[3.7] bpo-32861: urllib.robotparser fix incomplete __str__ methods. (GH-5711) (GH-6795)
The urllib.robotparser's __str__ representation now includes wildcard
entries and the "Crawl-delay" and "Request-rate" fields.
(cherry picked from commit bd08a0af2d88c590ede762102bd42da3437e9980)
Co-authored-by: Michael Lazar <lazar.michael22@gmail.com>
diff --git a/Lib/urllib/robotparser.py b/Lib/urllib/robotparser.py
index daac29c..883ef24 100644
--- a/Lib/urllib/robotparser.py
+++ b/Lib/urllib/robotparser.py
@@ -190,7 +190,10 @@
return self.default_entry.req_rate
def __str__(self):
- return ''.join([str(entry) + "\n" for entry in self.entries])
+ entries = self.entries
+ if self.default_entry is not None:
+ entries = entries + [self.default_entry]
+ return '\n'.join(map(str, entries)) + '\n'
class RuleLine:
@@ -222,10 +225,15 @@
def __str__(self):
ret = []
for agent in self.useragents:
- ret.extend(["User-agent: ", agent, "\n"])
- for line in self.rulelines:
- ret.extend([str(line), "\n"])
- return ''.join(ret)
+ ret.append(f"User-agent: {agent}")
+ if self.delay is not None:
+ ret.append(f"Crawl-delay: {self.delay}")
+ if self.req_rate is not None:
+ rate = self.req_rate
+ ret.append(f"Request-rate: {rate.requests}/{rate.seconds}")
+ ret.extend(map(str, self.rulelines))
+ ret.append('') # for compatibility
+ return '\n'.join(ret)
def applies_to(self, useragent):
"""check if this entry applies to the specified agent"""