经常需要在 binlog 中查找一些日志信息,于是写了一个简单的脚本。对于非常巨大的 binlog 文件,该脚本可能会速度慢,毕竟还是用的 list,暂时没想到好办法。
详细看代码:
#/usr/bin/python#2016-04-12#search string in the binlogs#usage:#put this file into binlog-dir,exec as:#"python test.py 111 123 update" or #"python test.py 111 123 update and insert" or #"python test.py 111 123 update or delete"#the nums are the binlog-num.import sysimport subprocess def find_str(files): for file in files: comm = "mysqlbinlog {0}".format(file) lines = subprocess.Popen(comm, shell=True, stdout=subprocess.PIPE) lines = lines.stdout.readlines() for line in lines: line = line.lower() if len(sys.argv) == 4: if sys.argv[3] + ' ' in line: yield line elif len(sys.argv) == 6 and sys.argv[4] == "and": if sys.argv[3] + ' ' in line and sys.argv[5] + ' ' in line: yield line elif len(sys.argv) == 6 and sys.argv[4] == "or": if sys.argv[3] + ' ' in line or sys.argv[5] + ' ' in line: yield line if __name__ == "__main__": start = sys.argv[1] end = sys.argv[2] files = ["updatelog.{0:06d}".format(i) for i in range(int(start), int(end)+1)] f = find_str(files) for i in f: print(i)
第二次改进版本,由于 py2 不支持 yield from 语句,gen_concatenate() 可能有点绕;详细看代码:
#/usr/bin/python#2016-04-12#search string in the binlogs#usage:#put this file into binlog-dir,exec as "python test.py 111 123 update" or "python test.py 111 123 update and insert" or "python test.py 111 123 update or delete"#the nums are the binlog-num.import sysimport subprocessdef find_str(files): print(sys.argv) for file in files: comm = "mysqlbinlog {0}".format(file) lines = subprocess.Popen(comm, shell=True, stdout=subprocess.PIPE) lines = lines.stdout.readlines() yield lines #此处返回的是生成器对象def gen_concatenate(lines): #将多个生成器对象迭代返回 for i in lines: for it in i: yield itdef gen_grep(lines): for line in lines: line = line.lower() if len(sys.argv) == 4: if sys.argv[3] + ' ' in line: yield line elif len(sys.argv) == 6 and sys.argv[4] == "and": if sys.argv[3] + ' ' in line and sys.argv[5] + ' ' in line: yield line elif len(sys.argv) == 6 and sys.argv[4] == "or": if sys.argv[3] + ' ' in line or sys.argv[5] + ' ' in line: yield lineif __name__ == "__main__": start = sys.argv[1] end = sys.argv[2] files = ["updatelog.{0:06d}".format(i) for i in range(int(start), int(end)+1)] f = find_str(files) lines = gen_concatenate(f) greplines = gen_grep(lines) for i in greplines: print(i)
脚本1理解起来更加容易,将实现功能全部封装在一个函数体内;
更新:
使用了 re 正则匹配,有时候日志里面记录的表名是带反引号的,比如`user`这样,见代码:
#/usr/bin/python#2016-04-27#search string in the binlogs#usage:#put this file into binlog-dir,exec as "python test.py 111 123 update" or "python test.py 111 123 update and insert" or "python test.py 111 123 update or delete"#the nums are the binlog-num.import sysimport subprocessimport redef find_str(files): print(sys.argv) for file in files: comm = "mysqlbinlog {0}".format(file) lines = subprocess.Popen(comm, shell=True, stdout=subprocess.PIPE) lines = lines.stdout.readlines() yield linesdef gen_concatenate(lines): for i in lines: for it in i: yield itdef gen_grep(lines): for line in lines: line = line.lower() if len(sys.argv) == 4: # if sys.argv[3] + ' ' in line: # yield line str1 = sys.argv[3] str1_match = re.search(str1, line, re.I) if str1_match is not None: yield line elif len(sys.argv) == 6 and sys.argv[4] == "and": # if sys.argv[3] + ' ' in line and sys.argv[5] + ' ' in line: # yield line str1 = sys.argv[3] str2 = sys.argv[5] if re.search(r'\b{}\b'.format(str1), line, re.I) is not None and re.search(r'\b{}\b'.format(str2), line, re.I): yield line elif len(sys.argv) == 6 and sys.argv[4] == "or": # if sys.argv[3] + ' ' in line or sys.argv[5] + ' ' in line: # yield line str1 = sys.argv[3] str2 = sys.argv[5] if re.search(r'\b{}\b'.format(str1), line, re.I) is not None and re.search(r'\b{}\b'.format(str2), line, re.I): yield lineif __name__ == "__main__": start = sys.argv[1] end = sys.argv[2] files = ["updatelog.{0:06d}".format(i) for i in range(int(start), int(end)+1)] f = find_str(files) lines = gen_concatenate(f) greplines = gen_grep(lines) for i in greplines: print(i)search.py
posted on 2016-04-12 22:21 阅读( ...) 评论( ...)