Mercurial > ~darius > hgwebdir.cgi > scrape-gm
annotate scrape-gm.py @ 13:e8550290e512
Update parser for new format.
| author | darius@Inchoate |
|---|---|
| date | Sun, 14 Dec 2008 18:48:15 +1030 |
| parents | ae9e833e4447 |
| children |
| rev | line source |
|---|---|
| 1 | 1 #!/usr/bin/env python |
| 2 | |
| 3 ############################################################################ | |
| 4 # Screen scraper for game-monitor.com | |
| 5 # | |
| 6 # Prints out matched player names agreated by server | |
| 7 # | |
| 8 ############################################################################ | |
| 9 # | |
|
11
22a51e8c0a69
Update copyright date. Remove CVS ID.
darius@inchoate.localdomain
parents:
10
diff
changeset
|
10 # Copyright (C) 2008 Daniel O'Connor. All rights reserved. |
| 1 | 11 # |
| 12 # Redistribution and use in source and binary forms, with or without | |
| 13 # modification, are permitted provided that the following conditions | |
| 14 # are met: | |
| 15 # 1. Redistributions of source code must retain the above copyright | |
| 16 # notice, this list of conditions and the following disclaimer. | |
| 17 # 2. Redistributions in binary form must reproduce the above copyright | |
| 18 # notice, this list of conditions and the following disclaimer in the | |
| 19 # documentation and/or other materials provided with the distribution. | |
| 20 # | |
| 21 # THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS ``AS IS'' AND | |
| 22 # ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE | |
| 23 # IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE | |
| 24 # ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE | |
| 25 # FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL | |
| 26 # DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS | |
| 27 # OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) | |
| 28 # HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT | |
| 29 # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY | |
| 30 # OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF | |
| 31 # SUCH DAMAGE. | |
| 32 # | |
| 33 ############################################################################ | |
| 34 | |
| 35 import re, time, datetime, urllib, sys, BeautifulSoup | |
| 36 | |
| 13 | 37 debug = False |
| 38 | |
| 1 | 39 class Server: |
| 40 alltags = re.compile('<[^>]*>') | |
| 41 vwhttags = re.compile('<(br|hr)>') | |
| 42 hwhttags = re.compile('\ ') | |
| 12 | 43 typetag = re.compile('<td><a href="/GameSearch/([^/]+)/.*</td>') |
| 44 | |
| 1 | 45 def __init__(self, description = "", ip = "", port = 0, mapname = "", |
| 46 updateage = 0, numplayers = 0, maxplayers = 0, players = []): | |
| 47 self.description = description | |
| 48 self.ip = ip | |
| 49 self.port = port | |
| 50 self.mapname = mapname | |
| 51 self.updateage = int(updateage) | |
| 52 self.players = [] | |
| 53 self.numplayers = numplayers | |
| 54 self.maxplayers = maxplayers | |
| 55 | |
| 56 def __init__(self, pcols, scols): | |
| 13 | 57 # pcols[1] = Player name |
| 58 # pcols[2] = Server description | |
| 1 | 59 # scols[0] = Players in server / max players |
| 13 | 60 # scols[1] = Server IP & port |
| 61 # scols[2] = Map name | |
| 62 # scols[3] = Game type | |
| 63 # scols[8] = Update age | |
| 64 if debug: | |
| 65 print "pcols = " + str(pcols) | |
| 66 print "scols = " + str(scols) | |
| 67 | |
| 68 self.pcountre = re.compile("([0-9]+)/([0-9]+)") | |
| 69 self.ipportre = re.compile("([0-9]+\.[0-9]+\.[0-9]+\.[0-9]+):([0-9]+)") | |
| 70 self.sdesc = re.compile(" +[0-9]+\. +(.*)") | |
| 71 | |
| 72 m = self.sdesc.match(pcols[2]) | |
| 73 if (m == None): | |
| 74 raise SyntaxError | |
| 75 self.description = m.group(1) | |
| 76 | |
| 77 m = self.ipportre.match(scols[1]) | |
| 78 if (m == None): | |
| 79 raise SyntaxError | |
| 80 | |
| 81 self.ip = m.group(1) | |
| 82 self.port = int(m.group(2)) | |
| 83 self.gametype = scols[3] | |
| 84 self.mapname = scols[2] | |
| 85 self.updateage = scols[8] | |
| 86 m = self.pcountre.match(scols[0]) | |
| 1 | 87 if (m == None): |
| 88 raise SyntaxError | |
| 89 | |
| 90 self.numplayers = int(m.group(1)) | |
| 91 self.maxplayers = int(m.group(2)) | |
| 92 self.players = [] | |
| 93 | |
| 94 def __str__(self): | |
| 95 plist = "" | |
| 96 for p in self.players: | |
| 97 plist = plist + " " + str(p) | |
| 98 | |
| 12 | 99 return "%s: %s (%s:%d) | Map: %s | Players: %d/%d : %s (%s old)" % \ |
| 100 (self.gametype, self.description, self.ip, self.port, self.mapname, | |
|
7
825302e32c35
Print out the server IP & port tuple as well as the name.
darius@inchoate.localdomain
parents:
6
diff
changeset
|
101 self.numplayers, self.maxplayers, plist, |
|
825302e32c35
Print out the server IP & port tuple as well as the name.
darius@inchoate.localdomain
parents:
6
diff
changeset
|
102 self.updateage) |
| 1 | 103 |
| 104 def GetTuple(scols): | |
| 105 return str(scols[2]) + ":" + str(scols[3]) | |
| 106 GetTuple = staticmethod(GetTuple) | |
| 107 | |
| 108 def FixTags(s): | |
| 12 | 109 # Mangle game type |
| 110 t = Server.typetag.match(s) | |
| 111 if t != None: | |
| 112 s = t.group(1) | |
| 1 | 113 s = re.sub(Server.vwhttags, '\n', s) |
| 114 s = re.sub(Server.hwhttags, '', s) | |
| 115 s = str(BeautifulSoup.BeautifulStoneSoup( \ | |
| 116 s, convertEntities = BeautifulSoup.BeautifulStoneSoup.XML_ENTITIES)) | |
| 117 s = re.sub(Server.alltags, '', s) | |
| 118 return(s) | |
| 119 FixTags = staticmethod(FixTags) | |
| 120 | |
| 121 def Scrape(handle): | |
| 122 s = BeautifulSoup.BeautifulSoup(handle) | |
| 123 | |
| 10 | 124 playertbl = s.find("table", "results") |
| 1 | 125 if (playertbl == None): |
| 13 | 126 if True: |
| 127 print "Unable to find results" | |
| 1 | 128 return None |
| 129 | |
| 12 | 130 servertbl = playertbl.findNext("table") |
| 1 | 131 |
| 132 playerrows = playertbl.findAll("tr") | |
| 133 serverrows = servertbl.findAll("tr") | |
| 134 | |
| 135 if (len(playerrows) != len(serverrows)): | |
| 136 print "Internal error 41223" | |
| 137 return | |
| 138 | |
| 139 servers = {} | |
| 140 for i in range(len(playerrows[1:])): | |
| 141 pcols = playerrows[i].findAll('td') | |
| 142 scols = serverrows[i].findAll('td') | |
| 13 | 143 if (len(pcols) != 3): |
| 144 if debug: | |
| 145 print "pcols has length %d, expected 3" % len(pcols) | |
| 1 | 146 continue |
| 147 | |
| 148 pcols = map(lambda c : Server.FixTags(str(c)), pcols) | |
| 149 scols = map(lambda c : Server.FixTags(str(c)), scols) | |
| 150 | |
| 151 stuple = Server.GetTuple(scols) | |
| 152 | |
| 153 if (stuple not in servers): | |
| 154 s = Server(pcols, scols) | |
| 155 servers[stuple] = s | |
| 156 | |
| 13 | 157 servers[stuple].addplayer(pcols[1]) |
| 1 | 158 |
| 159 return servers | |
| 160 Scrape = staticmethod(Scrape) | |
| 161 | |
| 162 def addplayer(self, pname): | |
| 163 self.players.append(pname) | |
| 164 | |
| 165 | |
| 13 | 166 if True: |
| 1 | 167 maxhits = 10 |
| 168 if (len(sys.argv) < 2): | |
| 169 print "Bad usage" | |
| 170 print sys.argv[0] + "search_string" | |
| 171 sys.exit(1) | |
| 172 | |
| 173 try: | |
| 174 #f = open("gm.html") | |
| 8 | 175 f = urllib.urlopen("http://www.game-monitor.com/search.php?location=AU&search=" + urllib.quote(sys.argv[1]) + "&type=player&location=AU") |
| 1 | 176 except IOError, e: |
| 177 print "Unable to fetch page - " + str(e) | |
| 178 sys.exit(0) | |
| 179 | |
| 180 servers = Server.Scrape(f) | |
| 181 del f | |
| 6 | 182 if (servers == None): |
| 183 print "No results available, please check manually" | |
| 184 elif (len(servers) == 0): | |
| 1 | 185 print "No players found" |
| 186 else: | |
| 12 | 187 tmp = [] |
| 188 for i in servers: | |
| 189 tmp.append(servers[i]) | |
| 190 tmp.sort() | |
| 1 | 191 i = 0 |
| 12 | 192 for s in tmp: |
| 1 | 193 i = i + 1 |
| 12 | 194 print s |
| 1 | 195 if (i >= maxhits): |
| 196 print "*** Stopping after " + str(maxhits) + " hits" | |
| 197 break |