chiark / gitweb /
WIP routesearch; before one query per pair
[ypp-sc-tools.db-test.git] / yarrg / yppedia-ocean-scraper
index 5f9c5f0..68efa38 100755 (executable)
@@ -37,11 +37,42 @@ signal.signal(signal.SIGINT, signal.SIG_DFL)
 import sys
 import os
 import urllib
-import urllib2
 import re as regexp
+import subprocess
 from optparse import OptionParser
 from BeautifulSoup import BeautifulSoup
 
+
+# For fuck's sake!
+import codecs
+import locale
+def fix_stdout():
+    sys.stdout = codecs.EncodedFile(sys.stdout, locale.getpreferredencoding())
+    def null_decode(input, errors='strict'):
+        return input, len(input)
+    sys.stdout.decode = null_decode
+# From
+#  http://ewx.livejournal.com/457086.html?thread=3016574
+#  http://ewx.livejournal.com/457086.html?thread=3016574
+# lightly modified.
+# See also Debian #415968.
+fix_stdout()
+
+
+# User agent:
+class YarrgURLopener(urllib.FancyURLopener):
+       base_version= urllib.URLopener().version
+       proc= subprocess.Popen(
+               ["./database-info-fetch", "useragentstringmap",
+                base_version, "manual islands/topology fetch"],
+               shell=False,
+               stderr=None,
+               stdout=subprocess.PIPE,
+               )
+       version = proc.communicate()[0].rstrip('\n');
+       assert(proc.returncode is not None and proc.returncode == 0)
+urllib._urlopener = YarrgURLopener()
+
 ocean = None
 soup = None
 opts = None
@@ -60,10 +91,9 @@ def fetch():
        url = ('http://yppedia.puzzlepirates.com/' +
                        (url_base % urllib.quote(ocean,'')))
        debug('fetching',url)
-       dataf = urllib2.urlopen(url)
+       dataf = urllib.urlopen(url)
        debug('fetched',dataf)
-       soup = BeautifulSoup(dataf,
-               convertEntities=BeautifulSoup.HTML_ENTITIES)
+       soup = BeautifulSoup(dataf)
 
 
 title_arch_re = regexp.compile('(\\S.*\\S) Archipelago \\((\\S+)\\)$')
@@ -86,13 +116,17 @@ def title_arch_ok(t):
 def parse_chart():
        ta = soup.find('textarea')
        debug('ta',ta)
-       rc = ta.renderContents()
-       debug('rc',rc)
-       return rc
+       s = ta.string
+       debug('s',s)
+       s = regexp.sub(r'\&lt\;', '<', s)
+       s = regexp.sub(r'\&gt\;', '>', s)
+       s = regexp.sub(r'\&quot\;', '"', s)
+       s = regexp.sub(r'\&amp\;', '&', s)
+       debug('s',s)
+       return s
 
 def parse_ocean():
-       firstarch = soup.find('a', attrs = {'title': title_arch_ok})
-       debug('fa',firstarch)
+       content = soup.find('div', attrs = {'id': 'content'})
 
        def findall_title_arch_ok(t):
                return t.findAll('a', attrs = {'title': title_arch_ok})
@@ -101,7 +135,7 @@ def parse_ocean():
                if u.name != 'table': return False
                return len(findall_title_arch_ok(u)) > 1
 
-       archestable = firstarch.findParent('table', attrs={'border':'1'})
+       archestable = content.findChild('table', attrs={'border':'1'})
        debug('at',archestable)
 
        archsoups = []