join(brands), ", ".join(companies), "
|
| 243 |
-
return "查無原廠", "-", "❌ 查無資料"
|
| 244 |
-
except Exception as e: return "執行失敗", "-", str(e)
|
| 245 |
|
| 246 |
# ==========================================
|
| 247 |
# 🚀 主執行中樞
|
|
@@ -253,7 +350,7 @@ def run_all_ten_countries(ing_en, ing_ja_manual, ing_de_manual):
|
|
| 253 |
ing_de = ing_de_manual if ing_de_manual else translate_lang(ing_en, 'de')
|
| 254 |
results = []
|
| 255 |
|
| 256 |
-
# 1. API 模組 (
|
| 257 |
usa_b, usa_c, usa_log = get_usa_originator(ing_en)
|
| 258 |
results.append(["🇺🇸 美國 (FDA)", usa_b, usa_c, usa_log])
|
| 259 |
|
|
@@ -294,7 +391,7 @@ def run_all_ten_countries(ing_en, ing_ja_manual, ing_de_manual):
|
|
| 294 |
# 🎨 UI 介面
|
| 295 |
# ==========================================
|
| 296 |
with gr.Blocks(title="十國原廠商品名智能檢索器") as demo:
|
| 297 |
-
gr.Markdown("## 🌐 跨國原廠商品名檢索器 (
|
| 298 |
|
| 299 |
with gr.Row():
|
| 300 |
ing_en = gr.Textbox(label="🧪 英文成分名 (必填)", placeholder="例如: Semaglutide")
|
|
@@ -306,7 +403,7 @@ with gr.Blocks(title="十國原廠商品名智能檢索器") as demo:
|
|
| 306 |
search_btn = gr.Button("🚀 啟動十國查詢", variant="primary")
|
| 307 |
|
| 308 |
result_table = gr.Dataframe(
|
| 309 |
-
headers=["國家", "🌟 判定為原廠的商品名", "🏭 藥廠名稱", "🛠️ 系統狀態"],
|
| 310 |
datatype=["str", "str", "str", "str"],
|
| 311 |
wrap=True, interactive=False
|
| 312 |
)
|
|
|
|
| 11 |
os.system("playwright install chromium")
|
| 12 |
|
| 13 |
# ==========================================
|
| 14 |
+
# 🛠️ 共用工具:翻譯與進階學名藥濾網
|
| 15 |
# ==========================================
|
| 16 |
def translate_lang(text, target_lang):
|
| 17 |
try:
|
|
|
|
| 22 |
return text
|
| 23 |
|
| 24 |
def is_generic(brand_name, company_name, ingredient):
|
| 25 |
+
"""通用學名藥濾網:排除成分名開頭、常見學名藥廠、以及歐洲平行輸入商"""
|
| 26 |
b_lower = brand_name.lower()
|
| 27 |
c_lower = company_name.lower()
|
| 28 |
i_lower = ingredient.lower()
|
| 29 |
|
| 30 |
+
# 新增德國平行輸入商 (Reimporte) 與法國/歐洲常見學名藥廠
|
| 31 |
+
generic_keywords = [
|
| 32 |
+
'sandoz', 'teva', 'apotex', 'ratiopharm', 'jamp', 'mint', 'pharmascience', 'sanis', 'sivem',
|
| 33 |
+
'auro', 'glenmark', 'taro', 'marcan', 'nora', 'mantra', 'reddy', 'mepha', 'axapharm',
|
| 34 |
+
'helvepharm', 'zentiva', 'spirig', 'aliud', 'puren', 'stada', 'eg ', '- gé', 'biogaran',
|
| 35 |
+
'arrow', 'viatris', 'zydus', 'kohlpharma', 'eurim', 'abacus', 'emra', 'cc pharma',
|
| 36 |
+
'orifarm', 'bb farma', 'fd pharma', 'mpa pharma', '1 4 u', '2care4', 'axicorp',
|
| 37 |
+
'nattermann', '1 0 1 carefarm', 'haemato'
|
| 38 |
+
]
|
| 39 |
|
| 40 |
if b_lower.startswith(i_lower) or i_lower in b_lower: return True
|
| 41 |
if any(gk in b_lower or gk in c_lower for gk in generic_keywords): return True
|
|
|
|
| 43 |
return False
|
| 44 |
|
| 45 |
def clean_brand_name(raw_name):
|
|
|
|
| 46 |
return re.split(r'(皮下注|錠|カプセル|顆粒|シロップ|OD|細粒|液|\d+)', raw_name)[0].replace('®', '').strip()
|
| 47 |
|
| 48 |
# ==========================================
|
|
|
|
| 52 |
log, brands, companies = [], set(), set()
|
| 53 |
try:
|
| 54 |
session = curl_req.Session(impersonate="chrome120")
|
| 55 |
+
log.append("1. 發送 GET 請求獲取 FDA Session Cookies...")
|
| 56 |
+
res_get = session.get("https://www.accessdata.fda.gov/scripts/cder/ob/index.cfm", timeout=30, verify=False)
|
| 57 |
+
log.append(f" -> 狀態碼: {res_get.status_code}, 內容長度: {len(res_get.text)}")
|
| 58 |
+
|
| 59 |
+
log.append("2. 發送 POST 請求提交搜尋...")
|
| 60 |
payload = {"drugname": ingredient, "discontinued": "RX,OTC,DISCN", "submit": "Search"}
|
| 61 |
res = session.post("https://www.accessdata.fda.gov/scripts/cder/ob/search_product.cfm", data=payload, timeout=30, verify=False)
|
| 62 |
+
log.append(f" -> 狀態碼: {res.status_code}, 內容長度: {len(res.text)}")
|
| 63 |
+
|
| 64 |
soup = BeautifulSoup(res.text, 'html.parser')
|
| 65 |
table = soup.find('table', id='example')
|
| 66 |
if table:
|
| 67 |
+
rows = table.find('tbody').find_all('tr') if table.find('tbody') else table.find_all('tr')
|
| 68 |
+
log.append(f"3. 找到結果表格,共 {len(rows)} 列資料。")
|
| 69 |
headers = [th.get_text(strip=True).lower() for th in table.find_all('th')]
|
| 70 |
brand_idx = next((i for i, h in enumerate(headers) if 'proprietary name' in h), 2)
|
| 71 |
rld_idx = next((i for i, h in enumerate(headers) if 'rld' in h), 8)
|
| 72 |
mfg_idx = next((i for i, h in enumerate(headers) if 'applicant holder' in h), 10)
|
| 73 |
+
|
| 74 |
+
for tr in rows:
|
| 75 |
tds = tr.find_all('td')
|
| 76 |
if len(tds) > max(rld_idx, brand_idx) and "RLD" in tds[rld_idx].get_text(strip=True).upper():
|
| 77 |
brands.add(tds[brand_idx].get_text(strip=True))
|
| 78 |
if len(tds) > mfg_idx: companies.add(tds[mfg_idx].get_text(strip=True))
|
| 79 |
+
else:
|
| 80 |
+
log.append("❌ HTML 中未發現 table#example,可能是無此藥或被阻擋。")
|
| 81 |
+
|
| 82 |
+
if brands: return ", ".join(brands), ", ".join(companies), "\n".join(log)
|
| 83 |
+
return "查無原廠", "-", "\n".join(log)
|
| 84 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 85 |
|
| 86 |
def get_belgium_originator(ingredient):
|
| 87 |
+
"""還原為您藥價工具中使用的 Test-Achats 後端 API 邏輯"""
|
| 88 |
log, brands = [], set()
|
| 89 |
try:
|
| 90 |
session = curl_req.Session(impersonate="chrome120")
|
| 91 |
+
log.append("1. 連線至 Test-Achats 後端 API...")
|
| 92 |
calc_id, ds = "5b55a713-dc8e-46bd-a82a-cea4641fb7b1", "c9852b22-1d58-4471-b8cc-a821763e821e"
|
| 93 |
url = f"https://www.test-achats.be/CalculatorsAPI/Medicine/GetSearchResults/{calc_id}?datasource={ds}&term={quote(ingredient)}"
|
| 94 |
res = session.get(url, headers={"X-Requested-With": "XMLHttpRequest"}, timeout=30, verify=False)
|
| 95 |
+
log.append(f" -> API 狀態碼: {res.status_code}")
|
| 96 |
+
|
| 97 |
+
exact_label = ingredient
|
| 98 |
if res.status_code == 200:
|
| 99 |
+
json_data = res.json()
|
| 100 |
+
log.append(f" -> API 回傳 {len(json_data)} 筆候選項目")
|
| 101 |
+
for item in json_data:
|
| 102 |
label = item.get('label', '')
|
| 103 |
+
if ingredient.lower() in label.lower():
|
| 104 |
+
exact_label = label
|
| 105 |
+
break
|
| 106 |
+
|
| 107 |
+
log.append(f"2. 使用標籤 '{exact_label}' 請求劑型資料 (/dosage)...")
|
| 108 |
+
post_headers = {"Content-Type": "application/x-www-form-urlencoded", "X-Requested-With": "XMLHttpRequest"}
|
| 109 |
+
res_dosage = session.post("https://www.test-achats.be/sante/maladies-et-medicaments/medicaments/calculateur/banque-de-donnees-medicaments/dosage", data=f"LabelValue={quote(exact_label)}", headers=post_headers, timeout=30, verify=False)
|
| 110 |
+
log.append(f" -> Dosage 狀態碼: {res_dosage.status_code}")
|
| 111 |
+
|
| 112 |
+
soup = BeautifulSoup(res_dosage.text, 'html.parser')
|
| 113 |
+
trs = soup.find_all('tr', attrs={'data-galenicform': True})
|
| 114 |
+
log.append(f"3. 找到 {len(trs)} 筆實際藥品資料。")
|
| 115 |
+
|
| 116 |
+
for tr in trs:
|
| 117 |
+
brand_tag = tr.find('a', class_='jq_detailPageLink') or tr.find('th', scope='row')
|
| 118 |
+
if brand_tag:
|
| 119 |
+
b_name = brand_tag.get_text(strip=True)
|
| 120 |
+
if not is_generic(b_name, "", ingredient):
|
| 121 |
+
brands.add(clean_brand_name(b_name))
|
| 122 |
+
|
| 123 |
+
if brands: return ", ".join(brands), "Test-Achats 無藥廠", "\n".join(log)
|
| 124 |
+
return "查無原廠", "-", "\n".join(log)
|
| 125 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 126 |
|
| 127 |
def get_france_originator(ingredient):
|
| 128 |
+
log, brands = [], set()
|
| 129 |
try:
|
| 130 |
session = curl_req.Session(impersonate="chrome120")
|
| 131 |
+
log.append("1. 發送 GET 至法國公共資料庫...")
|
| 132 |
res = session.get(f"https://base-donnees-publique.medicaments.gouv.fr/medicament/recherche/resultat?contains={quote(ingredient)}", timeout=30, verify=False)
|
| 133 |
+
log.append(f" -> 狀態碼: {res.status_code}, 內容長度: {len(res.text)}")
|
| 134 |
soup = BeautifulSoup(res.text, 'html.parser')
|
| 135 |
|
| 136 |
+
divs = soup.find_all('div', class_='result drug')
|
| 137 |
+
log.append(f"2. 找到 {len(divs)} 筆 'result drug' 區塊。")
|
| 138 |
+
for div in divs:
|
| 139 |
info_div = div.find('div', class_='infos')
|
| 140 |
if info_div and info_div.find('a'):
|
| 141 |
title = info_div.find('a').get_text(strip=True)
|
| 142 |
if not is_generic(title, "", ingredient) and '- gé' not in title.lower():
|
| 143 |
brands.add(clean_brand_name(title))
|
| 144 |
|
| 145 |
+
if brands: return ", ".join(brands), "法國系統無藥廠", "\n".join(log)
|
| 146 |
+
return "查無原廠", "-", "\n".join(log)
|
| 147 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 148 |
|
| 149 |
def get_australia_originator(ingredient):
|
| 150 |
+
log, brands = [], set()
|
| 151 |
try:
|
| 152 |
session = curl_req.Session(impersonate="chrome120")
|
| 153 |
+
log.append("1. 發送 GET 至澳洲 TGA...")
|
| 154 |
res = session.get(f"https://www.tga.gov.au/resources/artg?keywords={ingredient}", timeout=30, verify=False)
|
| 155 |
+
log.append(f" -> 狀態碼: {res.status_code}, 內容長度: {len(res.text)}")
|
| 156 |
soup = BeautifulSoup(res.text, 'html.parser')
|
| 157 |
+
|
| 158 |
+
articles = soup.find_all('article', class_='node--artg')
|
| 159 |
+
log.append(f"2. 找到 {len(articles)} 筆 ARTG 紀錄。")
|
| 160 |
cands = []
|
| 161 |
+
for article in articles:
|
| 162 |
title_tag, time_tag = article.find('h3'), article.find('time')
|
| 163 |
if title_tag and time_tag:
|
| 164 |
full_t = title_tag.get_text(strip=True)
|
|
|
|
| 169 |
cands.append({"brand": brand, "date": time_tag.get('datetime')})
|
| 170 |
if cands:
|
| 171 |
cands = sorted(cands, key=lambda x: x['date'])
|
| 172 |
+
log.append(f"✅ 依日期排序鎖定最早註冊: {cands[0]['date'][:10]}")
|
| 173 |
+
return cands[0]['brand'], "TGA資料庫", "\n".join(log)
|
| 174 |
+
return "查無原廠", "-", "\n".join(log)
|
| 175 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 176 |
|
| 177 |
def get_sweden_originator(ingredient):
|
| 178 |
+
log, brands, companies = [], set(), set()
|
| 179 |
try:
|
| 180 |
session = curl_req.Session(impersonate="chrome120")
|
| 181 |
+
log.append("1. 發送 GET 至瑞典 FASS...")
|
| 182 |
res = session.get(f"https://www.fass.se/LIF/search?query={quote(ingredient)}", timeout=30, verify=False)
|
| 183 |
+
log.append(f" -> 狀態碼: {res.status_code}, 內容長度: {len(res.text)}")
|
| 184 |
soup = BeautifulSoup(res.text, 'html.parser')
|
| 185 |
+
|
| 186 |
+
items = soup.find_all('li', class_=re.compile(r'search-result-item'))
|
| 187 |
+
log.append(f"2. 找到 {len(items)} 筆搜尋結果。")
|
| 188 |
+
for item in items:
|
| 189 |
title_tag, comp_tag = item.find('a', class_='product-name'), item.find('span', class_='company-name')
|
| 190 |
if title_tag and comp_tag:
|
| 191 |
title, comp = title_tag.get_text(strip=True), comp_tag.get_text(strip=True)
|
| 192 |
if not is_generic(title, comp, ingredient):
|
| 193 |
brands.add(clean_brand_name(title))
|
| 194 |
companies.add(comp)
|
| 195 |
+
if brands: return ", ".join(brands), ", ".join(companies), "\n".join(log)
|
| 196 |
+
return "查無原廠", "-", "\n".join(log)
|
| 197 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 198 |
|
| 199 |
# ==========================================
|
| 200 |
# 🚀 模組 B:使用 Playwright 抓取 (英、加、日、德、瑞)
|
| 201 |
# ==========================================
|
| 202 |
def get_uk_originator(ingredient, page):
|
| 203 |
+
log, brands, companies = [], set(), set()
|
| 204 |
try:
|
| 205 |
+
log.append("1. 前往 eMC...")
|
| 206 |
page.goto(f"https://www.medicines.org.uk/emc/search?q={ingredient}", timeout=30000)
|
| 207 |
page.wait_for_selector('.search-results-product-info-title-link', timeout=15000)
|
| 208 |
soup = BeautifulSoup(page.content(), 'html.parser')
|
| 209 |
+
links = soup.find_all('a', class_='search-results-product-info-title-link')
|
| 210 |
+
log.append(f"2. 找到 {len(links)} 筆標題,開始篩選。")
|
| 211 |
+
for link in links:
|
| 212 |
title = link.get_text(strip=True)
|
| 213 |
if not title.lower().startswith(ingredient.lower()):
|
| 214 |
brands.add(clean_brand_name(title))
|
| 215 |
p_div = link.find_parent(class_='search-results-product-info')
|
| 216 |
if p_div and p_div.find(class_='search-results-product-info-company'):
|
| 217 |
companies.add(p_div.find(class_='search-results-product-info-company').get_text(strip=True))
|
| 218 |
+
if brands: return ", ".join(brands), ", ".join(companies), "\n".join(log)
|
| 219 |
+
return "查無原廠", "-", "\n".join(log)
|
| 220 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 221 |
|
| 222 |
def get_canada_originator(ingredient, page):
|
| 223 |
+
log = []
|
| 224 |
try:
|
| 225 |
+
log.append("1. 前往 Canada DPD...")
|
| 226 |
page.goto("https://health-products.canada.ca/dpd-bdpp/index-eng.jsp", timeout=45000, wait_until="domcontentloaded")
|
| 227 |
page.locator('input[id="activeIngredient"]').fill(ingredient)
|
| 228 |
page.keyboard.press("Enter")
|
|
|
|
| 229 |
|
| 230 |
+
log.append("2. 等待結果表格或查無資料訊息...")
|
| 231 |
+
# 🟢 修正:同時等待 表格 或 錯誤提示,避免超時
|
| 232 |
+
page.wait_for_selector('table#results, .alert-info, .alert-warning', timeout=15000)
|
| 233 |
+
soup = BeautifulSoup(page.content(), 'html.parser')
|
| 234 |
table = soup.find('table', id='results')
|
|
|
|
| 235 |
|
| 236 |
+
if not table or not table.find('tbody'):
|
| 237 |
+
log.append("❌ 畫面上未出現表格,可能查無此藥。")
|
| 238 |
+
return "查無資料", "-", "\n".join(log)
|
| 239 |
+
|
| 240 |
+
rows = table.find('tbody').find_all('tr')
|
| 241 |
+
log.append(f"3. 找到 {len(rows)} 筆資料,進行黑名單過濾。")
|
| 242 |
+
all_cands = []
|
| 243 |
+
for tr in rows:
|
| 244 |
tds = tr.find_all('td')
|
| 245 |
if len(tds) >= 4:
|
| 246 |
comp = tds[2].get_text(strip=True)
|
| 247 |
if is_generic(tds[3].get_text(strip=True), comp, ingredient): continue
|
| 248 |
m = re.search(r'\d+', tds[1].get_text(strip=True))
|
| 249 |
if m and tds[1].find('a'):
|
| 250 |
+
all_cands.append({"company": comp, "product": tds[3].get_text(strip=True), "din": int(m.group())})
|
| 251 |
|
| 252 |
+
if not all_cands: return "查無原廠", "-", "\n".join(log) + "\n❌ 剩餘皆為學名藥廠"
|
| 253 |
|
| 254 |
all_cands = sorted(all_cands, key=lambda x: x['din'])
|
| 255 |
orig_comp = all_cands[0]['company']
|
| 256 |
brands = set([c['product'] for c in all_cands if c['company'] == orig_comp])
|
| 257 |
+
log.append(f"✅ 依 DIN 鎖定最古老藥廠: {orig_comp}")
|
| 258 |
+
return ", ".join(brands), orig_comp, "\n".join(log)
|
| 259 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 260 |
|
| 261 |
def get_japan_originator(ing_ja, page):
|
| 262 |
+
log, brands, companies = [], set(), set()
|
| 263 |
try:
|
| 264 |
+
log.append(f"1. 前往 PMDA (搜尋: {ing_ja})...")
|
| 265 |
page.goto("https://www.pmda.go.jp/PmdaSearch/iyakuSearch/", timeout=30000, wait_until="domcontentloaded")
|
|
|
|
|
|
|
|
|
|
| 266 |
|
| 267 |
+
# 尋找輸入框並按 Enter
|
| 268 |
+
page.locator('input#txtName, input[name="nameWord"], input[name="general_name"]').first.fill(ing_ja)
|
| 269 |
+
page.keyboard.press("Enter")
|
| 270 |
+
|
| 271 |
+
log.append("2. 等待表格 ResultList 或 錯誤訊息...")
|
| 272 |
+
# 🟢 修正:同時等待 表格 或 錯誤訊息,防止超時
|
| 273 |
+
page.wait_for_selector('table#ResultList, .errormsg, .non-result', timeout=15000)
|
| 274 |
+
log.append(f" -> 跳轉後網址: {page.url}")
|
| 275 |
+
|
| 276 |
+
soup = BeautifulSoup(page.content(), 'html.parser')
|
| 277 |
+
table = soup.find('table', id='ResultList')
|
| 278 |
+
|
| 279 |
+
if table:
|
| 280 |
+
rows = table.find_all('tr')
|
| 281 |
+
log.append(f"3. 找到表格,共 {len(rows)} 列。")
|
| 282 |
+
for tr in rows:
|
| 283 |
+
tds = tr.find_all('td')
|
| 284 |
+
if len(tds) >= 3:
|
| 285 |
+
title = tds[1].get_text(strip=True)
|
| 286 |
+
if not is_generic(title, "", ing_ja):
|
| 287 |
+
brands.add(clean_brand_name(title))
|
| 288 |
+
companies.add(tds[2].get_text(separator=" ", strip=True).replace('製造販売元/', ''))
|
| 289 |
+
else:
|
| 290 |
+
log.append("❌ 畫面上未出現 ResultList,查無此藥。")
|
| 291 |
+
|
| 292 |
+
if brands: return ", ".join(brands), ", ".join(companies), "\n".join(log)
|
| 293 |
+
return "查無原廠", "-", "\n".join(log)
|
| 294 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 295 |
|
| 296 |
def get_switzerland_originator(ing_de, page):
|
| 297 |
+
log, brands, companies = [], set(), set()
|
| 298 |
try:
|
| 299 |
+
log.append(f"1. 前往 Compendium (搜尋: {ing_de})...")
|
| 300 |
page.goto(f"https://compendium.ch/search?q={ing_de}", timeout=30000)
|
| 301 |
+
|
| 302 |
+
log.append("2. 等待網路請求完成 (networkidle)...")
|
| 303 |
page.wait_for_load_state('networkidle', timeout=20000)
|
| 304 |
|
| 305 |
+
soup = BeautifulSoup(page.content(), 'html.parser')
|
| 306 |
+
cards = soup.find_all('div', class_=re.compile('medicament-card'))
|
| 307 |
+
log.append(f"3. 畫面渲染完成,找到 {len(cards)} 筆卡片。")
|
| 308 |
+
|
| 309 |
+
for card in cards:
|
| 310 |
h3, strong = card.find('h3'), card.find('strong', class_='info')
|
| 311 |
if h3 and strong:
|
| 312 |
title, comp = h3.get_text(strip=True), strong.get_text(strip=True)
|
| 313 |
if not is_generic(title, comp, ing_de):
|
| 314 |
brands.add(clean_brand_name(title))
|
| 315 |
companies.add(comp)
|
| 316 |
+
if brands: return ", ".join(brands), ", ".join(companies), "\n".join(log)
|
| 317 |
+
return "查無原廠", "-", "\n".join(log) + "\n❌ 查無資料或全為學名藥"
|
| 318 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 319 |
|
| 320 |
def get_germany_originator(ing_de, page):
|
| 321 |
+
log, brands, companies = [], set(), set()
|
| 322 |
try:
|
| 323 |
+
log.append(f"1. 前往 Gelbe Liste (搜尋: {ing_de})...")
|
| 324 |
page.goto(f"https://www.gelbe-liste.de/profi-suche/results?substance={quote(ing_de)}", timeout=30000)
|
| 325 |
page.wait_for_selector('.product-list', timeout=15000)
|
| 326 |
+
|
| 327 |
+
soup = BeautifulSoup(page.content(), 'html.parser')
|
| 328 |
+
ul = soup.find('ul', class_='product-list')
|
| 329 |
if ul:
|
| 330 |
+
lis = ul.find_all('li')
|
| 331 |
+
log.append(f"2. 找到 {len(lis)} 筆結果,過濾平行輸入商。")
|
| 332 |
+
for li in lis:
|
| 333 |
h5, p_tag = li.find('h5'), li.find('p', class_='small')
|
| 334 |
if h5:
|
| 335 |
title, comp = h5.get_text(strip=True), (p_tag.get_text(strip=True) if p_tag else "-")
|
| 336 |
if not is_generic(title, comp, ing_de):
|
| 337 |
brands.add(clean_brand_name(title))
|
| 338 |
companies.add(comp)
|
| 339 |
+
if brands: return ", ".join(brands), ", ".join(companies), "\n".join(log)
|
| 340 |
+
return "查無原廠", "-", "\n".join(log) + "\n❌ 查無資料"
|
| 341 |
+
except Exception as e: return "執行失敗", "-", "\n".join(log) + f"\n錯誤: {str(e)}"
|
| 342 |
|
| 343 |
# ==========================================
|
| 344 |
# 🚀 主執行中樞
|
|
|
|
| 350 |
ing_de = ing_de_manual if ing_de_manual else translate_lang(ing_en, 'de')
|
| 351 |
results = []
|
| 352 |
|
| 353 |
+
# 1. API 模組 (不受瀏覽器阻擋)
|
| 354 |
usa_b, usa_c, usa_log = get_usa_originator(ing_en)
|
| 355 |
results.append(["🇺🇸 美國 (FDA)", usa_b, usa_c, usa_log])
|
| 356 |
|
|
|
|
| 391 |
# 🎨 UI 介面
|
| 392 |
# ==========================================
|
| 393 |
with gr.Blocks(title="十國原廠商品名智能檢索器") as demo:
|
| 394 |
+
gr.Markdown("## 🌐 跨國原廠商品名檢索器 (高階診斷版)")
|
| 395 |
|
| 396 |
with gr.Row():
|
| 397 |
ing_en = gr.Textbox(label="🧪 英文成分名 (必填)", placeholder="例如: Semaglutide")
|
|
|
|
| 403 |
search_btn = gr.Button("🚀 啟動十國查詢", variant="primary")
|
| 404 |
|
| 405 |
result_table = gr.Dataframe(
|
| 406 |
+
headers=["國家", "🌟 判定為原廠的商品名", "🏭 藥廠名稱", "🛠️ 系統狀態與詳細日誌"],
|
| 407 |
datatype=["str", "str", "str", "str"],
|
| 408 |
wrap=True, interactive=False
|
| 409 |
)
|