Coursera课程笔记----P4E.Capstone----Week 4&5

Spidering and Modeling Email Data（week4&5）

Mailing List - Gmane

• Crawl the archive of a mailing list
• Do some analysis / cleanup
• Visualize the data as word cloud and lines

code segment

gmane.py

import sqlite3
import time
import ssl
import urllib.request, urllib.parse, urllib.error
from urllib.parse import urljoin
from urllib.parse import urlparse
import re
from datetime import datetime, timedelta

# Not all systems have this so conditionally define parser
try:
    import dateutil.parser as parser
except:
    pass

def parsemaildate(md) :
    # See if we have dateutil
    try:
        pdate = parser.parse(tdate)
        test_at = pdate.isoformat()
        return test_at
    except:
        pass

    # Non-dateutil version - we try our best

    pieces = md.split()
    notz = " ".join(pieces[:4]).strip()

    # Try a bunch of format variations - strptime() is *lame*
    dnotz = None
    for form in [ '%d %b %Y %H:%M:%S', '%d %b %Y %H:%M:%S',
        '%d %b %Y %H:%M', '%d %b %Y %H:%M', '%d %b %y %H:%M:%S',
        '%d %b %y %H:%M:%S', '%d %b %y %H:%M', '%d %b %y %H:%M' ] :
        try:
            dnotz = datetime.strptime(notz, form)
            break
        except:
            continue

    if dnotz is None :
        # print 'Bad Date:',md
        return None

    iso = dnotz.isoformat()

    tz = "+0000"
    try:
        tz = pieces[4]
        ival = int(tz) # Only want numeric timezone values
        if tz == '-0000' : tz = '+0000'
        tzh = tz[:3]
        tzm = tz[3:]
        tz = tzh+":"+tzm
    except:
        pass

    return iso+tz

# Ignore SSL certificate errors
ctx = ssl.create_default_context()
ctx.check_hostname = False
ctx.verify_mode = ssl.CERT_NONE

conn = sqlite3.connect('content.sqlite')
cur = conn.cursor()

baseurl = "http://mbox.dr-chuck.net/sakai.devel/"

cur.execute('''CREATE TABLE IF NOT EXISTS Messages
    (id INTEGER UNIQUE, email TEXT, sent_at TEXT,
     subject TEXT, headers TEXT, body TEXT)''')

# Pick up where we left off
start = None
cur.execute('SELECT max(id) FROM Messages' )
try:
    row = cur.fetchone()
    if row is None :
        start = 0
    else:
        start = row[0]
except:
    start = 0

if start is None : start = 0

many = 0
count = 0
fail = 0
while True:
    if ( many < 1 ) :
        sval = input('How many messages:')
        if ( len(sval) < 1 ) : break
        many = int(sval)

    start = start + 1
    cur.execute('SELECT id FROM Messages WHERE id=?', (start,) )
    try:
        row = cur.fetchone()
        if row is not None : continue
    except:
        row = None

    many = many - 1
    url = baseurl + str(start) + '/' + str(start + 1)

    text = "None"
    try:
        # Open with a timeout of 30 seconds
        document = urllib.request.urlopen(url, None, 30, context=ctx)
        text = document.read().decode()
        if document.getcode() != 200 :
            print("Error code=",document.getcode(), url)
            break
    except KeyboardInterrupt:
        print('')
        print('Program interrupted by user...')
        break
    except Exception as e:
        print("Unable to retrieve or parse page",url)
        print("Error",e)
        fail = fail + 1
        if fail > 5 : break
        continue

    print(url,len(text))
    count = count + 1

    if not text.startswith("From "):
        print(text)
        print("Did not find From ")
        fail = fail + 1
        if fail > 5 : break
        continue

    pos = text.find("\n\n")
    if pos > 0 :
        hdr = text[:pos]
        body = text[pos+2:]
    else:
        print(text)
        print("Could not find break between headers and body")
        fail = fail + 1
        if fail > 5 : break
        continue

    email = None
    x = re.findall('\nFrom: .* <(\S+@\S+)>\n', hdr)
    if len(x) == 1 :
        email = x[0];
        email = email.strip().lower()
        email = email.replace("<","")
    else:
        x = re.findall('\nFrom: (\S+@\S+)\n', hdr)
        if len(x) == 1 :
            email = x[0];
            email = email.strip().lower()
            email = email.replace("<","")

    date = None
    y = re.findall('\Date: .*, (.*)\n', hdr)
    if len(y) == 1 :
        tdate = y[0]
        tdate = tdate[:26]
        try:
            sent_at = parsemaildate(tdate)
        except:
            print(text)
            print("Parse fail",tdate)
            fail = fail + 1
            if fail > 5 : break
            continue

    subject = None
    z = re.findall('\Subject: (.*)\n', hdr)
    if len(z) == 1 : subject = z[0].strip().lower();

    # Reset the fail counter
    fail = 0
    print("   ",email,sent_at,subject)
    cur.execute('''INSERT OR IGNORE INTO Messages (id, email, sent_at, subject, headers, body)
        VALUES ( ?, ?, ?, ?, ?, ? )''', ( start, email, sent_at, subject, hdr, body))
    if count % 50 == 0 : conn.commit()
    if count % 100 == 0 : time.sleep(1)

conn.commit()
cur.close()

gmodel.py

import sqlite3
import time
import re
import zlib
from datetime import datetime, timedelta

# Not all systems have this
try:
    import dateutil.parser as parser
except:
    pass

dnsmapping = dict()
mapping = dict()

def fixsender(sender,allsenders=None) :
    global dnsmapping
    global mapping
    if sender is None : return None
    sender = sender.strip().lower()
    sender = sender.replace('<','').replace('>','')

    # Check if we have a hacked gmane.org from address
    if allsenders is not None and sender.endswith('gmane.org') :
        pieces = sender.split('-')
        realsender = None
        for s in allsenders:
            if s.startswith(pieces[0]) :
                realsender = sender
                sender = s
                # print(realsender, sender)
                break
        if realsender is None :
            for s in mapping:
                if s.startswith(pieces[0]) :
                    realsender = sender
                    sender = mapping[s]
                    # print(realsender, sender)
                    break
        if realsender is None : sender = pieces[0]

    mpieces = sender.split("@")
    if len(mpieces) != 2 : return sender
    dns = mpieces[1]
    x = dns
    pieces = dns.split(".")
    if dns.endswith(".edu") or dns.endswith(".com") or dns.endswith(".org") or dns.endswith(".net") :
        dns = ".".join(pieces[-2:])
    else:
        dns = ".".join(pieces[-3:])
    # if dns != x : print(x,dns)
    # if dns != dnsmapping.get(dns,dns) : print(dns,dnsmapping.get(dns,dns))
    dns = dnsmapping.get(dns,dns)
    return mpieces[0] + '@' + dns

def parsemaildate(md) :
    # See if we have dateutil
    try:
        pdate = parser.parse(md)
        test_at = pdate.isoformat()
        return test_at
    except:
        pass

    # Non-dateutil version - we try our best

    pieces = md.split()
    notz = " ".join(pieces[:4]).strip()

    # Try a bunch of format variations - strptime() is *lame*
    dnotz = None
    for form in [ '%d %b %Y %H:%M:%S', '%d %b %Y %H:%M:%S',
        '%d %b %Y %H:%M', '%d %b %Y %H:%M', '%d %b %y %H:%M:%S',
        '%d %b %y %H:%M:%S', '%d %b %y %H:%M', '%d %b %y %H:%M' ] :
        try:
            dnotz = datetime.strptime(notz, form)
            break
        except:
            continue

    if dnotz is None :
        # print('Bad Date:',md)
        return None

    iso = dnotz.isoformat()

    tz = "+0000"
    try:
        tz = pieces[4]
        ival = int(tz) # Only want numeric timezone values
        if tz == '-0000' : tz = '+0000'
        tzh = tz[:3]
        tzm = tz[3:]
        tz = tzh+":"+tzm
    except:
        pass

    return iso+tz

# Parse out the info...
def parseheader(hdr, allsenders=None):
    if hdr is None or len(hdr) < 1 : return None
    sender = None
    x = re.findall('\nFrom: .* <(\S+@\S+)>\n', hdr)
    if len(x) >= 1 :
        sender = x[0]
    else:
        x = re.findall('\nFrom: (\S+@\S+)\n', hdr)
        if len(x) >= 1 :
            sender = x[0]

    # normalize the domain name of Email addresses
    sender = fixsender(sender, allsenders)

    date = None
    y = re.findall('\nDate: .*, (.*)\n', hdr)
    sent_at = None
    if len(y) >= 1 :
        tdate = y[0]
        tdate = tdate[:26]
        try:
            sent_at = parsemaildate(tdate)
        except Exception as e:
            # print('Date ignored ',tdate, e)
            return None

    subject = None
    z = re.findall('\nSubject: (.*)\n', hdr)
    if len(z) >= 1 : subject = z[0].strip().lower()

    guid = None
    z = re.findall('\nMessage-ID: (.*)\n', hdr)
    if len(z) >= 1 : guid = z[0].strip().lower()

    if sender is None or sent_at is None or subject is None or guid is None :
        return None
    return (guid, sender, subject, sent_at)

conn = sqlite3.connect('index.sqlite')
cur = conn.cursor()

cur.execute('''DROP TABLE IF EXISTS Messages ''')
cur.execute('''DROP TABLE IF EXISTS Senders ''')
cur.execute('''DROP TABLE IF EXISTS Subjects ''')
cur.execute('''DROP TABLE IF EXISTS Replies ''')

cur.execute('''CREATE TABLE IF NOT EXISTS Messages
    (id INTEGER PRIMARY KEY, guid TEXT UNIQUE, sent_at INTEGER,
     sender_id INTEGER, subject_id INTEGER,
     headers BLOB, body BLOB)''')
cur.execute('''CREATE TABLE IF NOT EXISTS Senders
    (id INTEGER PRIMARY KEY, sender TEXT UNIQUE)''')
cur.execute('''CREATE TABLE IF NOT EXISTS Subjects
    (id INTEGER PRIMARY KEY, subject TEXT UNIQUE)''')
cur.execute('''CREATE TABLE IF NOT EXISTS Replies
    (from_id INTEGER, to_id INTEGER)''')

conn_1 = sqlite3.connect('mapping.sqlite')
cur_1 = conn_1.cursor()

cur_1.execute('''SELECT old,new FROM DNSMapping''')
for message_row in cur_1 :
    dnsmapping[message_row[0].strip().lower()] = message_row[1].strip().lower()

mapping = dict()
cur_1.execute('''SELECT old,new FROM Mapping''')
for message_row in cur_1 :
    old = fixsender(message_row[0])
    new = fixsender(message_row[1])
    mapping[old] = fixsender(new)

# Done with mapping.sqlite
conn_1.close()

# Open the main content (Read only)
conn_1 = sqlite3.connect('file:content.sqlite?mode=ro', uri=True)
cur_1 = conn_1.cursor()

allsenders = list()
cur_1.execute('''SELECT email FROM Messages''')
for message_row in cur_1 :
    sender = fixsender(message_row[0])
    if sender is None : continue
    if 'gmane.org' in sender : continue
    if sender in allsenders: continue
    allsenders.append(sender)

print("Loaded allsenders",len(allsenders),"and mapping",len(mapping),"dns mapping",len(dnsmapping))

cur_1.execute('''SELECT headers, body, sent_at
    FROM Messages ORDER BY sent_at''')

senders = dict()
subjects = dict()
guids = dict()

count = 0

for message_row in cur_1 :
    hdr = message_row[0]
    parsed = parseheader(hdr, allsenders)
    if parsed is None: continue
    (guid, sender, subject, sent_at) = parsed

    # Apply the sender mapping
    sender = mapping.get(sender,sender)

    count = count + 1
    if count % 250 == 1 : print(count,sent_at, sender)
    # print(guid, sender, subject, sent_at)

    if 'gmane.org' in sender:
        print("Error in sender ===", sender)

    sender_id = senders.get(sender,None)
    subject_id = subjects.get(subject,None)
    guid_id = guids.get(guid,None)

    if sender_id is None :
        cur.execute('INSERT OR IGNORE INTO Senders (sender) VALUES ( ? )', ( sender, ) )
        conn.commit()
        cur.execute('SELECT id FROM Senders WHERE sender=? LIMIT 1', ( sender, ))
        try:
            row = cur.fetchone()
            sender_id = row[0]
            senders[sender] = sender_id
        except:
            print('Could not retrieve sender id',sender)
            break
    if subject_id is None :
        cur.execute('INSERT OR IGNORE INTO Subjects (subject) VALUES ( ? )', ( subject, ) )
        conn.commit()
        cur.execute('SELECT id FROM Subjects WHERE subject=? LIMIT 1', ( subject, ))
        try:
            row = cur.fetchone()
            subject_id = row[0]
            subjects[subject] = subject_id
        except:
            print('Could not retrieve subject id',subject)
            break
    # print(sender_id, subject_id)
    cur.execute('INSERT OR IGNORE INTO Messages (guid,sender_id,subject_id,sent_at,headers,body) VALUES ( ?,?,?,datetime(?),?,? )',
            ( guid, sender_id, subject_id, sent_at,
            zlib.compress(message_row[0].encode()), zlib.compress(message_row[1].encode())) )
    conn.commit()
    cur.execute('SELECT id FROM Messages WHERE guid=? LIMIT 1', ( guid, ))
    try:
        row = cur.fetchone()
        message_id = row[0]
        guids[guid] = message_id
    except:
        print('Could not retrieve guid id',guid)
        break

cur.close()
cur_1.close()

gbasic.py

import sqlite3
import time
import zlib

howmany = int(input("How many to dump? "))

conn = sqlite3.connect('index.sqlite')
cur = conn.cursor()

cur.execute('SELECT id, sender FROM Senders')
senders = dict()
for message_row in cur :
    senders[message_row[0]] = message_row[1]

cur.execute('SELECT id, subject FROM Subjects')
subjects = dict()
for message_row in cur :
    subjects[message_row[0]] = message_row[1]

# cur.execute('SELECT id, guid,sender_id,subject_id,headers,body FROM Messages')
cur.execute('SELECT id, guid,sender_id,subject_id,sent_at FROM Messages')
messages = dict()
for message_row in cur :
    messages[message_row[0]] = (message_row[1],message_row[2],message_row[3],message_row[4])

print("Loaded messages=",len(messages),"subjects=",len(subjects),"senders=",len(senders))

sendcounts = dict()
sendorgs = dict()
for (message_id, message) in list(messages.items()):
    sender = message[1]
    sendcounts[sender] = sendcounts.get(sender,0) + 1
    pieces = senders[sender].split("@")
    if len(pieces) != 2 : continue
    dns = pieces[1]
    sendorgs[dns] = sendorgs.get(dns,0) + 1

print('')
print('Top',howmany,'Email list participants')

x = sorted(sendcounts, key=sendcounts.get, reverse=True)
for k in x[:howmany]:
    print(senders[k], sendcounts[k])
    if sendcounts[k] < 10 : break

print('')
print('Top',howmany,'Email list organizations')

x = sorted(sendorgs, key=sendorgs.get, reverse=True)
for k in x[:howmany]:
    print(k, sendorgs[k])
    if sendorgs[k] < 10 : break