From 960e848f0d32399824d61dce2ff5736bb596ed44 Mon Sep 17 00:00:00 2001 From: Berker Peksag Date: Thu, 8 Oct 2015 12:27:06 +0300 Subject: Issue #16099: RobotFileParser now supports Crawl-delay and Request-rate extensions. Patch by Nikolay Bogoychev. --- Doc/library/urllib.robotparser.rst | 30 ++++++++++++- Doc/whatsnew/3.6.rst | 8 ++++ Lib/test/test_robotparser.py | 90 +++++++++++++++++++++++++++++--------- Lib/urllib/robotparser.py | 39 ++++++++++++++++- Misc/ACKS | 1 + Misc/NEWS | 5 ++- 6 files changed, 147 insertions(+), 26 deletions(-) diff --git a/Doc/library/urllib.robotparser.rst b/Doc/library/urllib.robotparser.rst index f179de2..c2e1bef 100644 --- a/Doc/library/urllib.robotparser.rst +++ b/Doc/library/urllib.robotparser.rst @@ -53,15 +53,41 @@ structure of :file:`robots.txt` files, see http://www.robotstxt.org/orig.html. Sets the time the ``robots.txt`` file was last fetched to the current time. + .. method:: crawl_delay(useragent) -The following example demonstrates basic use of the RobotFileParser class. + Returns the value of the ``Crawl-delay`` parameter from ``robots.txt`` + for the *useragent* in question. If there is no such parameter or it + doesn't apply to the *useragent* specified or the ``robots.txt`` entry + for this parameter has invalid syntax, return ``None``. + + .. versionadded:: 3.6 + + .. method:: request_rate(useragent) + + Returns the contents of the ``Request-rate`` parameter from + ``robots.txt`` in the form of a :func:`~collections.namedtuple` + ``(requests, seconds)``. If there is no such parameter or it doesn't + apply to the *useragent* specified or the ``robots.txt`` entry for this + parameter has invalid syntax, return ``None``. + + .. versionadded:: 3.6 + + +The following example demonstrates basic use of the :class:`RobotFileParser` +class:: >>> import urllib.robotparser >>> rp = urllib.robotparser.RobotFileParser() >>> rp.set_url("http://www.musi-cal.com/robots.txt") >>> rp.read() + >>> rrate = rp.request_rate("*") + >>> rrate.requests + 3 + >>> rrate.seconds + 20 + >>> rp.crawl_delay("*") + 6 >>> rp.can_fetch("*", "http://www.musi-cal.com/cgi-bin/search?city=San+Francisco") False >>> rp.can_fetch("*", "http://www.musi-cal.com/") True - diff --git a/Doc/whatsnew/3.6.rst b/Doc/whatsnew/3.6.rst index dd35c9a..3080820 100644 --- a/Doc/whatsnew/3.6.rst +++ b/Doc/whatsnew/3.6.rst @@ -119,6 +119,14 @@ datetime (Contributed by Ashley Anderson in :issue:`12006`.) +urllib.robotparser +------------------ + +:class:`~urllib.robotparser.RobotFileParser` now supports ``Crawl-delay`` and +``Request-rate`` extensions. +(Contributed by Nikolay Bogoychev in :issue:`16099`.) + + Optimizations ============= diff --git a/Lib/test/test_robotparser.py b/Lib/test/test_robotparser.py index d01266f..90b3072 100644 --- a/Lib/test/test_robotparser.py +++ b/Lib/test/test_robotparser.py @@ -1,6 +1,7 @@ import io import unittest import urllib.robotparser +from collections import namedtuple from urllib.error import URLError, HTTPError from urllib.request import urlopen from test import support @@ -12,7 +13,8 @@ except ImportError: class RobotTestCase(unittest.TestCase): - def __init__(self, index=None, parser=None, url=None, good=None, agent=None): + def __init__(self, index=None, parser=None, url=None, good=None, + agent=None, request_rate=None, crawl_delay=None): # workaround to make unittest discovery work (see #17066) if not isinstance(index, int): return @@ -25,6 +27,8 @@ class RobotTestCase(unittest.TestCase): self.url = url self.good = good self.agent = agent + self.request_rate = request_rate + self.crawl_delay = crawl_delay def runTest(self): if isinstance(self.url, tuple): @@ -34,6 +38,18 @@ class RobotTestCase(unittest.TestCase): agent = self.agent if self.good: self.assertTrue(self.parser.can_fetch(agent, url)) + self.assertEqual(self.parser.crawl_delay(agent), self.crawl_delay) + # if we have actual values for request rate + if self.request_rate and self.parser.request_rate(agent): + self.assertEqual( + self.parser.request_rate(agent).requests, + self.request_rate.requests + ) + self.assertEqual( + self.parser.request_rate(agent).seconds, + self.request_rate.seconds + ) + self.assertEqual(self.parser.request_rate(agent), self.request_rate) else: self.assertFalse(self.parser.can_fetch(agent, url)) @@ -43,15 +59,17 @@ class RobotTestCase(unittest.TestCase): tests = unittest.TestSuite() def RobotTest(index, robots_txt, good_urls, bad_urls, - agent="test_robotparser"): + request_rate, crawl_delay, agent="test_robotparser"): lines = io.StringIO(robots_txt).readlines() parser = urllib.robotparser.RobotFileParser() parser.parse(lines) for url in good_urls: - tests.addTest(RobotTestCase(index, parser, url, 1, agent)) + tests.addTest(RobotTestCase(index, parser, url, 1, agent, + request_rate, crawl_delay)) for url in bad_urls: - tests.addTest(RobotTestCase(index, parser, url, 0, agent)) + tests.addTest(RobotTestCase(index, parser, url, 0, agent, + request_rate, crawl_delay)) # Examples from http://www.robotstxt.org/wc/norobots.html (fetched 2002) @@ -65,14 +83,18 @@ Disallow: /foo.html good = ['/','/test.html'] bad = ['/cyberworld/map/index.html','/tmp/xxx','/foo.html'] +request_rate = None +crawl_delay = None -RobotTest(1, doc, good, bad) +RobotTest(1, doc, good, bad, request_rate, crawl_delay) # 2. doc = """ # robots.txt for http://www.example.com/ User-agent: * +Crawl-delay: 1 +Request-rate: 3/15 Disallow: /cyberworld/map/ # This is an infinite virtual URL space # Cybermapper knows where to go. @@ -83,8 +105,10 @@ Disallow: good = ['/','/test.html',('cybermapper','/cyberworld/map/index.html')] bad = ['/cyberworld/map/index.html'] +request_rate = None # The parameters should be equal to None since they +crawl_delay = None # don't apply to the cybermapper user agent -RobotTest(2, doc, good, bad) +RobotTest(2, doc, good, bad, request_rate, crawl_delay) # 3. doc = """ @@ -95,14 +119,18 @@ Disallow: / good = [] bad = ['/cyberworld/map/index.html','/','/tmp/'] +request_rate = None +crawl_delay = None -RobotTest(3, doc, good, bad) +RobotTest(3, doc, good, bad, request_rate, crawl_delay) # Examples from http://www.robotstxt.org/wc/norobots-rfc.html (fetched 2002) # 4. doc = """ User-agent: figtree +Crawl-delay: 3 +Request-rate: 9/30 Disallow: /tmp Disallow: /a%3cd.html Disallow: /a%2fb.html @@ -115,8 +143,17 @@ bad = ['/tmp','/tmp.html','/tmp/a.html', '/~joe/index.html' ] -RobotTest(4, doc, good, bad, 'figtree') -RobotTest(5, doc, good, bad, 'FigTree Robot libwww-perl/5.04') +request_rate = namedtuple('req_rate', 'requests seconds') +request_rate.requests = 9 +request_rate.seconds = 30 +crawl_delay = 3 +request_rate_bad = None # not actually tested, but we still need to parse it +crawl_delay_bad = None # in order to accommodate the input parameters + + +RobotTest(4, doc, good, bad, request_rate, crawl_delay, 'figtree' ) +RobotTest(5, doc, good, bad, request_rate_bad, crawl_delay_bad, + 'FigTree Robot libwww-perl/5.04') # 6. doc = """ @@ -125,14 +162,18 @@ Disallow: /tmp/ Disallow: /a%3Cd.html Disallow: /a/b.html Disallow: /%7ejoe/index.html +Crawl-delay: 3 +Request-rate: 9/banana """ good = ['/tmp',] # XFAIL: '/a%2fb.html' bad = ['/tmp/','/tmp/a.html', '/a%3cd.html','/a%3Cd.html',"/a/b.html", '/%7Ejoe/index.html'] +crawl_delay = 3 +request_rate = None # since request rate has invalid syntax, return None -RobotTest(6, doc, good, bad) +RobotTest(6, doc, good, bad, None, None) # From bug report #523041 @@ -140,12 +181,16 @@ RobotTest(6, doc, good, bad) doc = """ User-Agent: * Disallow: /. +Crawl-delay: pears """ good = ['/foo.html'] -bad = [] # Bug report says "/" should be denied, but that is not in the RFC +bad = [] # bug report says "/" should be denied, but that is not in the RFC + +crawl_delay = None # since crawl delay has invalid syntax, return None +request_rate = None -RobotTest(7, doc, good, bad) +RobotTest(7, doc, good, bad, crawl_delay, request_rate) # From Google: http://www.google.com/support/webmasters/bin/answer.py?hl=en&answer=40364 @@ -154,12 +199,15 @@ doc = """ User-agent: Googlebot Allow: /folder1/myfile.html Disallow: /folder1/ +Request-rate: whale/banana """ good = ['/folder1/myfile.html'] bad = ['/folder1/anotherfile.html'] +crawl_delay = None +request_rate = None # invalid syntax, return none -RobotTest(8, doc, good, bad, agent="Googlebot") +RobotTest(8, doc, good, bad, crawl_delay, request_rate, agent="Googlebot") # 9. This file is incorrect because "Googlebot" is a substring of # "Googlebot-Mobile", so test 10 works just like test 9. @@ -174,12 +222,12 @@ Allow: / good = [] bad = ['/something.jpg'] -RobotTest(9, doc, good, bad, agent="Googlebot") +RobotTest(9, doc, good, bad, None, None, agent="Googlebot") good = [] bad = ['/something.jpg'] -RobotTest(10, doc, good, bad, agent="Googlebot-Mobile") +RobotTest(10, doc, good, bad, None, None, agent="Googlebot-Mobile") # 11. Get the order correct. doc = """ @@ -193,12 +241,12 @@ Disallow: / good = [] bad = ['/something.jpg'] -RobotTest(11, doc, good, bad, agent="Googlebot") +RobotTest(11, doc, good, bad, None, None, agent="Googlebot") good = ['/something.jpg'] bad = [] -RobotTest(12, doc, good, bad, agent="Googlebot-Mobile") +RobotTest(12, doc, good, bad, None, None, agent="Googlebot-Mobile") # 13. Google also got the order wrong in #8. You need to specify the @@ -212,7 +260,7 @@ Disallow: /folder1/ good = ['/folder1/myfile.html'] bad = ['/folder1/anotherfile.html'] -RobotTest(13, doc, good, bad, agent="googlebot") +RobotTest(13, doc, good, bad, None, None, agent="googlebot") # 14. For issue #6325 (query string support) @@ -224,7 +272,7 @@ Disallow: /some/path?name=value good = ['/some/path'] bad = ['/some/path?name=value'] -RobotTest(14, doc, good, bad) +RobotTest(14, doc, good, bad, None, None) # 15. For issue #4108 (obey first * entry) doc = """ @@ -238,7 +286,7 @@ Disallow: /another/path good = ['/another/path'] bad = ['/some/path'] -RobotTest(15, doc, good, bad) +RobotTest(15, doc, good, bad, None, None) # 16. Empty query (issue #17403). Normalizing the url first. doc = """ @@ -250,7 +298,7 @@ Disallow: /another/path? good = ['/some/path?'] bad = ['/another/path?'] -RobotTest(16, doc, good, bad) +RobotTest(16, doc, good, bad, None, None) class RobotHandler(BaseHTTPRequestHandler): diff --git a/Lib/urllib/robotparser.py b/Lib/urllib/robotparser.py index 4fbb0cb..4ac553a 100644 --- a/Lib/urllib/robotparser.py +++ b/Lib/urllib/robotparser.py @@ -10,7 +10,9 @@ http://www.robotstxt.org/norobots-rfc.txt """ -import urllib.parse, urllib.request +import collections +import urllib.parse +import urllib.request __all__ = ["RobotFileParser"] @@ -120,10 +122,29 @@ class RobotFileParser: if state != 0: entry.rulelines.append(RuleLine(line[1], True)) state = 2 + elif line[0] == "crawl-delay": + if state != 0: + # before trying to convert to int we need to make + # sure that robots.txt has valid syntax otherwise + # it will crash + if line[1].strip().isdigit(): + entry.delay = int(line[1]) + state = 2 + elif line[0] == "request-rate": + if state != 0: + numbers = line[1].split('/') + # check if all values are sane + if (len(numbers) == 2 and numbers[0].strip().isdigit() + and numbers[1].strip().isdigit()): + req_rate = collections.namedtuple('req_rate', + 'requests seconds') + entry.req_rate = req_rate + entry.req_rate.requests = int(numbers[0]) + entry.req_rate.seconds = int(numbers[1]) + state = 2 if state == 2: self._add_entry(entry) - def can_fetch(self, useragent, url): """using the parsed robots.txt decide if useragent can fetch url""" if self.disallow_all: @@ -153,6 +174,18 @@ class RobotFileParser: # agent not found ==> access granted return True + def crawl_delay(self, useragent): + for entry in self.entries: + if entry.applies_to(useragent): + return entry.delay + return None + + def request_rate(self, useragent): + for entry in self.entries: + if entry.applies_to(useragent): + return entry.req_rate + return None + def __str__(self): return ''.join([str(entry) + "\n" for entry in self.entries]) @@ -180,6 +213,8 @@ class Entry: def __init__(self): self.useragents = [] self.rulelines = [] + self.delay = None + self.req_rate = None def __str__(self): ret = [] diff --git a/Misc/ACKS b/Misc/ACKS index 9678bc3..221d6e2 100644 --- a/Misc/ACKS +++ b/Misc/ACKS @@ -151,6 +151,7 @@ Finn Bock Paul Boddie Matthew Boedicker Robin Boerdijk +Nikolay Bogoychev David Bolen Wouter Bolsterlee Gawain Bolton diff --git a/Misc/NEWS b/Misc/NEWS index daa40fc..b543088 100644 --- a/Misc/NEWS +++ b/Misc/NEWS @@ -1,4 +1,4 @@ -+++++++++++ ++++++++++++ Python News +++++++++++ @@ -46,6 +46,9 @@ Core and Builtins Library ------- +- Issue #16099: RobotFileParser now supports Crawl-delay and Request-rate + extensions. Patch by Nikolay Bogoychev. + - Issue #25316: distutils raises OSError instead of DistutilsPlatformError when MSVC is not installed. -- cgit v0.12