Mercurial > ~darius > hgwebdir.cgi > scrape-gm
annotate scrape-gm.py @ 10:0e18c714b69d
Update for new page layout.
Luckily no substative changes! :)
| author | darius@inchoate.localdomain |
|---|---|
| date | Fri, 29 Feb 2008 19:19:14 +1030 |
| parents | fbaf34d7bab7 |
| children | 22a51e8c0a69 |
| 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 # | |
| 6 | 8 # $Id: scrape-gm.py,v 1.3 2007/11/18 08:54:07 darius Exp $ |
| 1 | 9 ############################################################################ |
| 10 # | |
| 11 # Copyright (C) 2007 Daniel O'Connor. All rights reserved. | |
| 12 # | |
| 13 # Redistribution and use in source and binary forms, with or without | |
| 14 # modification, are permitted provided that the following conditions | |
| 15 # are met: | |
| 16 # 1. Redistributions of source code must retain the above copyright | |
| 17 # notice, this list of conditions and the following disclaimer. | |
| 18 # 2. Redistributions in binary form must reproduce the above copyright | |
| 19 # notice, this list of conditions and the following disclaimer in the | |
| 20 # documentation and/or other materials provided with the distribution. | |
| 21 # | |
| 22 # THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS ``AS IS'' AND | |
| 23 # ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE | |
| 24 # IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE | |
| 25 # ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE | |
| 26 # FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL | |
| 27 # DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS | |
| 28 # OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) | |
| 29 # HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT | |
| 30 # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY | |
| 31 # OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF | |
| 32 # SUCH DAMAGE. | |
| 33 # | |
| 34 ############################################################################ | |
| 35 | |
| 36 import re, time, datetime, urllib, sys, BeautifulSoup | |
| 37 | |
| 38 class Server: | |
| 39 alltags = re.compile('<[^>]*>') | |
| 40 vwhttags = re.compile('<(br|hr)>') | |
| 41 hwhttags = re.compile('\ ') | |
| 42 | |
| 43 def __init__(self, description = "", ip = "", port = 0, mapname = "", | |
| 44 updateage = 0, numplayers = 0, maxplayers = 0, players = []): | |
| 45 self.description = description | |
| 46 self.ip = ip | |
| 47 self.port = port | |
| 48 self.mapname = mapname | |
| 49 self.updateage = int(updateage) | |
| 50 self.players = [] | |
| 51 self.numplayers = numplayers | |
| 52 self.maxplayers = maxplayers | |
| 53 | |
| 54 def __init__(self, pcols, scols): | |
| 55 # pcols[2] = Player name | |
| 56 # pcols[3] = Server description | |
| 57 # scols[0] = Players in server / max players | |
| 58 # scols[2] = Server IP | |
| 59 # scols[3] = Server port | |
| 60 # scols[4] = Map name | |
| 61 # scols[10] = Update age | |
| 62 self.tuplere = re.compile("\[?([0-9]+)/([0-9]+)\]?") | |
| 63 self.description = pcols[3] | |
| 64 self.ip = scols[2] | |
| 65 self.port = int(scols[3]) | |
| 66 self.mapname = scols[4] | |
| 67 self.updateage = scols[10] | |
| 68 m = self.tuplere.match(scols[0]) | |
| 69 if (m == None): | |
| 70 raise SyntaxError | |
| 71 | |
| 72 self.numplayers = int(m.group(1)) | |
| 73 self.maxplayers = int(m.group(2)) | |
| 74 self.players = [] | |
| 75 | |
| 76 def __str__(self): | |
| 77 plist = "" | |
| 78 for p in self.players: | |
| 79 plist = plist + " " + str(p) | |
| 80 | |
|
7
825302e32c35
Print out the server IP & port tuple as well as the name.
darius@inchoate.localdomain
parents:
6
diff
changeset
|
81 return "%s (%s:%d) | Map: %s | Players: %d/%d : %s (%s old)" % \ |
|
825302e32c35
Print out the server IP & port tuple as well as the name.
darius@inchoate.localdomain
parents:
6
diff
changeset
|
82 (self.description, self.ip, self.port, self.mapname, |
|
825302e32c35
Print out the server IP & port tuple as well as the name.
darius@inchoate.localdomain
parents:
6
diff
changeset
|
83 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
|
84 self.updateage) |
| 1 | 85 |
| 86 def GetTuple(scols): | |
| 87 return str(scols[2]) + ":" + str(scols[3]) | |
| 88 GetTuple = staticmethod(GetTuple) | |
| 89 | |
| 90 def FixTags(s): | |
| 91 s = re.sub(Server.vwhttags, '\n', s) | |
| 92 s = re.sub(Server.hwhttags, '', s) | |
| 93 s = str(BeautifulSoup.BeautifulStoneSoup( \ | |
| 94 s, convertEntities = BeautifulSoup.BeautifulStoneSoup.XML_ENTITIES)) | |
| 95 s = re.sub(Server.alltags, '', s) | |
| 96 return(s) | |
| 97 FixTags = staticmethod(FixTags) | |
| 98 | |
| 99 def Scrape(handle): | |
| 100 s = BeautifulSoup.BeautifulSoup(handle) | |
| 101 | |
| 10 | 102 playertbl = s.find("table", "results") |
| 1 | 103 if (playertbl == None): |
| 6 | 104 #print "Unable to find results" |
| 1 | 105 return None |
| 106 | |
| 10 | 107 servertbl = playertbl.findNext("table", "results") |
| 1 | 108 |
| 109 playerrows = playertbl.findAll("tr") | |
| 110 serverrows = servertbl.findAll("tr") | |
| 111 | |
| 112 if (len(playerrows) != len(serverrows)): | |
| 113 print "Internal error 41223" | |
| 114 return | |
| 115 | |
| 116 servers = {} | |
| 117 for i in range(len(playerrows[1:])): | |
| 118 pcols = playerrows[i].findAll('td') | |
| 119 scols = serverrows[i].findAll('td') | |
| 120 if (len(pcols) != 4): | |
| 121 continue | |
| 122 | |
| 123 pcols = map(lambda c : Server.FixTags(str(c)), pcols) | |
| 124 scols = map(lambda c : Server.FixTags(str(c)), scols) | |
| 125 | |
| 126 stuple = Server.GetTuple(scols) | |
| 127 | |
| 128 if (stuple not in servers): | |
| 129 s = Server(pcols, scols) | |
| 130 servers[stuple] = s | |
| 131 | |
| 132 servers[stuple].addplayer(pcols[2]) | |
| 133 | |
| 134 return servers | |
| 135 Scrape = staticmethod(Scrape) | |
| 136 | |
| 137 def addplayer(self, pname): | |
| 138 self.players.append(pname) | |
| 139 | |
| 140 | |
| 141 if (1): | |
| 142 maxhits = 10 | |
| 143 if (len(sys.argv) < 2): | |
| 144 print "Bad usage" | |
| 145 print sys.argv[0] + "search_string" | |
| 146 sys.exit(1) | |
| 147 | |
| 148 try: | |
| 149 #f = open("gm.html") | |
| 8 | 150 f = urllib.urlopen("http://www.game-monitor.com/search.php?location=AU&search=" + urllib.quote(sys.argv[1]) + "&type=player&location=AU") |
| 1 | 151 except IOError, e: |
| 152 print "Unable to fetch page - " + str(e) | |
| 153 sys.exit(0) | |
| 154 | |
| 155 servers = Server.Scrape(f) | |
| 156 del f | |
| 6 | 157 if (servers == None): |
| 158 print "No results available, please check manually" | |
| 159 elif (len(servers) == 0): | |
| 1 | 160 print "No players found" |
| 161 else: | |
| 162 i = 0 | |
| 163 for s in servers: | |
| 164 i = i + 1 | |
| 165 print servers[s] | |
| 166 if (i >= maxhits): | |
| 167 print "*** Stopping after " + str(maxhits) + " hits" | |
| 168 break |