gendummydata.py 8.87 KB
Newer Older
1
#!/usr/bin/python3
eric's avatar
eric committed
2
3
4
"""
usage: gendummydata.py outputfilename.sql
"""
eric's avatar
eric committed
5
6
7
8
9
10
11
#
# This script seeds the AUR database with dummy data for
# use during development/testing.  It uses random entries
# from /usr/share/dict/words to create user accounts and
# package names.  It generates the SQL statements to
# insert these users/packages into the AUR database.
#
12
13
14
15
import random
import time
import os
import sys
16
import io
17
import logging
18

19
LOG_LEVEL = logging.DEBUG # logging level. set to logging.INFO to reduce output
eric's avatar
eric committed
20
SEED_FILE = "/usr/share/dict/words"
21
22
23
24
DB_HOST   = os.getenv("DB_HOST", "localhost")
DB_NAME   = os.getenv("DB_NAME", "AUR")
DB_USER   = os.getenv("DB_USER", "aur")
DB_PASS   = os.getenv("DB_PASS", "aur")
eric's avatar
eric committed
25
USER_ID   = 5          # Users.ID of first bogus user
eric's avatar
eric committed
26
PKG_ID    = 1          # Packages.ID of first package
Loui Chang's avatar
Loui Chang committed
27
MAX_USERS = 300        # how many users to 'register'
eric's avatar
eric committed
28
29
MAX_DEVS  = .1         # what percentage of MAX_USERS are Developers
MAX_TUS   = .2         # what percentage of MAX_USERS are Trusted Users
Loui Chang's avatar
Loui Chang committed
30
MAX_PKGS  = 900       # how many packages to load
31
32
PKG_DEPS  = (1, 5)     # min/max depends a package has
PKG_SRC   = (1, 3)     # min/max sources a package has
eric's avatar
eric committed
33
PKG_CMNTS = (1, 5)     # min/max number of comments a package has
34
CATEGORIES_COUNT = 17  # the number of categories from aur-schema
eric's avatar
eric committed
35
VOTING    = (0, .30)   # percentage range for package voting
36
37
OPEN_PROPOSALS = 5 # number of open trusted user proposals
CLOSE_PROPOSALS = 15 # number of closed trusted user proposals
eric's avatar
eric committed
38
39
40
RANDOM_TLDS = ("edu", "com", "org", "net", "tw", "ru", "pl", "de", "es")
RANDOM_URL = ("http://www.", "ftp://ftp.", "http://", "ftp://")
RANDOM_LOCS = ("pub", "release", "files", "downloads", "src")
41
FORTUNE_FILE = "/usr/share/fortune/cookie"
42

43
44
45
46
# setup logging
logformat = "%(levelname)s: %(message)s"
logging.basicConfig(format=logformat, level=LOG_LEVEL)
log = logging.getLogger()
eric's avatar
eric committed
47
48

if len(sys.argv) != 2:
49
	log.error("Missing output filename argument")
eric's avatar
eric committed
50
51
52
53
54
	raise SystemExit

# make sure the seed file exists
#
if not os.path.exists(SEED_FILE):
55
	log.error("Please install the 'words' Arch package")
56
	raise SystemExit
eric's avatar
eric committed
57

58
59
# make sure comments can be created
#
60
if not os.path.exists(FORTUNE_FILE):
61
62
63
	log.error("Please install the 'fortune-mod' Arch package")
	raise SystemExit

eric's avatar
eric committed
64
65
66
67
68
69
70
71
# track what users/package names have been used
#
seen_users = {}
seen_pkgs = {}
user_keys = []

# some functions to generate random data
#
72
def genVersion():
eric's avatar
eric committed
73
74
75
	ver = []
	ver.append("%d" % random.randrange(0,10))
	ver.append("%d" % random.randrange(0,20))
eric's avatar
eric committed
76
	if random.randrange(0,2) == 0:
eric's avatar
eric committed
77
		ver.append("%d" % random.randrange(0,100))
78
	return ".".join(ver) + "-%d" % random.randrange(1,11)
eric's avatar
eric committed
79
def genCategory():
80
	return random.randrange(1,CATEGORIES_COUNT)
eric's avatar
eric committed
81
82
def genUID():
	return seen_users[user_keys[random.randrange(0,len(user_keys))]]
83
84
def genFortune():
	return fortunes[random.randrange(0,len(fortunes))].replace("'", "")
eric's avatar
eric committed
85
86
87
88


# load the words, and make sure there are enough words for users/pkgs
#
89
log.debug("Grabbing words from seed file...")
eric's avatar
eric committed
90
91
92
93
94
95
96
97
98
99
100
101
102
103
fp = open(SEED_FILE, "r")
contents = fp.readlines()
fp.close()
if MAX_USERS > len(contents):
	MAX_USERS = len(contents)
if MAX_PKGS > len(contents):
	MAX_PKGS = len(contents)
if len(contents) - MAX_USERS > MAX_PKGS:
	need_dupes = 0
else:
	need_dupes = 1

# select random usernames
#
104
log.debug("Generating random user names...")
eric's avatar
eric committed
105
106
107
user_id = USER_ID
while len(seen_users) < MAX_USERS:
	user = random.randrange(0, len(contents))
eric's avatar
eric committed
108
109
	word = contents[user].replace("'", "").replace(".","").replace(" ", "_")
	word = word.strip().lower()
110
	if word not in seen_users:
eric's avatar
eric committed
111
112
		seen_users[word] = user_id
		user_id += 1
113
user_keys = list(seen_users.keys())
eric's avatar
eric committed
114
115
116

# select random package names
#
117
log.debug("Generating random package names...")
eric's avatar
eric committed
118
119
120
num_pkgs = PKG_ID
while len(seen_pkgs) < MAX_PKGS:
	pkg = random.randrange(0, len(contents))
eric's avatar
eric committed
121
122
	word = contents[pkg].replace("'", "").replace(".","").replace(" ", "_")
	word = word.strip().lower()
eric's avatar
eric committed
123
	if not need_dupes:
124
		if word not in seen_pkgs and word not in seen_users:
eric's avatar
eric committed
125
126
127
			seen_pkgs[word] = num_pkgs
			num_pkgs += 1
	else:
128
		if word not in seen_pkgs:
eric's avatar
eric committed
129
130
131
			seen_pkgs[word] = num_pkgs
			num_pkgs += 1

eric's avatar
eric committed
132
133
134
135
# free up contents memory
#
contents = None

136
137
138
139
140
141
142
# developer/tu IDs
#
developers = []
trustedusers = []
has_devs = 0
has_tus = 0

143
144
145
146
147
# Just let python throw the errors if any happen
#
out = open(sys.argv[1], "w")
out.write("BEGIN;\n")

eric's avatar
eric committed
148
149
# Begin by creating the User statements
#
150
log.debug("Creating SQL statements for users.")
eric's avatar
eric committed
151
for u in user_keys:
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
	account_type = 1  # default to normal user
	if not has_devs or not has_tus:
		account_type = random.randrange(1, 4)
		if account_type == 3 and not has_devs:
			# this will be a dev account
			#
			developers.append(seen_users[u])
			if len(developers) >= MAX_DEVS * MAX_USERS:
				has_devs = 1
		elif account_type == 2 and not has_tus:
			# this will be a trusted user account
			#
			trustedusers.append(seen_users[u])
			if len(trustedusers) >= MAX_TUS * MAX_USERS:
				has_tus = 1
		else:
			# a normal user account
			#
			pass
171

172
173
174
	s = ("INSERT INTO Users (ID, AccountTypeID, Username, Email, Passwd)"
		 " VALUES (%d, %d, '%s', '%s@example.com', MD5('%s'));\n")
	s = s % (seen_users[u], account_type, u, u, u)
eric's avatar
eric committed
175
	out.write(s)
176
177
178
179
180

log.debug("Number of developers: %d" % len(developers))
log.debug("Number of trusted users: %d" % len(trustedusers))
log.debug("Number of users: %d" % (MAX_USERS-len(developers)-len(trustedusers)))
log.debug("Number of packages: %d" % MAX_PKGS)
eric's avatar
eric committed
181

182
183
184
185
186
log.debug("Gathering text from fortune file...")
fp = open(FORTUNE_FILE, "r")
fortunes = fp.read().split("%\n")
fp.close()

eric's avatar
eric committed
187
188
# Create the package statements
#
189
log.debug("Creating SQL statements for packages.")
eric's avatar
eric committed
190
count = 0
191
for p in list(seen_pkgs.keys()):
eric's avatar
eric committed
192
	NOW = int(time.time())
193
194
195
196
197
	if count % 2 == 0:
		muid = developers[random.randrange(0,len(developers))]
	else:
		muid = trustedusers[random.randrange(0,len(trustedusers))]
	if count % 20 == 0: # every so often, there are orphans...
198
		muid = "NULL"
199

eric's avatar
eric committed
200
201
	uuid = genUID() # the submitter/user

202
203
204
205
	s = ("INSERT INTO Packages (ID, Name, Version, CategoryID,"
		 " SubmittedTS, SubmitterUID, MaintainerUID) VALUES "
		 " (%d, '%s', '%s', %d, %d, %d, %s);\n")
	s = s % (seen_pkgs[p], p, genVersion(), genCategory(), NOW, uuid, muid)
206

eric's avatar
eric committed
207
208
209
	out.write(s)
	count += 1

eric's avatar
eric committed
210
211
212
213
214
	# create random comments for this package
	#
	num_comments = random.randrange(PKG_CMNTS[0], PKG_CMNTS[1])
	for i in range(0, num_comments):
		now = NOW + random.randrange(400, 86400*3)
215
216
		s = ("INSERT INTO PackageComments (PackageID, UsersID,"
			 " Comments, CommentTS) VALUES (%d, %d, '%s', %d);\n")
217
		s = s % (seen_pkgs[p], genUID(), genFortune(), now)
eric's avatar
eric committed
218
219
		out.write(s)

eric's avatar
eric committed
220
221
# Cast votes
#
222
track_votes = {}
223
log.debug("Casting votes for packages.")
eric's avatar
eric committed
224
for u in user_keys:
225
226
	num_votes = random.randrange(int(len(seen_pkgs)*VOTING[0]),
			int(len(seen_pkgs)*VOTING[1]))
eric's avatar
eric committed
227
228
	pkgvote = {}
	for v in range(num_votes):
229
		pkg = random.randrange(1, len(seen_pkgs) + 1)
230
		if pkg not in pkgvote:
231
232
233
			s = ("INSERT INTO PackageVotes (UsersID, PackageID)"
				 " VALUES (%d, %d);\n")
			s = s % (seen_users[u], pkg)
eric's avatar
eric committed
234
			pkgvote[pkg] = 1
235
			if pkg not in track_votes:
236
237
				track_votes[pkg] = 0
			track_votes[pkg] += 1
eric's avatar
eric committed
238
239
			out.write(s)

240
241
# Update statements for package votes
#
242
for p in list(track_votes.keys()):
243
244
	s = "UPDATE Packages SET NumVotes = %d WHERE ID = %d;\n"
	s = s % (track_votes[p], p)
245
246
	out.write(s)

247
248
# Create package dependencies and sources
#
249
log.debug("Creating statements for package depends/sources.")
250
for p in list(seen_pkgs.keys()):
251
252
253
254
	num_deps = random.randrange(PKG_DEPS[0], PKG_DEPS[1])
	this_deps = {}
	i = 0
	while i != num_deps:
255
		dep = random.choice([k for k in seen_pkgs])
256
		if dep not in this_deps:
257
			s = "INSERT INTO PackageDepends VALUES (%d, '%s', NULL);\n"
258
			s = s % (seen_pkgs[p], dep)
259
260
261
262
263
264
265
266
267
268
269
			out.write(s)
			i += 1

	num_sources = random.randrange(PKG_SRC[0], PKG_SRC[1])
	for i in range(num_sources):
		src_file = user_keys[random.randrange(0, len(user_keys))]
		src = "%s%s.%s/%s/%s-%s.tar.gz" % (
				RANDOM_URL[random.randrange(0,len(RANDOM_URL))],
				p, RANDOM_TLDS[random.randrange(0,len(RANDOM_TLDS))],
				RANDOM_LOCS[random.randrange(0,len(RANDOM_LOCS))],
				src_file, genVersion())
270
271
		s = "INSERT INTO PackageSources VALUES (%d, '%s');\n"
		s = s % (seen_pkgs[p], src)
272
273
		out.write(s)

274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
# Create trusted user proposals
#
log.debug("Creating SQL statements for trusted user proposals.")
count=0
for t in range(0, OPEN_PROPOSALS+CLOSE_PROPOSALS):
	now = int(time.time())
	if count < CLOSE_PROPOSALS:
		start =  now - random.randrange(3600*24*7, 3600*24*21)
		end = now - random.randrange(0, 3600*24*7)
	else:
		start = now
		end = now + random.randrange(3600*24, 3600*24*7)
	if count % 5 == 0: # Don't make the vote about anyone once in a while
		user = ""
	else:
		user = user_keys[random.randrange(0,len(user_keys))]
	suid = trustedusers[random.randrange(0,len(trustedusers))]
	s = ("INSERT INTO TU_VoteInfo (Agenda, User, Submitted, End,"
	" SubmitterID) VALUES ('%s', '%s', %d, %d, %d);\n")
293
	s = s % (genFortune(), user, start, end, suid)
294
295
296
	out.write(s)
	count += 1

eric's avatar
eric committed
297
298
# close output file
#
299
out.write("COMMIT;\n")
eric's avatar
eric committed
300
301
out.write("\n")
out.close()
302
log.debug("Done.")