adding PDF parser for bieber

tools/getBieber.py

@@ -0,0 +1,73 @@
+import sys
+import urllib
+from cStringIO import StringIO
+from pdfminer.pdfinterp import PDFResourceManager, process_pdf
+from pdfminer.pdfdevice import PDFDevice
+from pdfminer.converter import TextConverter
+import re
+
+
+TEMPLATE = """  {
+        "company": "Bieber",
+        "departureTime": "%s",
+        "departureLocation": "%s",
+        "arrivalTime": "%s",
+        "arrivalLocation": "%s"
+    },
+"""
+
+def pdf2txt(buff):
+    txt = StringIO()
+    rsrcmgr = PDFResourceManager(caching=True)
+    device = TextConverter(rsrcmgr, txt, codec='utf-8', laparams=None)
+    process_pdf(rsrcmgr, device, buff, set(), maxpages=0, password='', caching=True, check_extractable=True)
+    return txt.getvalue()
+
+def brittle_extractor(blob, from_city, to_city):
+    """ brittle_extractor -- A Brittle extractor, it's pretty becuase it depends almost entirely
+                             on the format of the PDF not changing, but hey it's better than typing
+                             all this shit by hand.
+        Arguments:
+        from_city -- name of depature city (e.g. Hellertown)
+        to_city -- name of destination city (e.g. New York City)
+    """
+    blob = blob.lower()
+    scheduleNumbers = re.findall("SCHEDULE NUMBER(([\d]{3})+)",txt)
+    scheduleNumbers = re.findall('(\d\d\d)', scheduleNumbers[0][0])
+    week = [x for x in scheduleNumbers if x[0]=='1']
+    weekend = [x for x in scheduleNumbers if x[0]=='2']
+    #print len(scheduleNumbers), "schedules"
+    #print len(week), 'week day'
+    #print len(weekend), 'week end'
+
+    groups = re.findall('([^0-9]+)((\d{3,4}[ap])+)',blob)
+    src = [] #3 6
+    dst = [] #4 5
+    for i in xrange(len(groups)):
+        if from_city.lower() in groups[i][0]:
+            src.append(i)
+        if to_city.lower() in groups[i][0]:
+            dst.append(i)
+
+    if src[0] < dst[0]:
+        a,b = groups[src[0]], groups[dst[0]]
+    else:
+        a,b = groups[src[1]], groups[dst[1]]
+
+    a,b = re.findall('(\d{3,4}[ap])',a[1]), re.findall('(\d{3,4}[ap])',b[1])
+    n = len(week)
+    for lv,ar in zip(a,b)[:n]:
+        print TEMPLATE%(lv,from_city,ar,to_city)
+
+
+
+if __name__ == '__main__':
+    if len(sys.argv) != 2:
+        print "usage: python getBieber.py http://url/to/schedule.pdf > bieber.json"
+        print "       eg, http://www.biebertourways.com/active/2013-2-26/NYC%203-2013.pdf"
+    else:
+        pdf = StringIO(urllib.urlopen(sys.argv[1]).read())
+        txt = pdf2txt(pdf)
+        txt = txt.decode('ascii','ignore')
+        s = brittle_extractor(txt, 'Hellertown', 'PABT')
+        s = brittle_extractor(txt, 'PABT', 'Hellertown')