=Start=<\/p>\n\n\n\n
\u5f53\u524d\u9700\u8981\u7528 Python \u6216 Java \u5bf9 SQL\u8bed\u53e5 \u8fdb\u884c\u89e3\u6790\uff0c\u4ee5\u63d0\u53d6\u51fa\u5176\u4e2d\u7684 \u8868\u540d\u3001\u5217\u540d \u7b49\u4fe1\u606f\uff0c\u65b9\u4fbf\u540e\u7eed\u8fdb\u884c\u7cbe\u51c6\u7684\u5b89\u5168\u95ee\u9898\u5206\u6790\u3002<\/p>\n\n\n\n
\u5982\u679c\u4ec5\u4ec5\u53ea\u662f\u63d0\u53d6\u8868\u540d\u7684\u8bdd\u8fd8\u6bd4\u8f83\u7b80\u5355\uff08\u76f4\u63a5\u7528\u5b57\u7b26\u4e32\u5207\u5206\u3001\u5339\u914d\u7b49\u5904\u7406\u5c31\u53ef\u4ee5\u641e\u5b9a\uff09\uff0c\u4f46\u5982\u679c\u8fd8\u8981\u63d0\u53d6\u5217\u540d\uff08\u4e14SQL\u6bd4\u8f83\u590d\u6742\uff09\u7684\u8bdd\uff0c\u5c31\u4e0d\u90a3\u4e48\u5bb9\u6613\u641e\u5b9a\u4e86\uff0c\u8fd9\u65f6\u5c31\u9700\u8981\u5bf9SQL\u8fdb\u884c\u8bed\u6cd5\u5206\u6790\u4ee5\u8fdb\u884c\u51c6\u786e\u63d0\u53d6\u3002\u95ee\u4e86\u4e00\u4e0b\u522b\u4eba\uff0c\u8bf4 ANTLR4 \u53ef\u4ee5\u505a\uff0c\u4e8e\u662f\u5728\u7f51\u4e0a\u641c\u4e86\u4e00\u4e9b\u8d44\u6599\uff0c\u5728\u6b64\u5b9e\u8df5\u6574\u7406\u4e00\u4e0b\uff0c\u65b9\u4fbf\u4ee5\u540e\u53c2\u8003\u3002<\/p>\n\n\n\n
\u8fd9\u91cc\u5148\u4ecb\u7ecd\u4e00\u4e0b\u5982\u4f55\u901a\u8fc7\u5b57\u7b26\u4e32\u5207\u5206\u548c\u6b63\u5219\u7684\u65b9\u5f0f\u6765\u63d0\u53d6SQL\u4e2d\u7684\u8868\u540d\uff08\u5217\u540d\u63d0\u53d6\u6bd4\u8f83\u590d\u6742\uff0c\u4e0d\u592a\u597d\u5b9e\u73b0\uff0c\u540e\u7eed\u627e\u5230\u597d\u7684\u65b9\u6cd5\u4e86\u518d\u8865\u5145\u8fdb\u6765\uff09\u3002<\/p>\n\n\n\n
\u65b9\u6cd5\u4e00\uff08\u5b57\u7b26\u4e32\u5207\u5206\uff09\uff1a<\/p>\n\n\n\n
def tables_in_query(sql):\n replace_list = ['\\n', '(', ')', '*', '=', '`']\n for i in replace_list:\n sql = sql.replace(i, ' ')\n sql = sql.split()\n res = set()\n for i in range(1, len(sql)):\n if sql[i-1].lower() in ['from', 'join'] and sql[i].lower() != 'select': \n res.add(sql[i])\n # print(res) # ['table1', 'table2', 'table3', 'table4', 'table5']\n return list(res)<\/code><\/pre>\n\n\n\n\u65b9\u6cd5\u4e8c\uff08\u6b63\u5219\u63d0\u53d6\uff09\uff1a<\/p>\n\n\n\n
import re\n\ndef tables_in_query2(sql_str):\n # remove the \/* *\/ comments (\u5148\u79fb\u9664\u5757\u6ce8\u91ca)\n q = re.sub(r\"\/\\*[^*]*\\*+(?:[^*\/][^*]*\\*+)*\/\", \"\", sql_str)\n\n # remove whole line -- and # comments (\u518d\u79fb\u9664\u6574\u884c\u6ce8\u91ca)\n lines = [line for line in q.splitlines() if not re.match(\"^\\s*(--|#)\", line)]\n\n # remove trailing -- and # comments (\u518d\u79fb\u9664\u884c\u5c3e\u6ce8\u91ca)\n q = \" \".join([re.split(\"--|#\", line)[0] for line in lines])\n\n # split on blanks, parens and semicolons (\u7528\u300c\u7a7a\u767d\u7b26\u3001\u62ec\u53f7\u3001\u5206\u53f7\u300d\u4f5c\u4e3a\u5206\u9694\u7b26\u8fdb\u884c\u5207\u5206)\n tokens = re.split(r\"[\\s)(;`]+\", q)\n\n # scan the tokens. if we see a FROM or JOIN, we set the get_next\n # flag, and grab the next one (unless it's SELECT).\n # \u626b\u63cf\u7ecf\u8fc7\u4e0a\u8ff0\u5206\u8bcd\u7684\u5217\u8868\uff0c\u5f53\u770b\u5230 FROM\/JOIN \u5173\u952e\u5b57\u65f6\uff0c\u83b7\u53d6\u4e0b\u4e00\u4e2a\u975e\u300c\u7a7a\/select\u300d\u7684\u4f5c\u4e3a\u8868\u540d\n\n result = set()\n get_next = False\n for tok in tokens:\n if get_next:\n if tok.lower() not in [\"\", \"select\"]:\n result.add(tok)\n get_next = False\n get_next = tok.lower() in [\"from\", \"join\"]\n\n return result<\/code><\/pre>\n\n\n\n\u5176\u5b9e\u6b63\u5219\u7684\u8fd9\u4e2a\u65b9\u6cd5\u8003\u8651\u7684\u633a\u5b8c\u6574\u7684\uff08\u6bd4\u5982\u5728\u79fb\u9664\u6ce8\u91ca\u90a3\u91cc\uff09\uff0c\u4f46\u662f\u4e0a\u9762\u7684\u5b57\u7b26\u4e32\u7248\u672c\u4f1a\u66f4\u7b80\u5355\u800c\u4e14\u6548\u7387\u66f4\u9ad8\uff08\u5728\u7ebf\u4e0aMySQL\u64cd\u4f5c\u4e2d\u4e00\u822c\u4e0d\u4f1a\u6709\u6ce8\u91ca\uff09\uff0c\u6240\u4ee5\u53ef\u4ee5\u6309\u9700\u9009\u62e9\u3002\u4e0b\u9762\u63d0\u4f9b\u4e00\u4e2a\u6d4b\u8bd5\u6548\u679c\u7528\u7684SQL\uff1a<\/p>\n\n\n\n
sql = \"\"\"SELECT a.time_updated_server\/1000,
content,
nick,
name
FROM table1 a
JOIN table2 b ON a.sender_id = b.user_id
JOIN table3 c ON a.channel_id = c.channel_id
JOIN table4 d ON c.store_id = d.store_id
WHERE sender_id NOT IN
(SELECT user_id
FROM table5
WHERE store_id IN ('agent_store:1', 'ask:1'))
AND to_timestamp(a.time_updated_server\/1000)::date >= '2014-05-01'
GROUP BY 1,2,3,4
HAVING sum(1) > 500
ORDER BY 1 ASC
\"\"\"<\/code><\/p>\n\n\n\n\u63d0\u53d6\u5217\u540d\u7684\u65b9\u5f0f\u5176\u5b9e\u7167\u7740\u8868\u540d\u7684\u65b9\u6cd5\u6539\u4e00\u4e0b\u5e94\u8be5\u4e5fOK\uff0c\u4f46\u8fd9\u91cc\u6682\u65f6\u5c31\u4e0d\u653e\u51fa\u6765\u4e86\uff0c\u6709\u9700\u8981\u7684\u53ef\u4ee5\u6309\u9700\u8fdb\u884c\u4fee\u6539\u3002<\/p>\n\n\n\n
<\/p>\n\n\n\n
\u53c2\u8003\u94fe\u63a5\uff1a<\/h5>\n\n\n\n