更具体并使用字符类 ([...]):
(?P<host>[\d.]+)[-\s]+
(?P<user_name>\w+)\s+
\[(?P<time>[^][]+)\]\s+
"(?P<request>[^"]+)"
见a demo on regex101.com。
或者 - 完全使用解析器:
from parsimonious.grammar import Grammar
from parsimonious.nodes import NodeVisitor
string = '30.95.91.251 - larson8319 [21/Jun/2019:16:02:02 -0700] "PUT /one-to-one/whiteboard HTTP/1.0" 401 7270'
class LogVisitor(NodeVisitor):
grammar = Grammar(
r"""
entry = ip user time request rest
ip = ~"[\d.]+" junk
user = ~"\w+" junk
time = "[" ~"[^][]+" "]" junk
request = '"' ~"[^\"]+" '"'
junk = ~"[-\s]*"
rest = ~".*"
"""
)
def generic_visit(self, node, visited_children):
return visited_children or node
def visit_entry(self, node, visited_children):
ip, user, time, request, *_ = visited_children
return dict([ip, user, time, request])
def __clean__(self, node, visited_children, first=False):
if first:
_, what, *_ = visited_children
else:
what, *_ = visited_children
return what.text
def visit_ip(self, node, visited_children):
return ('ip', self.__clean__(node, visited_children))
def visit_user(self, node, visited_children):
return ('user', self.__clean__(node, visited_children))
def visit_time(self, node, visited_children):
return ('time', self.__clean__(node, visited_children, True))
def visit_request(self, node, visited_children):
return ('request', self.__clean__(node, visited_children, True))
lv = LogVisitor()
result = lv.parse(string)
print(result)
这会产生
{'ip': '30.95.91.251', 'user': 'larson8319', 'time': '21/Jun/2019:16:02:02 -0700', 'request': 'PUT /one-to-one/whiteboard HTTP/1.0'}