import urllib.parse
import re
import unicodedata

html_path = '/home/x79/sisvietnamvn_01/BV_DHYD_HCM/Các chuyên khoa tại Bệnh viện Đại học Y Dược TP. Hồ Chí Minh.html'
with open(html_path, 'r', encoding='utf-8') as f:
    text = f.read()

def slugify(value):
    value = unicodedata.normalize('NFKD', value).encode('ascii', 'ignore').decode('utf-8')
    value = re.sub(r'[^\w\s-]', '', value).strip().lower()
    return re.sub(r'[-\s]+', '-', value)

db_slugs = [
    "chuyen-khoa-tai-mui-hong", "khoa-chan-doan-hinh-anh", "khoa-cap-cuu",
    "khoa-dinh-duong-tiet-che", "khoa-duoc", "khoa-kham-benh",
    "don-vi-kiem-soat-nhiem-khuan", "khoa-ngoai-tong-hop", "tim-mach", 
    "phau-thuat-gay-me-hoi-suc", "khoa-than-kinh-dot-quy", 
    "khoa-vat-ly-tri-lieu-phuc-hoi-chuc-nang", "khoa-xet-nghiem",
    "phong-quan-ly-van-hanh", "don-vi-can-thiep-mach-dsa",
    "don-vi-cap-cuu-ngoai-vien", "don-vi-kham-suc-khoe-ngoai-vien"
]

manual_mapping = {
    "noi tong hop": "tim-mach",
    "phau thuat gay me hoi suc phong mo": "phau-thuat-gay-me-hoi-suc",
    "dsa": "don-vi-can-thiep-mach-dsa",
    "tai mui hong": "chuyen-khoa-tai-mui-hong",
    "kiem soat nhiem khuan": "don-vi-kiem-soat-nhiem-khuan",
    "kham suc khoe ngoai vien": "don-vi-kham-suc-khoe-ngoai-vien",
    "cap cuu ngoai vien": "don-vi-cap-cuu-ngoai-vien",
    "chan doan hinh anh": "khoa-chan-doan-hinh-anh",
    "vat ly tri lieu phuc hoi chuc nang": "khoa-vat-ly-tri-lieu-phuc-hoi-chuc-nang",
    "phuc hoi chuc nang": "khoa-vat-ly-tri-lieu-phuc-hoi-chuc-nang",
    "kham benh": "khoa-kham-benh",
    "cap cuu": "khoa-cap-cuu",
    "than kinh": "khoa-than-kinh-dot-quy",
    "ngoai tong hop": "khoa-ngoai-tong-hop",
    "xet nghiem": "khoa-xet-nghiem",
    "dinh duong tiet che": "khoa-dinh-duong-tiet-che"
}

chunks = text.split('href="https://bvdaihoc.com.vn/chuyen-khoa/')

icons = {}
for chunk in chunks[1:]:
    slug_in_href = chunk.split('"')[0]
    if '?' in slug_in_href: continue
    
    # Try to find absolute img src in srcset
    src = None
    srcset_m = re.search(r'srcset=\"([^\"]+)\"', chunk)
    if srcset_m:
        # get the first url from srcset
        srcsets = srcset_m.group(1).split(', ')
        first = srcsets[0].split(' ')[0]
        if first.startswith('/_next/image?url='):
            src = urllib.parse.unquote(first.replace('/_next/image?url=', '').split('&')[0])

    if not src:
        # look for absolute console.bvdaihoc url anywhere in chunk
        abs_m = re.search(r'https%3A%2F%2Fconsole.bvdaihoc.com.vn%2Fuploads%2Fchuyen-khoa%2F[^&\s\"\'\\]+', chunk)
        if abs_m:
            src = urllib.parse.unquote(abs_m.group(0))

    if src:
        title = slug_in_href
        text_match = re.findall(r'>([^<]+)<', chunk[:1000])
        texts = [t.strip() for t in text_match if len(t.strip()) > 3 and not t.strip().startswith('&')]
        if texts:
            title = texts[0]
            
        s_text = slugify(title)
        matched = None
        for key, val in manual_mapping.items():
            if slugify(key) in s_text:
                matched = val
                break
        
        if not matched:
            for s in db_slugs:
                if s_text in s or s in s_text or s_text.replace("khoa-", "") in s:
                    matched = s
                    break
        
        if matched:
            icons[matched] = src
        else:
            s_slug = slugify(slug_in_href)
            for s in db_slugs:
                if s_slug in s or s in s_slug:
                    icons[s] = src
                    break

switch_code = '                <th:block th:switch="${spec.slug}">\n'
for slug, src in sorted(icons.items()):
    switch_code += f'                  <!-- {slug} -->\n'
    switch_code += f'                  <th:block th:case="\'{slug}\'">\n'
    switch_code += f'                    <img src="{src}" alt="Icon" class="w-12 h-12 object-contain mx-auto" />\n'
    switch_code += f'                  </th:block>\n'

switch_code += '                  <!-- Default -->\n'
switch_code += '                  <th:block th:case="*">\n'
switch_code += '                    <div class="w-12 h-12 rounded-full bg-primary-100 flex items-center justify-center text-primary-600 mx-auto">\n'
switch_code += '                      <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="w-6 h-6">\n'
switch_code += '                        <path d="M8 2v4M16 2v4M3 10h18M5 4h14a2 2 0 012 2v14a2 2 0 01-2 2H5a2 2 0 01-2-2V6a2 2 0 012-2zM9 14h6M12 11v6"/>\n'
switch_code += '                      </svg>\n'
switch_code += '                    </div>\n'
switch_code += '                  </th:block>\n'
switch_code += '                </th:block>'

with open('/tmp/icons_block_img.txt', 'w', encoding='utf-8') as f:
    f.write(switch_code)
    
print("Wrote block to /tmp/icons_block_img.txt")