MEGA -- DB

This commit is contained in:
gauthiier
2019-07-11 13:21:42 +02:00
parent 3703dcc169
commit 4197cd4d32
25 changed files with 663 additions and 1657 deletions
View File
+257
View File
@@ -0,0 +1,257 @@
import email, email.parser
import os, json, gzip, re
import mysql.connector as mariadb
import archive.sql, archive.util
from datetime import date, datetime
from dateutil import parser
import terminal.progress
def load_from_file(filename, archive_name, archive_dir):
if not filename.endswith('.json.gz'):
file_path = os.path.join(archive_dir, filename + '.json.gz')
else:
file_path = os.path.join(archive_dir, filename)
if os.path.isfile(file_path):
with gzip.open(file_path, 'r') as fp:
json_data = json.load(fp)
return (json_data, archive_name)
else:
#list of all "filename[...].json.gz" in archive_dir
files = sorted([f for f in os.listdir(archive_dir) if os.path.isfile(os.path.join(archive_dir, f)) and f.startswith(filename) and f.endswith('.json.gz')])
if files:
filename = files[-1] # take the most recent (listed alpha-chronological)
file_path = os.path.join(archive_dir, filename)
if os.path.isfile(file_path):
with gzip.open(file_path, 'r') as fp:
json_data = json.load(fp)
return (json_data, archive_name) # <--- this makes no sense....
else:
#list of all json files in archive_dir/filename
dir_path = os.path.join(archive_dir, filename)
if not os.path.isdir(dir_path):
return None
files = [os.path.join(dir_path, f) for f in os.listdir(dir_path) if os.path.isfile(os.path.join(dir_path, f)) and f.endswith('.json')]
if not files:
return None
# load all json files
threads = []
for file_path in files:
with open(file_path, 'r') as fp:
json_data = json.load(fp)
threads.append(json_data)
return (threads, archive_name)
def connect_db(database, host, user, password):
try:
con = mariadb.connect(host=host, user=user, password=password, database=database)
except mariadb.Error as error:
print("Error: {}".format(error))
if error.errno == 1049:
if util.y_n_question("Table " + archive_name + " does not exist. Create it?"):
print("creating")
else:
print("not creating")
return None
finally:
return con
class Archive:
data = None # "raw" json data
db_con = None
def __init__(self, archive_name, archive_dir):
if isinstance(archive_name, str):
# need a filename or a dir name....
print("reading archive " + archive_name, end='')
(self.data, self.archive_name) = load_from_file(archive_name, archive_name, archive_dir)
print(" - done.")
def __init__(self, archive_name, database, host, user, password):
self.archive_name = archive_name
self.db_con = connect_db(database, host, user, password)
def __init__(self, archive_name, config):
self.archive_name = archive_name
self.db_con = connect_db(config['database'], config['host'], config['user'], config['password'])
def __enter__(self):
return self
def __exit__(self, exc_type, exc_value, traceback):
if self.db_con is not None:
self.db_con.close()
def create_db(self, host, database, user, password):
print("creating table: " + self.archive_name, end='')
self.db_con = connect_db(database, host, user, password)
if self.db_con is None:
return
try:
cursor = self.db_con.cursor()
cursor.execute(archive.sql.CREATE.format(self.archive_name))
except mariadb.Error as error:
print("Error: {}".format(error))
finally:
cursor.close()
print(" - done.")
def insert_db(self, host, database, user, password):
self.db_con = connect_db(database, host, user, password)
if self.db_con is None:
return
try:
cursor = self.db_con.cursor()
progress = terminal.progress.ProgressBar(self.archive_name, len(self.data), fmt=terminal.progress.ProgressBar.FULL)
for t in self.data:
n_inserted = self.recursive_insert_db(cursor, t["threads"])
# print(" - insert: " + str(n_inserted), end='')
if n_inserted > 0:
self.db_con.commit()
progress.current += 1
progress()
progress.done()
self.db_con.commit()
except mariadb.Error as error:
pass
# print("Error: {}".format(error))
finally:
cursor.close()
def recursive_insert_db(self, cursor, thread):
n_inserted = 0
for m in thread:
try:
from_ = archive.util.format_from(m)
author_name_ = archive.util.format_author(m)
to_ = archive.util.format_to(m)
date_ = archive.util.format_date(m, self.archive_name)
if date_ is None or from_ is None:
# print("\nerrorororororo")
# print(m['from'] + " -- " + m['date'])
continue
cursor.execute(archive.sql.INSERT, (from_,author_name_,to_,m["subject"],date_,m["content-type"],m["content"],m["url"]))
n_inserted += 1
if "follow-up" in m:
n_inserted += self.recursive_insert_db(cursor, m["follow-up"])
except mariadb.Error as error:
if error.errno == 1062:
#duplication continue <------------------------- look this up...
# print("\nError: {}".format(error))
continue
return n_inserted
def content_search(self, term, bool=True):
if self.db_con is None:
print("Not connection to database...")
return
try:
cursor = self.db_con.cursor(buffered=True)
if bool:
cursor.execute(archive.sql.CONTENT_QUERY_BOOLEAN.format(self.archive_name, term))
else:
cursor.execute(archive.sql.CONTENT_QUERY.format(self.archive_name, term))
# print(cursor.rowcount)
results = []
for (from_, author_name_, subject_, date_, url_) in cursor:
results.append((from_, author_name_, subject_, date_, url_))
# print("{} {} {}".format(from_, str(date_), url_))
return results
except mariadb.Error as error:
print("Error: {}".format(error))
finally:
cursor.close()
def from_search(self, term, bool=True):
if self.db_con is None:
print("Not connection to database...")
return
try:
cursor = self.db_con.cursor(buffered=True)
if bool:
cursor.execute(archive.sql.FROM_QUERY_BOOLEAN.format(self.archive_name, term))
else:
cursor.execute(archive.sql.FROM_QUERY.format(self.archive_name, term))
# print(cursor.rowcount)
results = []
for (from_, author_name_, subject_, date_, url_) in cursor:
results.append((from_, author_name_, subject_, date_, url_))
# print("{} {} {}".format(from_, str(date_), url_))
return results
except mariadb.Error as error:
print("Error: {}".format(error))
finally:
cursor.close()
# analysis
def longest_field(self, field, thread, max_length=0):
import archive.util
for m in thread:
if not field in m:
if "threads" in m:
max_length = self.longest_field(field, m["threads"], max_length)
continue
if m[field] is None:
continue
if field == "from":
m[field] = archive.util.format_from(m)
elif field == "author_name":
m[field] = archive.util.format_author(m)
elif field == "to":
m[field] = archive.util.format_to(m)
elif field == "date":
m[field] = str(archive.util.format_date(m, self.archive_name))
if m[field] is None:
continue
l = len(m[field])
if l > max_length:
max_length = l
print(">> " + m[field])
if "follow-up" in m:
max_length = self.longest_field(field, m["follow-up"], max_length)
return max_length
+31
View File
@@ -0,0 +1,31 @@
CREATE = "CREATE TABLE `{}` (" \
"`from_` varchar(85) NOT NULL," \
"`author_name_` varchar(200) NOT NULL," \
"`to_` text(60)," \
"`subject_` varchar(3500) NOT NULL," \
"`date_` datetime NOT NULL," \
"`content_type_` varchar(15) NOT NULL," \
"`content_` mediumtext NOT NULL," \
"`url_` varchar(100) NOT NULL," \
"PRIMARY KEY(`from_`, `date_`)," \
"FULLTEXT (`subject_`, `content_`)," \
"FULLTEXT (`from_`, `author_name_`)" \
") ENGINE = InnoDB;"
INSERT = ("INSERT INTO nettime_l"
"(from_, author_name_, to_, subject_, date_, content_type_, content_, url_) "
"VALUES (%s, %s, %s, %s, %s, %s, %s, %s)")
CONTENT_QUERY_BOOLEAN = ("SELECT from_, author_name_, subject_, date_, url_ FROM {} "
"WHERE MATCH(subject_, content_) AGAINST('{}' IN BOOLEAN MODE) ORDER BY date_")
CONTENT_QUERY_NL = ("SELECT from_, author_name_, subject_, date_, url_ FROM {} "
"WHERE MATCH(subject_, content_) AGAINST('{}') ORDER BY date_")
FROM_QUERY_BOOLEAN = ("SELECT from_, author_name_, subject_, date_, url_ FROM {} "
"WHERE MATCH(from_, author_name_) AGAINST('{}' IN BOOLEAN MODE) ORDER BY date_")
FROM_QUERY_NL = ("SELECT from_, author_name_, subject_, date_, url_ FROM {} "
"WHERE MATCH(from_, author_name_) AGAINST('{}') ORDER BY date_")
# SELECT from_, author_name_, subject_, date_, url_ FROM nettime_l WHERE MATCH(content_) AGAINST('%s' IN BOOLEAN MODE)
+225
View File
@@ -0,0 +1,225 @@
import email, datetime, sys
import hashlib
import dateparser
def format_content(msg):
return msg['content']
def format_url(msg):
return msg['url']
def format_author(msg):
if 'author_name' not in msg or msg['author_name'] is None:
return None
author_str = msg['author_name'].replace('"', '')
if "by way of" in author_str:
toks = author_str.split("by way of")
if toks[0] == "":
author_str = format_from(msg)
elif toks[0][-1] == "(":
author_str = toks[0][:-1].strip()
else:
author_str = toks[0]
if ("(" in author_str) or ("<" in author_str):
# ex. zx {AT} xyz.net (Michel Foucault) OR Michel Foucault (c'estcommeca.com) OR Michel Foucault <zx {AT} xyz.net>
# print("±±±±±±")
# print("name: " + author_str)
# print("from: " + msg['from'])
if not '@' in author_str.lower().replace('{at}', '@').replace(' at ', '@'):
author_str = author_str.split('(')[0].strip()
else:
author_str = email.utils.parseaddr(author_str)[0]
# print(" Name:" + author_str.replace('"', ''))
# print(" From:" + format_from(msg))
if " ," in author_str:
# nettime's_roving_reporter , thing.net {AT} bbs.thing.net
author_str = author_str.split(' ,')[0]
return author_str
def format_from_token(from_str, sep):
from_addr = email.utils.parseaddr(from_str)[1]
if sep not in from_addr:
tok = from_str.split()
try:
at = tok.index(sep)
from_addr = ''.join([tok[at-1], '{AT}', tok[at+1]])
if from_addr.startswith('<') or from_addr.endswith('>'):
from_addr = from_addr.strip('<').strip('>')
except ValueError:
print(tok)
print("error formating 'from' " + from_str + " -- expecting sep: " + sep)
return None
else:
from_addr = from_addr.replace(sep, '{AT}')
return "".join(from_addr.lower().split())
def format_from(msg):
if 'from' not in msg or msg['from'] is None:
return None
from_str = msg['from']
if " {AT} " in from_str:
return format_from_token(from_str, '{AT}')
elif " at " in from_str:
return format_from_token(from_str, 'at')
elif "@" in from_str:
return format_from_token(from_str, '@')
else:
return "".join(from_str.split())
def format_to(msg):
if "to" not in msg or msg["to"] is None:
return None
to_str = msg["to"]
toks = email.utils.parseaddr(to_str)
# print(toks)
if len(toks) == 2:
to_str = toks[1]
return "".join(to_str.lower().split())
# returns utc timestamp --- old...
def format_date_utc(msg, archive_name):
if 'date' not in msg or msg['date'] is None:
return None
date_str = msg['date'].replace('.', '')
time_tz = None
try:
date_tz = email.utils.parsedate_tz(date_str)
time_tz = email.utils.mktime_tz(date_tz) #utc timestamp
except TypeError:
print("Format Date TypeError")
print(" > " + date_str)
return None
except ValueError:
print("Format Date ValueError")
print(" > " + date_str)
return None
finally:
return time_tz
def format_date(msg, archive_name):
if 'date' not in msg or msg['date'] is None:
return None
# date_str = msg['date'].replace('.', '')
date_str = msg['date']
# fix Thu, 01 Aug 2002 17:33:08 +0900 (JST)
if '(' in date_str:
date_str = date_str.split('(')[0].rstrip()
date_time = dateparser.parse(date_str)
if date_time is None:
# random stuff...
fix = False
toks = date_str.split()
if len(toks[-1]) == 5 or len(toks[-1]) == 4:
# ex. Thu, 24 Jan 2002 15:21:31 -0000
if toks[-1] in ['+0000', '-0000', '0000']:
date_str = date_str[:-5]
fix = True
# ex. Fri, 25 Jan 2002 13:21:49 +1050
elif toks[-1][-2] == '5':
d = list(date_str)
d[-2] = '3'
date_str = "".join(d)
fix = True
if toks[-1][-1] != '0':
#ex. 'Fri,', '20', 'Jun', '1997', '02:58:59', '-0005'
date_str = date_str[:-5]
fix = True
if 'Fru' in toks[0]:
date_str = date_str.replace('Fru', 'Fri')
fix = True
elif 'Thur' in toks[0]:
date_str = date_str.replace('Thur', 'Thu')
fix = True
if not fix:
# print("----")
return None
date_time = dateparser.parse(date_str)
if date_time is None:
if 'GMT' in date_str:
# ex. 'Mon,', '15', 'Jan', '96', '02:55', 'GMT+0100'
date_str = date_str.split('GMT')[0].rstrip()
fix = True
if 'METDST' in toks[-1]:
# ex. 'Sat,', '3', 'May', '97', '21:07', 'METDST'
date_str = date_str.replace('METDST', 'MET')
fix = True
if not fix:
# print("++++")
return None
date_time = dateparser.parse(date_str)
return date_time
# else:
# print(date_str)
# date_time = datetime.datetime.fromtimestamp(time_tz)
min_d = datetime.datetime.strptime(min_date(archive_name), "%d/%m/%Y")
max_d = datetime.datetime.now()
date_time_naive = date_time.replace(tzinfo=None)
if date_time_naive < min_d or date_time_naive > max_d:
return None
return date_time
def format_subject(msg, archive_name):
if 'subject' not in msg or msg['subject'] is None:
return None
return msg['subject']
def format_id(msg, archive_name):
if "message-id" in msg:
return msg['message-id']
else:
# create hash with author_name + date
s = msg['author_name'] + msg['date']
sha = hashlib.sha1(s.encode('utf-8'))
return sha.hexdigest()
# format='%d/%m/%Y'
def min_date(archive_name):
if "nettime" in archive_name:
return '01/10/1995'
elif archive_name == "spectre":
return '01/08/2001'
elif archive_name == "empyre":
return '01/01/2002'
elif archive_name == "crumb":
return '01/02/2001'