Conception connection, inspired by http://www.smbc-comics.com/index.php?db=comics&id=2922#comic

conception.py

#!/usr/bin/python
#
# Usage:
#
# $ ./conception.py 1969-12-25
# searching for most recent event before 1969-04-03
# <li><a href="/wiki/April_1" title="April 1">April 1</a>
# &#226;&#8364;&#8220; The <a href="/wiki/Hawker_Siddeley_Harrier" title="Hawker Siddeley Harrier">Hawker Siddeley Harrier</a> enters
# service with the <a href="/wiki/Royal_Air_Force" title="Royal Air Force">Royal Air Force</a>.</li>
#
# If python complains about ImportError, install tidy/lxml python libs
#
# $ aptitude install python-tidylib python-lxml

from bisect import bisect
from lxml import etree
import os.path
from functools import partial
import sys
import tidy
import time
import urllib2

# TODO: better estimate from your birth weight and location
AV_PREG_LENGTH=266*24*3600

TIDY_OPTS = dict(output_xhtml=1,
                 add_xml_decl=1,
                 tidy_mark=0)

def get_content(url):
    # http://stackoverflow.com/questions/120061/fetch-a-wikipedia-article-with-python
    opener = urllib2.build_opener()
    opener.addheaders = [('User-agent', 'Mozilla/5.0')]
    infile = opener.open(url)
    text = infile.read()
    return etree.HTML(str(tidy.parseString(text, **TIDY_OPTS)))

def md_to_key(d):
    return "%02d.%02d" % (d.tm_mon, d.tm_mday)

def parse_li(parent, b_i, e_i, l):
    try:
        # ensure node is inside "events" section
        p = l
        while p.getparent() != parent:
            p = p.getparent()
        assert p.getparent() == parent
        i = parent.index(p)
        if not (b_i < i < e_i): return None
        n = l.find(".//a[@href]")
        s = os.path.basename(n.get("href"))
        d = time.strptime(s, "%B_%d")
        return (md_to_key(d), l)
    except Exception, e:
        return None

def get_events(y):
    dom = get_content("http://en.wikipedia.org/wiki/%s" % y)
    # find delimiters of "events" section
    eventsh2 = dom.find(".//span[@id='Events']/..")
    nexth2 = eventsh2
    while True:
        nexth2 = nexth2.getnext()
        if nexth2.tag == 'h2': break
    parent = eventsh2.getparent()
    assert parent == nexth2.getparent()
    b_i = parent.index(eventsh2)
    e_i = parent.index(nexth2)
    return sorted(filter(None, map(partial(parse_li, parent, b_i, e_i), dom.findall(".//li"))))

if __name__ == '__main__':
    d = time.localtime(time.mktime(time.strptime(sys.argv[1], "%Y-%m-%d")) - AV_PREG_LENGTH)
    print time.strftime("searching for most recent event before %Y-%m-%d", d)
    dd = get_events(d.tm_year)
    i = bisect(dd, (md_to_key(d), None))
    if i > 0:
        ev = dd[i-1]
    else:
        # need to check previous year
        dd = get_events(d.tm_year - 1)
        ev = dd[-1]
    print etree.tostring(ev[1])