sbnews/docker/crawler/src/parser/nuxt_parser.py

"""
从懂球帝 liveDetail 页面的 window.__NUXT__ IIFE 中解析比赛详情数据
"""
import re
from typing import Optional, Dict, Any


def parse_js_args(args_str: str) -> list:
    """解析 JS IIFE 的实参列表，返回 Python 值列表"""
    results = []
    i = 0
    n = len(args_str)
    while i < n:
        c = args_str[i]
        if c in (' ', '\n', '\r', '\t'):
            i += 1
            continue
        if c == ',':
            i += 1
            continue
        if c == '"':
            j = i + 1
            while j < n:
                if args_str[j] == '\\':
                    j += 2
                    continue
                if args_str[j] == '"':
                    break
                j += 1
            raw = args_str[i:j+1]
            inner = raw[1:-1]
            if '\\u' in inner:
                val = inner.encode('utf-8').decode('unicode_escape')
            else:
                val = inner.replace('\\n', '\n').replace('\\t', '\t').replace('\\"', '"').replace('\\\\', '\\')
            results.append(val)
            i = j + 1
            continue
        if args_str[i:i+4] == 'true':
            results.append(True)
            i += 4
            continue
        if args_str[i:i+5] == 'false':
            results.append(False)
            i += 5
            continue
        if args_str[i:i+4] == 'null':
            results.append(None)
            i += 4
            continue
        if args_str[i:i+6] == 'void 0':
            results.append(None)
            i += 6
            continue
        if c in '0123456789.-':
            j = i + 1
            while j < n and args_str[j] in '0123456789.eE+-':
                j += 1
            num_str = args_str[i:j]
            try:
                val = int(num_str)
            except ValueError:
                val = float(num_str)
            results.append(val)
            i = j
            continue
        if c == '{':
            depth = 1
            j = i + 1
            while j < n and depth > 0:
                if args_str[j] == '{':
                    depth += 1
                elif args_str[j] == '}':
                    depth -= 1
                elif args_str[j] == '"':
                    j += 1
                    while j < n and args_str[j] != '"':
                        if args_str[j] == '\\':
                            j += 1
                        j += 1
                j += 1
            results.append(args_str[i:j])
            i = j
            continue
        if c == '[':
            depth = 1
            j = i + 1
            while j < n and depth > 0:
                if args_str[j] == '[':
                    depth += 1
                elif args_str[j] == ']':
                    depth -= 1
                elif args_str[j] == '"':
                    j += 1
                    while j < n and args_str[j] != '"':
                        if args_str[j] == '\\':
                            j += 1
                        j += 1
                j += 1
            results.append(args_str[i:j])
            i = j
            continue
        j = i
        while j < n and args_str[j] not in ',)]}':
            j += 1
        token = args_str[i:j].strip()
        results.append(token)
        i = j

    return results


def _decode_str(val_expr: str) -> str:
    """解码 JS 字符串字面量（去掉引号）"""
    inner = val_expr[1:-1]
    if '\\u' in inner:
        try:
            return inner.encode('utf-8').decode('unicode_escape')
        except Exception:
            return inner
    return inner.replace('\\n', '\n').replace('\\t', '\t').replace('\\"', '"').replace('\\\\', '\\')


def _resolve_value(val_expr: str, var_map: dict):
    """将 JS 赋值右侧表达式解析为 Python 值"""
    if val_expr in var_map:
        return var_map[val_expr]
    if val_expr.startswith('"') and val_expr.endswith('"'):
        return _decode_str(val_expr)
    if val_expr == 'true':
        return True
    if val_expr == 'false':
        return False
    if val_expr == 'null':
        return None
    try:
        return int(val_expr)
    except ValueError:
        pass
    try:
        return float(val_expr)
    except ValueError:
        pass
    return val_expr


def parse_nuxt_match_sample(html: str) -> Optional[Dict[str, Any]]:
    """从 liveDetail HTML 中解析 matchSample 对象"""
    pattern = r'window\.__NUXT__\s*=\s*\(function\(([^)]+)\)\s*\{(.+)\}\((.+)\)\);'
    m = re.search(pattern, html, re.DOTALL)
    if not m:
        return None

    params = [p.strip() for p in m.group(1).split(',')]
    body = m.group(2)
    args_str = m.group(3)

    args = parse_js_args(args_str)

    var_map = {}
    for i, p in enumerate(params):
        if i < len(args):
            var_map[p] = args[i]

    ms_var_match = re.search(r'(\w+)\.match_id\s*=\s*(\w+)', body)
    if not ms_var_match:
        return None

    ms_var = ms_var_match.group(1)

    assign_pattern = re.compile(rf'\b{re.escape(ms_var)}\.(\w+)\s*=\s*(.+?)\s*;', re.MULTILINE)
    match_sample = {}

    for am in assign_pattern.finditer(body):
        field = am.group(1)
        val_expr = am.group(2).strip()
        match_sample[field] = _resolve_value(val_expr, var_map)

    return match_sample if match_sample else None