Correct various errors:

- Use substring search, not re search for user-agent and paths.
- Consider * entry last. Unquote, then requote URLs.
- Treat empty Disallow as "allow everything".
Add test cases. Fixes #523041

Lib/robotparser.py

@@ -22,6 +22,7 @@ def _debug(msg):
 class RobotFileParser:
     def __init__(self, url=''):
         self.entries = []
+        self.default_entry = None
         self.disallow_all = 0
         self.allow_all = 0
         self.set_url(url)
@@ -72,7 +73,11 @@ class RobotFileParser:
                     entry = Entry()
                     state = 0
                 elif state==2:
-                    self.entries.append(entry)
+                    if "*" in entry.useragents:
+                        # the default entry is considered last
+                        self.default_entry = entry
+                    else:
+                        self.entries.append(entry)
                     entry = Entry()
                     state = 0
             # remove optional comment and strip line
@@ -85,7 +90,7 @@ class RobotFileParser:
             line = line.split(':', 1)
             if len(line) == 2:
                 line[0] = line[0].strip().lower()
-                line[1] = line[1].strip()
+                line[1] = urllib.unquote(line[1].strip())
                 if line[0] == "user-agent":
                     if state==2:
                         _debug("line %d: warning: you should insert a blank"
@@ -128,10 +133,13 @@ class RobotFileParser:
             return 1
         # search for given user agent matches
         # the first match counts
-        url = urllib.quote(urlparse.urlparse(url)[2]) or "/"
+        url = urllib.quote(urlparse.urlparse(urllib.unquote(url))[2]) or "/"
         for entry in self.entries:
             if entry.applies_to(useragent):
                 return entry.allowance(url)
+        # try the default entry last
+        if self.default_entry:
+            return self.default_entry.allowance(url)
         # agent not found ==> access granted
         return 1
 
@@ -147,11 +155,14 @@ class RuleLine:
     """A rule line is a single "Allow:" (allowance==1) or "Disallow:"
        (allowance==0) followed by a path."""
     def __init__(self, path, allowance):
+        if path == '' and not allowance:
+            # an empty value means allow all
+            allowance = 1
         self.path = urllib.quote(path)
         self.allowance = allowance
 
     def applies_to(self, filename):
-        return self.path=="*" or re.match(self.path, filename)
+        return self.path=="*" or filename.startswith(self.path)
 
     def __str__(self):
         return (self.allowance and "Allow" or "Disallow")+": "+self.path
@@ -180,8 +191,7 @@ class Entry:
                 # we have the catch-all agent
                 return 1
             agent = agent.lower()
-            # don't forget to re.escape
-            if re.search(re.escape(useragent), agent):
+            if useragent.find(agent) != -1:
                 return 1
         return 0
 

Lib/test/test_robotparser.py

@@ -0,0 +1,141 @@
+import unittest, StringIO, robotparser, test_support
+
+class RobotTestCase(unittest.TestCase):
+    def __init__(self, index, parser, url, good, agent):
+        unittest.TestCase.__init__(self)
+        if good:
+            self.str = "RobotTest(%d, good, %s)" % (index, url)
+        else:
+            self.str = "RobotTest(%d, bad, %s)" % (index, url)
+        self.parser = parser
+        self.url = url
+        self.good = good
+        self.agent = agent
+
+    def runTest(self):
+        if isinstance(self.url, tuple):
+            agent, url = self.url
+        else:
+            url = self.url
+            agent = self.agent
+        if self.good:
+            self.failUnless(self.parser.can_fetch(agent, url))
+        else:
+            self.failIf(self.parser.can_fetch(agent, url))
+
+    def __str__(self):
+        return self.str
+
+tests = unittest.TestSuite()
+
+def RobotTest(index, robots_txt, good_urls, bad_urls,
+              agent="test_robotparser"):
+    
+        lines = StringIO.StringIO(robots_txt).readlines()
+        parser = robotparser.RobotFileParser()
+        parser.parse(lines)
+        for url in good_urls:
+            tests.addTest(RobotTestCase(index, parser, url, 1, agent))
+        for url in bad_urls:
+            tests.addTest(RobotTestCase(index, parser, url, 0, agent))
+
+# Examples from http://www.robotstxt.org/wc/norobots.html (fetched 2002)
+
+# 1.
+doc = """
+User-agent: *
+Disallow: /cyberworld/map/ # This is an infinite virtual URL space
+Disallow: /tmp/ # these will soon disappear
+Disallow: /foo.html
+"""
+
+good = ['/','/test.html']
+bad = ['/cyberworld/map/index.html','/tmp/xxx','/foo.html']
+
+RobotTest(1, doc, good, bad)
+
+# 2.
+doc = """
+# robots.txt for http://www.example.com/
+
+User-agent: *
+Disallow: /cyberworld/map/ # This is an infinite virtual URL space
+
+# Cybermapper knows where to go.
+User-agent: cybermapper
+Disallow:
+
+"""
+
+good = ['/','/test.html',('cybermapper','/cyberworld/map/index.html')]
+bad = ['/cyberworld/map/index.html']
+
+RobotTest(2, doc, good, bad)
+
+# 3.
+doc = """
+# go away
+User-agent: *
+Disallow: /
+"""
+
+good = []
+bad = ['/cyberworld/map/index.html','/','/tmp/']
+
+RobotTest(3, doc, good, bad)
+
+# Examples from http://www.robotstxt.org/wc/norobots-rfc.html (fetched 2002)
+
+# 4.
+doc = """
+User-agent: figtree
+Disallow: /tmp
+Disallow: /a%3cd.html
+Disallow: /a%2fb.html
+Disallow: /%7ejoe/index.html
+"""
+
+good = [] # XFAIL '/a/b.html'
+bad = ['/tmp','/tmp.html','/tmp/a.html',
+       '/a%3cd.html','/a%3Cd.html','/a%2fb.html',
+       '/~joe/index.html'
+       ]
+
+RobotTest(4, doc, good, bad, 'figtree')
+RobotTest(5, doc, good, bad, 'FigTree Robot libwww-perl/5.04')
+
+# 6.
+doc = """
+User-agent: *
+Disallow: /tmp/
+Disallow: /a%3Cd.html
+Disallow: /a/b.html
+Disallow: /%7ejoe/index.html
+"""
+
+good = ['/tmp',] # XFAIL: '/a%2fb.html'
+bad = ['/tmp/','/tmp/a.html',
+       '/a%3cd.html','/a%3Cd.html',"/a/b.html",
+       '/%7Ejoe/index.html'] 
+
+RobotTest(6, doc, good, bad)
+
+# From bug report #523041
+
+# 7.
+doc = """
+User-Agent: *
+Disallow: /.
+"""
+
+good = ['/foo.html']
+bad = [] # Bug report says "/" should be denied, but that is not in the RFC
+
+RobotTest(7, doc, good, bad)
+
+def test_main():
+    test_support.run_suite(tests)
+
+if __name__=='__main__':
+    test_support.Verbose = 1
+    test_support.run_suite(tests)