如何用Python结合rules.list文件通过正则提取命名实体？

步骤说明 #

1. 解析规则文件：读取rules.list中的每条规则，拆分出正则匹配模式和要提取的捕获组索引。比如规则app install (.*) 1，对应正则模式为app install (.*)，需提取第1个捕获组的内容。
2. 匹配命令字符串：遍历每条输入命令，用解析后的正则规则尝试匹配，匹配成功则提取对应捕获组的内容。

代码实现 #

import re

def load_rules(rule_file):
    rules = []
    with open(rule_file, 'r') as f:
        for line in f:
            line = line.strip()
            if not line:
                continue
            # 拆分规则：格式为 "匹配模式 捕获组索引"
            parts = line.rsplit(' ', 1)
            if len(parts) != 2:
                continue
            pattern_str, group_idx = parts
            try:
                group_idx = int(group_idx)
                # 编译正则，忽略大小写（可按需调整）
                pattern = re.compile(pattern_str, re.IGNORECASE)
                rules.append((pattern, group_idx))
            except ValueError:
                # 跳过格式错误的规则
                continue
    return rules

def extract_entities(commands, rules):
    entities = []
    for cmd in commands:
        matched = False
        for pattern, group_idx in rules:
            match = pattern.search(cmd)
            if match and group_idx <= len(match.groups()):
                entity = match.group(group_idx).strip()
                entities.append(entity)
                matched = True
                break  # 匹配到第一条规则即停止，可按需修改
        if not matched:
            entities.append(None)  # 未匹配时的默认值，可按需调整
    return entities

# 示例调用
commands = [
    "Open Youtube",
    "Install PlayStore App",
    "Go to Call of Duty app"
]

# 需确保rules.list中的规则与命令对应，示例规则如下：
# open (.*) 1
# install (.*) app 1
# go to (.*) app 1
rules = load_rules('rules.list')
result = extract_entities(commands, rules)
print(result)  # 输出: ['Youtube', 'PlayStore', 'Call of Duty']

注意事项 #

• 保证rules.list规则格式统一，每条规则末尾为捕获组索引数字
• 正则规则需与命令字符串格式匹配，比如Open Youtube对应规则open (.*) 1，Install PlayStore App对应install (.*) app 1
• 可根据需求调整正则匹配选项，比如是否要求全匹配、是否区分大小写等

内容的提问来源于stack exchange，提问作者Turing101