parser.py

import re
from logger import get_logger

log = get_logger()

def parse_html(html: str, final_url: str, word: str) -> str:
    """
    High-Performance Text Boundary Token Mapper optimized for reading 
    direct username landing pages and search loop arrays cleanly.
    """
    if not html or not html.strip():
        log.warning(f"🔍 Empty text payload passed to evaluation parser engine for target: {word}")
        return "AVAILABLE"

    # Flatten trailing breaks, tabs, and layout spaces into single line strings
    clean_html = " ".join(html.split())
    word_clean = word.strip().replace("@", "").lower()
    final_url_lower = final_url.lower()

    # ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    # ENGINE 1: DIRECT DOM PROFILE SCANNING
    # ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    if '/username/' in final_url_lower and '?query=' not in final_url_lower:
        log.info(f"🔎 Converted Parser executing Profile Mode parsing path for word: {word_clean}")
        
        status_match = re.search(r'class="tm-section-header-status[^"]*"[^>]*>\s*([^<]+?)\s*</span>', clean_html, re.IGNORECASE)
        if status_match:
            s = status_match.group(1).strip().lower()
            log.info(f"🔮 Direct Class Marker found for '{word_clean}': '{s}'")
            if "sold" in s: return "SOLD"
            if "taken" in s: return "TAKEN"
            if "auction" in s: return "ON_AUCTION"
            if "available" in s: return "AVAILABLE"
            if "sale" in s or "purchase" in s: return "FOR_SALE"

        if 'is already taken' in clean_html.lower() or 'make an offer' in clean_html.lower():
            log.info(f"🔮 Found fallback text marker 'taken/offer' for: {word_clean}")
            return "TAKEN"

    # ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    # ENGINE 2: REDIRECTED SEARCH RESULTS ROW LOOKUP
    # ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    log.info(f"🔎 Converted Parser executing Grid Mode search loop path for word: {word_clean}")
    search_regex = re.compile(rf'>@{word_clean}<.*?class="[^"]*status[^"]*"[^>]*>\s*([^<]+?)\s*<', re.IGNORECASE)
    match = search_regex.search(clean_html)
    
    if match:
        s = match.group(1).strip().lower()
        log.info(f"🔮 Search Grid Table status mapped for '{word_clean}': '{s}'")
        if "auction" in s or "bidding" in s: return "ON_AUCTION"
        if "sold" in s: return "SOLD"
        if "unavailable" in s: return "UNAVAILABLE"
        if "taken" in s or "offer" in s: return "TAKEN"
        if "sale" in s or "purchase" in s: return "FOR_SALE"

    # ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    # ENGINE 3: GLOBAL LAYOUT REGEX FALLBACK SCANS
    # ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    log.info(f"⚠️ Row boundaries missed for '{word_clean}'. Executing global text scans...")
    fallback_text = clean_html.lower()
    
    if "on auction" in fallback_text: return "ON_AUCTION"
    if "sold for" in fallback_text or "recently sold" in fallback_text: return "SOLD"
    if "taken" in fallback_text or "make an offer" in fallback_text: return "TAKEN"
    if "for sale" in fallback_text or "purchase" in fallback_text: return "FOR_SALE"
    if "unavailable" in fallback_text: return "UNAVAILABLE"

    return "AVAILABLE"