weblog-model.lua 50.3 KB
Newer Older
Ted Trask's avatar
Ted Trask committed
1 2 3 4
module(..., package.seeall)

-- Load libraries
require("modelfunctions")
5 6 7
fs = require("acf.fs")
format = require("acf.format")
validator = require("acf.validator")
Ted Trask's avatar
Ted Trask committed
8
require("luasql.postgres")
9 10
require("posix")
require("subprocess")
Ted Trask's avatar
Ted Trask committed
11

12
local DatabaseName = "webproxylog"
Ted Trask's avatar
Ted Trask committed
13 14 15 16 17 18
local DatabaseOwner = "weblogowner"
local DatabaseUser = "webloguser"

local path = "PATH=/usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/sbin:/sbin "
local env
local con
19
local configfile = "/etc/weblog/weblog.conf"
Ted Trask's avatar
Ted Trask committed
20 21
local configcontent = fs.read_file(configfile) or ""
local config = format.parse_ini_file(configcontent, "") or {}
22
local goodwordslist = "/etc/weblog/goodwords"
23
local goodwords
24
local badwordslist = "/etc/weblog/badwords"
25
local badwords
Ted Trask's avatar
Ted Trask committed
26
local ignorewordslist = "/etc/weblog/ignorewords"
27
local ignorewords
Ted Trask's avatar
Ted Trask committed
28
local files = {badwordslist, goodwordslist, ignorewordslist, configfile}
Ted Trask's avatar
Ted Trask committed
29 30

local database_creation_script = {
31
	"CREATE TABLE dbhistlog (logdatetime timestamp(3) without time zone NOT NULL, msgtext text)",
32 33
	"CREATE TABLE pubweblog(sourcename character varying(40), clientip inet NOT NULL, clientuserid character varying(64) NOT NULL, logdatetime timestamp(3) without time zone NOT NULL, uri text NOT NULL, bytes bigint NOT NULL, reason text, score integer, shortreason text, badyesno int, deniedyesno int, bypassyesno int, wordloc text, goodwordloc text, selected boolean, id serial)",
	"CREATE TABLE pubweblog_history(sourcename character varying(40), clientip inet NOT NULL, clientuserid character varying(64) NOT NULL, logdatetime timestamp(3) without time zone NOT NULL, uri text NOT NULL, bytes bigint NOT NULL, reason text, score integer, shortreason text, badyesno int, deniedyesno int, bypassyesno int, wordloc text, goodwordloc text, selected boolean, id int)",
34
	"CREATE TABLE weblog(sourcename character varying(40), clientip inet NOT NULL, clientuserid character varying(64) NOT NULL, logdatetime timestamp(3) without time zone NOT NULL, uri text NOT NULL, bytes bigint NOT NULL, reason text, score integer, shortreason text, badyesno int, deniedyesno int, bypassyesno int, wordloc text, goodwordloc text)",
Ted Trask's avatar
Ted Trask committed
35
	"CREATE TABLE source (sourcename character varying(40) NOT NULL, method character varying(100) NOT NULL, userid character varying(32), passwd character varying(255), source character varying(255) NOT NULL, tzislocal boolean, enabled boolean)",
36
	"CREATE TABLE usagestat (sourcename character varying(40) NOT NULL, date timestamp(0) without time zone NOT NULL, numrequest integer, numblock integer)",
Ted Trask's avatar
Ted Trask committed
37 38 39 40 41
	"ALTER TABLE ONLY source ADD CONSTRAINT source_pkey PRIMARY KEY (sourcename)",
	"CREATE INDEX dbhistlogdatetimeidx ON dbhistlog USING btree (logdatetime)",
	"CREATE INDEX pubweblogclientdateidx ON pubweblog USING btree (logdatetime, clientuserid)",
	"CREATE INDEX pubweblogclientuserididx ON pubweblog USING btree (clientuserid)",
	"CREATE INDEX pubwebloglogdatetimeidx ON pubweblog USING btree (logdatetime)",
42 43 44
	"CREATE INDEX pubweblog_historyclientdateidx ON pubweblog_history USING btree (logdatetime, clientuserid)",
	"CREATE INDEX pubweblog_historyclientuserididx ON pubweblog_history USING btree (clientuserid)",
	"CREATE INDEX pubweblog_historylogdatetimeidx ON pubweblog_history USING btree (logdatetime)",
45 46
	"GRANT SELECT ON dbhistlog TO "..DatabaseUser,
	"GRANT SELECT ON pubweblog TO "..DatabaseUser,
47
	"GRANT SELECT ON pubweblog_history TO "..DatabaseUser,
48 49
	"GRANT SELECT, UPDATE, INSERT, DELETE ON source TO "..DatabaseUser,
	"GRANT SELECT ON usagestat TO "..DatabaseUser,
Ted Trask's avatar
Ted Trask committed
50 51 52 53 54 55 56 57 58 59 60 61 62
}

-- ################################################################################
-- DATABASE FUNCTIONS

local function assert (v, m)
	if not v then
		m = m or "Assertion failed!"
		error(m, 0)
	end
	return v, m
end

63 64
-- Escape special characters in sql statements and truncate to length
local escape = function(sql, length)
Ted Trask's avatar
Ted Trask committed
65
	sql = sql or ""
66
	if length then sql = string.sub(sql, 1, length) end
67
	return con:escape(sql)
Ted Trask's avatar
Ted Trask committed
68 69 70 71 72
end

-- List the postgres databases on this system
local listdatabases = function()
	local dbs = {}
73
	local result = modelfunctions.run_executable({"psql", "-U", "postgres", "-tl"})
Ted Trask's avatar
Ted Trask committed
74 75 76 77 78 79 80 81 82 83 84
	for line in string.gmatch(result, "[^\n]+") do
		dbs[#dbs+1] = string.match(line, "^ (%S+)")
	end
	return dbs
end

-- Create the necessary database
local createdatabase = function(password)
	local result = {}

	-- First, create the users
85 86 87 88 89 90 91 92 93
	local cmd = "CREATE USER "..DatabaseOwner.." WITH PASSWORD '"..password.."'"
	local cmdresult, errtxt = modelfunctions.run_executable({"psql", "-U", "postgres", "-c", cmd}, true)
	table.insert(result, errtxt)
	table.insert(result, cmdresult)

	cmd = "CREATE USER "..DatabaseUser
	cmdresult, errtxt = modelfunctions.run_executable({"psql", "-U", "postgres", "-c", cmd}, true)
	table.insert(result, errtxt)
	table.insert(result, cmdresult)
Timo Teräs's avatar
Timo Teräs committed
94

Ted Trask's avatar
Ted Trask committed
95
	-- Create the database
96 97 98 99
	cmd = "CREATE DATABASE "..DatabaseName.." WITH OWNER "..DatabaseOwner
	cmdresult, errtxt = modelfunctions.run_executable({"psql", "-U", "postgres", "-c", cmd}, true)
	table.insert(result, errtxt)
	table.insert(result, cmdresult)
Timo Teräs's avatar
Timo Teräs committed
100

Ted Trask's avatar
Ted Trask committed
101 102 103 104 105 106 107
	return table.concat(result, "\n")
end

-- Delete the database and roles
local deletedatabase = function()
	local result = {}

108 109 110 111 112 113 114 115 116 117 118 119 120 121
	local cmd = "DROP DATABASE "..DatabaseName
	local cmdresult, errtxt = modelfunctions.run_executable({"psql", "-U", "postgres", "-c", cmd}, true)
	table.insert(result, errtxt)
	table.insert(result, cmdresult)
	
	cmd = "DROP ROLE "..DatabaseUser
	cmdresult, errtxt = modelfunctions.run_executable({"psql", "-U", "postgres", "-c", cmd}, true)
	table.insert(result, errtxt)
	table.insert(result, cmdresult)
	
	cmd = "DROP ROLE "..DatabaseOwner
	cmdresult, errtxt = modelfunctions.run_executable({"psql", "-U", "postgres", "-c", cmd}, true)
	table.insert(result, errtxt)
	table.insert(result, cmdresult)
Timo Teräs's avatar
Timo Teräs committed
122

Ted Trask's avatar
Ted Trask committed
123 124 125 126 127
	return table.concat(result, "\n")
end

-- Run an SQL script
local runSQLscript = function(filename)
128 129
	local result, errtxt = modelfunctions.run_executable({"psql", "-U", "postgres", "-f", filename, DatabaseName}, true)
	return errtxt or result
Ted Trask's avatar
Ted Trask committed
130 131 132
end

-- Create the database and tables
133
-- pg_dump -U postgres -c webproxylog > makeweblog.postgres
Ted Trask's avatar
Ted Trask committed
134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
--runSQLscript("/root/work/weblog/makeweblog.postgres")

local databaseconnect = function(username, password)
	if not con then
		-- create environment object
		env = assert (luasql.postgres())
		-- connect to data source
		con = assert (env:connect(DatabaseName, username, password))
	end
end

local databasedisconnect = function()
	if env then
		env:close()
		env = nil
	end
	if con then
		con:close()
		con = nil
	end
end

local logme = function(message)
157
	local sql = string.format("INSERT INTO dbhistlog VALUES ('now', '%s')", escape(message))
Ted Trask's avatar
Ted Trask committed
158 159 160 161 162 163
	local res = assert (con:execute(sql))
end

local listhistorylogentries = function()
	local entries = {}
	-- retrieve a cursor
Timo Teräs's avatar
Timo Teräs committed
164
	cur = assert (con:execute"SELECT logdatetime, msgtext from dbhistlog ORDER BY logdatetime")
Ted Trask's avatar
Ted Trask committed
165 166 167 168 169 170 171 172 173 174
	row = cur:fetch ({}, "a")
	while row do
		entries[#entries+1] = {logdatetime = row.logdatetime, msgtext = row.msgtext}
		row = cur:fetch (row, "a")
	end
	-- close everything
	cur:close()
	return entries
end

175
local importlogentry = function(entry, sourcename)
176
	if entry then
177
		local sql = string.format("INSERT INTO weblog VALUES ('%s', '%s', '%s', '%s', '%s', '%s','%s','%s','%s','%s','%s','%s','%s')",
178
			escape(sourcename), escape(entry.clientip), escape(entry.clientuserid, 64):lower(),
179
			escape(entry.logdatetime), escape(entry.URL), escape(entry.bytes), escape(entry.reason), escape(entry.score or "0"), escape(entry.shortreason), escape(entry.badyesno or "0"), escape(entry.deniedyesno or "0"), escape(entry.bypassyesno or "0"), escape(entry.wordloc), escape(entry.goodwordloc))
Ted Trask's avatar
Ted Trask committed
180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230
		local res = assert (con:execute(sql))
	end
end

local listsourceentries = function(sourcename)
	local sources = {}
	-- retrieve a cursor
	local sql = "SELECT sourcename, method, userid, passwd, source, tzislocal, enabled FROM source"
	if sourcename then
		sql = sql .. " WHERE sourcename='" .. escape(sourcename) .. "'"
	end
	sql = sql .. " ORDER BY sourcename"
	cur = assert (con:execute(sql))
	row = cur:fetch ({}, "a")
	while row do
		row.tzislocal = (row.tzislocal == "t")
		row.enabled = (row.enabled == "t")
		sources[#sources+1] = row
		row = cur:fetch ({}, "a")
	end
	cur:close()
	return sources
end

local importsourceentry = function(source)
	local sql = string.format("INSERT INTO source VALUES ('%s', '%s', '%s', '%s', '%s', '%s', '%s')",
		escape(source.sourcename), escape(source.method), escape(source.userid), escape(source.passwd),
		escape(source.source), escape(tostring(source.tzislocal):upper()), escape(tostring(source.enabled):upper()))
	local res = assert (con:execute(sql))
	return res
end

local updatesourceentry = function(source)
	local sql = string.format("UPDATE source SET method='%s', userid='%s', passwd='%s', source='%s', tzislocal='%s', enabled='%s' WHERE sourcename='%s'",
		escape(source.method), escape(source.userid), escape(source.passwd), escape(source.source),
		escape(tostring(source.tzislocal):upper()), escape(tostring(source.enabled):upper()),
		escape(source.sourcename))
	local res = assert (con:execute(sql))
	return res
end

local deletesourceentry = function(sourcename)
	local sql = string.format("DELETE FROM source WHERE sourcename='%s'", escape(sourcename))
	local res = assert (con:execute(sql))
	return res
end

-- Generate usage statistics from weblog and blocklog
local updateusagestats = function()
	-- update the usagestat table from weblog
	-- (will result in multiple rows where logs rotated on partial hours)
231 232 233
	local sql = "insert into usagestat select weblog.sourcename, " ..
		"date_trunc('hour', weblog.logdatetime) as date, " ..
		"count(*), SUM(deniedyesno) from weblog group by sourcename,date"
Ted Trask's avatar
Ted Trask committed
234 235 236
	local res = assert (con:execute(sql))
end

237 238 239 240 241 242 243 244 245 246 247 248 249 250 251
-- Move weblog into pubweblog, and truncate weblog
local importpubweblog = function()
	local sql = "ANALYZE"
	res = assert (con:execute(sql))

	-- Move weblog to pubweblog
	sql= "insert into pubweblog select * from weblog"
	res = assert (con:execute(sql))
	logme("importpubweblog imported " .. res .. " new rows into database.")

	-- Truncate the staging table
	assert (con:execute("truncate weblog"))
	logme("truncated staging table")
end

Ted Trask's avatar
Ted Trask committed
252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268
-- Delete useage stats from more than a year ago
local groomusagestat = function()
	local res = assert (con:execute("delete from usagestat where " ..
		"date < (now() - INTERVAL '1 year')"))
	logme("removed " .. res .. " old usage status lines")
end

-- Delete history log information from more than a month ago
local groomdbhistlog = function()
	local res = assert (con:execute("delete from dbhistlog where " ..
		"logdatetime < (now() - INTERVAL '1 month')"))
	logme("removed " .. res .. " old dbhistlog lines")
end

-- Delete old junk from pub tables
local groompublogs = function()
	local purgedays = config.purgedays or 30
Ted Trask's avatar
Ted Trask committed
269

Ted Trask's avatar
Ted Trask committed
270 271 272
	local now = os.time()

	local temp = os.date("%Y-%m-%d %H:%M:%S", now - purgedays*86400)
273
	logme("Purgedate is " .. temp .. ". Nothing will exist in pubweblog from before purgedate.")
Ted Trask's avatar
Ted Trask committed
274

275
	-- Move flagged records to histoy and then purge anything older than purgedate
276
	sql = "Insert into pubweblog_history select * from pubweblog where logdatetime < '" .. escape(temp) .."' and (badyesno > 0 or deniedyesno > 0 or bypassyesno > 0 or selected = 'true')"
Ted Trask's avatar
Ted Trask committed
277
	res = assert (con:execute(sql))
278
	logme("Moved " .. res .. " old records to history")
Ted Trask's avatar
Ted Trask committed
279

280
	sql = "Delete from pubweblog where logdatetime < '" .. escape(temp) .."'"
Ted Trask's avatar
Ted Trask committed
281
	res = assert (con:execute(sql))
282 283 284 285 286
	logme("Deleted " .. res .. " old records from pubweblog")
	
	sql = "delete from pubweblog_history where logdatetime < (now() - INTERVAL '1 year')"
	res = assert (con:execute(sql))
	logme("Deleted " .. res .. " old records from pubweblog_history")
Ted Trask's avatar
Ted Trask committed
287 288
end

289
local generatewhereclause = function(clientuserid, starttime, endtime, clientip, badyesno, deniedyesno, bypassyesno, score, urisearch, selected)
Ted Trask's avatar
Ted Trask committed
290 291 292
	local sql = ""
	local where = {}
	if clientuserid and clientuserid ~= "" then
293
		where[#where+1] = "clientuserid LIKE '%"..escape(clientuserid).."%'"
Ted Trask's avatar
Ted Trask committed
294 295 296 297 298 299 300 301 302 303
	end
	if starttime and starttime ~= "" then
		where[#where+1] = "logdatetime >= '"..escape(starttime).."'"
	end
	if endtime and endtime ~= "" then
		where[#where+1] = "logdatetime <= '"..escape(endtime).."'"
	end
	if clientip and clientip ~= "" then
		where[#where+1] = "clientip = '"..escape(clientip).."'"
	end
304 305
	if badyesno then
		where[#where+1] = "badyesno = '1'"
306
	end
307 308
	if deniedyesno then
		where[#where+1] = "deniedyesno = '1'"
309
	end
310 311
	if bypassyesno then
		where[#where+1] = "bypassyesno = '1'"
312 313 314 315
	end
	if score and score ~= "" then
		where[#where+1] = "score >= '"..escape(score).."'"
	end
Ted Trask's avatar
Ted Trask committed
316
	if urisearch and urisearch ~= "" then
317 318
	    	where[#where+1] = "lower(uri) LIKE '%"..escape(urisearch).."%'"
	end
319
	if selected then
320 321
		where[#where+1] = "selected = 'true'"
	end
Ted Trask's avatar
Ted Trask committed
322

Ted Trask's avatar
Ted Trask committed
323 324 325
	if #where > 0 then
		sql = " WHERE " .. table.concat(where, " AND ")
	end
Ted Trask's avatar
Ted Trask committed
326

Ted Trask's avatar
Ted Trask committed
327 328 329
	return sql
end

330
local listlogentries = function(activelog, clientuserid, starttime, endtime, clientip, badyesno, deniedyesno, bypassyesno, score, urisearch, sortby, selected)
Ted Trask's avatar
Ted Trask committed
331 332
	local entries = {}
	-- retrieve a cursor
333
	local sql = "SELECT * FROM "..escape(activelog)
334
	sql = sql .. generatewhereclause(clientuserid, starttime, endtime, clientip, badyesno, deniedyesno, bypassyesno, score, urisearch, selected)
335
	sql = sql .. " ORDER BY "..escape(sortby)
Ted Trask's avatar
Ted Trask committed
336 337 338
	cur = assert (con:execute(sql))
	row = cur:fetch ({}, "a")
	while row do
Ted Trask's avatar
Ted Trask committed
339 340
		if config.shorturi == "true" then
			shorturi=string.gsub(row.uri, "[;?].*", "...")
341
		end
342
		entries[#entries+1] = {sourcename=row.sourcename, clientip=row.clientip, clientuserid=row.clientuserid, logdatetime=row.logdatetime, uri=row.uri, shorturi=shorturi, bytes=row.bytes, reason=row.reason, score=row.score, shortreason=row.shortreason, badyesno=row.badyesno, deniedyesno=row.deniedyesno, bypassyesno=row.bypassyesno, wordloc=row.wordloc, id=row.id, selected=row.selected }
343 344 345
		if (config.shortreason ~= "true") then
			entries[#entries].shortreason = nil
		end
Ted Trask's avatar
Ted Trask committed
346 347 348 349 350 351 352
		row = cur:fetch (row, "a")
	end
	-- close everything
	cur:close()
	return entries
end

353
local groupflaggedlogentries = function(starttime, endtime, groupby)
354 355 356
	groupby = groupby or "clientuserid"
	local entries = {}
	-- retrieve a cursor
357
	local sql = "SELECT "..escape(groupby)..", COUNT(*) as numrecords, SUM(CASE WHEN (bypassyesno > '0' OR deniedyesno > '0' OR badyesno > '0') THEN 1 ELSE 0 END) as numflagged, sum(score) AS numhits, sum(CASE WHEN deniedyesno > '0' THEN 1 ELSE 0 END) AS numdenied, sum(CASE WHEN bypassyesno > '0' THEN 1 ELSE 0 END) AS numbypassed, max(score) as maxscore from pubweblog"
358
	sql = sql .. generatewhereclause(nil, starttime, endtime)
359
	sql = sql .. " GROUP BY " ..escape(groupby).. " ORDER BY numflagged DESC"
360 361 362
	cur = assert (con:execute(sql))
	row = cur:fetch ({}, "a")
	while row do
363
		entries[#entries+1] = {numrecords=row.numrecords, numflagged=row.numflagged, numhits=row.numhits, numdenied=row.numdenied, numbypassed=row.numbypassed, maxscore=row.maxscore}
364 365 366 367 368 369 370 371
		entries[#entries][groupby] = row[groupby]
		row = cur:fetch (row, "a")
	end
	-- close everything
	cur:close()
	return entries
end

Ted Trask's avatar
Ted Trask committed
372 373 374 375
local listusagestats = function()
	local entries = {}
	-- retrieve a cursor
	local sql = "SELECT sourcename, date, sum(numrequest) AS numrequest, sum(numblock) AS numblock " ..
376
		"FROM usagestat GROUP BY sourcename, date ORDER BY date, sourcename"
Ted Trask's avatar
Ted Trask committed
377 378 379 380 381 382 383 384 385 386 387
	cur = assert (con:execute(sql))
	row = cur:fetch ({}, "a")
	while row do
		entries[#entries+1] = {sourcename=row.sourcename, date=row.date, numrequest=row.numrequest, numblock=row.numblock}
		row = cur:fetch (row, "a")
	end
	-- close everything
	cur:close()
	return entries
end

388 389 390
local testdatabaseentry = function(datatype, value)
	local success = true
	local errtxt
391
	local sql = "CREATE TEMP TABLE testing ( test "..escape(datatype).." DEFAULT '"..escape(value).."' ) ON COMMIT DROP"
392 393 394 395 396 397 398 399 400 401
	local res, err = pcall(function()
		assert (con:execute(sql))
	end)
	if not res then
		success = false
		errtxt = string.gsub(err or "", "\n.*", "")
	end
	return success, errtxt
end

402 403 404 405 406
local convertdatabaseentry = function(datatype, value)
	local success = true
	local errtxt
	local result = value
	local res, err = pcall(function()
407
		local sql = "CREATE TEMP TABLE testing ( test "..escape(datatype).." )"
408
		assert (con:execute(sql))
409
		sql = "INSERT INTO testing VALUES ('"..escape(value).."')"
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428
		assert (con:execute(sql))
		sql = "SELECT * FROM testing"
		local cur = assert (con:execute(sql))
		local row = cur:fetch ({}, "a")
		if row then
			result = row.test
		end
	end)
	if not res then
		success = false
		errtxt = string.gsub(err or "", "\n.*", "")
	end
	local res, err = pcall(function()
		local sql = "DROP TABLE testing"
		assert (con:execute(sql))
	end)
	return success, errtxt, result
end

Ted Trask's avatar
Ted Trask committed
429 430 431
local printtableentries = function(tablename)
	-- retrieve a cursor
	local count = 0
432
	cur = assert (con:execute("SELECT * from "..escape(tablename)))
Ted Trask's avatar
Ted Trask committed
433 434 435 436 437
	-- print all rows, the rows will be indexed by field names
	row = cur:fetch ({}, "a")
	while row do
		count = count + 1
		for name,val in pairs(row) do
Ted Trask's avatar
Ted Trask committed
438
			logevent(name.." = "..val..", ")
Ted Trask's avatar
Ted Trask committed
439 440 441 442 443
		end
		row = cur:fetch (row, "a")
	end
	-- close everything
	cur:close()
Ted Trask's avatar
Ted Trask committed
444
	logevent("Table "..tablename.." contains "..count.." rows")
Ted Trask's avatar
Ted Trask committed
445 446 447 448 449
end

-- ################################################################################
-- LOG FILE FUNCTIONS

450
local function checkwords(logentry)
451 452
	local goodwordloc={}
	local badwordloc={}
453 454
	local wrdcnt=0
	local isbad=0
455

456
	--check for ignored records first
457
	for i,thisline in ipairs(ignorewords) do
458 459 460
		if not thisline then
			break
		end
461
		_,instcnt = string.lower(logentry.URL):gsub(format.escapemagiccharacters(thisline), " ")
462
		if instcnt ~= 0 then
463
			logentry.ignoreme = true
464
			--logme("ignoring...")
465
			break
466 467 468
		end
	end

469
	if not logentry.ignoreme then
470 471 472 473 474
		--proceed with record analysis
		for i,thisline in ipairs(badwords) do
			if not thisline then
				break
			end
Ted Trask's avatar
Ted Trask committed
475

476
			_,instcnt = string.lower(logentry.URL):gsub(format.escapemagiccharacters(thisline), " ")
477 478 479 480
			if instcnt ~= 0 then
				-- logme("instcnt =  "..instcnt)
				isbad=1
				wrdcnt= wrdcnt + instcnt
481
				badwordloc[#badwordloc+1] = thisline
482
			end
483
		end
Ted Trask's avatar
Ted Trask committed
484

485
		--check for DansGuardian actions
486
		if (logentry.reason and logentry.reason ~= "") then
Ted Trask's avatar
Ted Trask committed
487 488 489 490 491 492 493 494 495 496
			if string.find(logentry.reason,"DENIED") then
				-- logme("*Denied*")
				logentry.deniedyesno=1
			elseif string.find(logentry.URL,"GBYPASS") then
				-- logme("GBYPASS")
				logentry.bypassyesno=1
			elseif string.find(logentry.reason,"OVERRIDE") then
				-- logme("*OVERRIDE*")
				logentry.bypassyesno=1
			end
497
		end
498 499 500

		--check for Squark actions
		if (logentry.squarkaction and logentry.squarkaction ~= "") then
Ted Trask's avatar
Ted Trask committed
501 502 503 504 505 506
			--logme("squarkaction="..logentry.squarkaction)
			if string.find(logentry.squarkaction, "blocked") then
				logentry.deniedyesno=1
			elseif string.find(logentry.squarkaction,"overridden") then
				logentry.bypassyesno=1
			end
507 508
		end

509 510 511 512 513 514
		--check for Squark category
		if (logentry.squarkcategory and logentry.squarkcategory ~= "") then
			logentry.reason = logentry.squarkcategory
			logentry.shortreason = logentry.squarkcategory
		end

515 516 517 518
		for i,goodline in ipairs(goodwords) do
			if not goodline then
				break
			end
519
			_,instcnt = string.lower(logentry.URL):gsub(format.escapemagiccharacters(goodline), " ")
520 521
			--if string.find(logentry.URL,goodline) then
			if instcnt ~= 0 then
522
				if wrdcnt >= instcnt then
523
					wrdcnt = wrdcnt - instcnt
524 525
				else
					wrdcnt = 0
526
				end
527
				goodwordloc[#goodwordloc+1] = goodline
528
			end
529 530
		end
	end
Ted Trask's avatar
Ted Trask committed
531 532 533 534
	-- Reset bad to reduce number of bad hits if score is zero
	-- if wrdcnt == 0 then
	-- isbad=0
	-- end
535 536 537

	logentry.score=wrdcnt
	logentry.badyesno=isbad
538 539
	logentry.wordloc=table.concat(badwordloc,"|")
	logentry.gwordloc=table.concat(goodwordloc,"|")
540 541 542 543
end

local function parsesquidlog(line)
	-- Format of squid log (space separated):
544
	-- time elapsed remotehost code/status bytes method URL rfc931 peerstatus/peerhost ? squarkcategory/squarkaction
545 546 547 548 549 550
	local words = {}

	for word in string.gmatch(line, "%S+") do
		words[#words+1] = word
	end

551 552 553
	local logentry = {logdatetime=words[1],
		elapsed=words[2],
		clientip=words[3],
Ted Trask's avatar
Ted Trask committed
554 555
		code=string.match(words[4] or "", "^[^/]*"),
		status=string.match(words[4] or "", "[^/]*$"),
556
		bytes=words[5],
Ted Trask's avatar
Ted Trask committed
557
		method=words[6],
558
		URL=words[7],
559
		clientuserid=words[8],
Ted Trask's avatar
Ted Trask committed
560
		peerstatus=string.match(words[9] or "", "^[^/]*"),
561 562 563
		peerhost=string.match(words[9] or "", "[^/]*$"),
		squarkcategory=string.match(words[11] or "", "^[^,]*"),
		squarkaction=string.match(words[11] or "", "[^,]*$")}
Ted Trask's avatar
Ted Trask committed
564

565
	checkwords(logentry)
566

Natanael Copa's avatar
Natanael Copa committed
567 568 569 570 571 572
	-- Don't care about TCP_DENIED so apps like dropbox, nokia connection
	-- suite does not flood our logs.
	if logentry.code == nil or logentry.code == "TCP_DENIED" then
		return nil
	end

Ted Trask's avatar
Ted Trask committed
573 574
	-- Don't care about local requests (from DG) (this check also removes blank lines)
	if logentry.clientip and logentry.clientip ~= "127.0.0.1" then
575 576
		logentry.logdatetime = os.date("%Y-%m-%d %H:%M:%S", logentry.logdatetime)..string.match(logentry.logdatetime, "%..*")
		return logentry
Ted Trask's avatar
Ted Trask committed
577 578 579 580
	end
	return nil
end

581
local function parsedglog(line)
582
	local words = format.string_to_table(line, "\t")
583 584 585 586 587 588 589
	local logentry = {logdatetime=words[1],
			clientuserid=words[2],
			clientip=words[3],
			URL=words[4],
			reason=words[5],
			method=words[6],
			bytes=words[7],
590 591
			shortreason=words[9],
			deniedyesno=1}
Ted Trask's avatar
Ted Trask committed
592

593
	checkwords(logentry)
Ted Trask's avatar
Ted Trask committed
594 595 596 597 598

	if logentry.reason and logentry.reason ~= "" then
           	if logentry.shortreason == "" then
           		logentry.shortreason = logentry.reason
           	end
599
		--logentry.score = string.match(logentry.reason, "^.*: ([0-9]+) ")
600
		logentry.logdatetime = string.gsub(logentry.logdatetime, "%.", "-")
Ted Trask's avatar
Ted Trask committed
601 602 603 604 605
	   	return logentry
	end
	return nil
end

Ted Trask's avatar
Ted Trask committed
606 607 608 609 610 611 612 613 614
-- ################################################################################
-- DOWNLOAD FILE FUNCTIONS

-- must do apk_add wget first

local connecttosource = function(source, cookiesfile)
	local success = false
	logme("Connecting to source "..source.sourcename)
	if source.method == "http" or source.method == "https" then
615 616
		fs.write_file(cookiesfile, "password="..source.passwd.."&userid="..source.userid.."&Logon=Logon&submit=true")
		local resultpage, errtxt = modelfunctions.run_executable({"wget", "-O", "-", "--no-check-certificate", "--save-cookies", cookiesfile, "--keep-session-cookies", "--post-file", cookiesfile, source.method.."://"..source.source.."/cgi-bin/acf/acf-util/logon/logon"})
Ted Trask's avatar
Ted Trask committed
617 618
		if resultpage == "" then
			logme("Failed to connect to "..source.sourcename)
619
		elseif string.find(resultpage, "/acf%-util/logon/logon") then
620
			logme("Failed to log on to "..source.sourcename)
Ted Trask's avatar
Ted Trask committed
621 622 623
		else
			success = true
		end
624 625
	elseif source.method == "local" then
		success = true
Ted Trask's avatar
Ted Trask committed
626 627 628 629 630 631 632
	end
	return success
end

local getlogcandidates = function(source, cookiesfile)
	local candidates = {}
	if source.method == "http" or source.method == "https" then
633 634
		local resultpage = modelfunctions.run_executable({"wget", "-O", "-", "--no-check-certificate", "--load-cookies", cookiesfile, source.method.."://"..source.source.."/cgi-bin/acf/alpine-baselayout/logfiles/status"})
		for file in string.gmatch(resultpage, "download%?[^\"]*name=([^\"]+)") do
Ted Trask's avatar
Ted Trask committed
635 636
			candidates[#candidates+1] = file
		end
637 638
	elseif source.method == "local" then
		candidates = fs.find_files_as_array(nil, source.source)
Ted Trask's avatar
Ted Trask committed
639 640 641 642
	end
	return candidates
end

643
local openlogfile = function(source, cookiesfile, logfile)
644
	local handle, handle2, errtxt
Ted Trask's avatar
Ted Trask committed
645
	if source.method == "http" or source.method == "https" then
646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678
		local cmd = {"wget", "-O", "-", "--no-check-certificate", "--load-cookies", cookiesfile, "--post-data", "submit=true&viewtype=stream&name="..logfile.."&filename="..logfile, source.method.."://"..source.source.."/cgi-bin/acf/alpine-baselayout/logfiles/download"}
		cmd.stdin = "/dev/null"
		cmd.stdout = subprocess.PIPE
		cmd.stderr = "/dev/null"
		local res, err = pcall(function()
	                -- For security, set the path
	                posix.setenv("PATH", "/usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/sbin:/sbin")
	                local proc, errmsg, errno = subprocess.popen(cmd)
			if proc then
				if string.find(logfile, "%.gz$") then
					local cmd2 = {"gunzip", "-c"}
					-- Pipe the output of wget into gunzip
					cmd2.stdin = proc.stdout
					cmd2.stdout = subprocess.PIPE
					cmd2.stderr = "/dev/null"
	                		local proc2, errmsg2, errno2 = subprocess.popen(cmd2)
					if proc2 then
						handle = proc2.stdout
						handle2 = proc.stdout
					else
						proc.stdout:close()
						errtxt = errmsg2 or "Unknown failure"
					end
				else
					handle = proc.stdout
				end
				
			else
				errtxt = errmsg or "Unknown failure"
			end
		end)
		if not res or err then
			errtxt = err or "Unknown failure"
679
		end
680
	elseif source.method == "local" then
681
		if string.find(logfile, "%.gz$") then
682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698
			local res, err = pcall(function()
	                	-- For security, set the path
		                posix.setenv("PATH", "/usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/sbin:/sbin")
				local cmd = {"gunzip", "-c", logfile}
				cmd.stdin = "/dev/null"
				cmd.stdout = subprocess.PIPE
				cmd.stderr = "/dev/null"
	               		local proc, errmsg, errno = subprocess.popen(cmd)
				if proc then
					handle = proc.stdout
				else
					errtxt = errmsg or "Unknown failure"
				end
			end)
			if not res or err then
				errtxt = err or "Unknown failure"
			end
699
		else
700
			handle = io.open(logfile)
701
		end
Ted Trask's avatar
Ted Trask committed
702
	end
703
	return handle, handle2
Ted Trask's avatar
Ted Trask committed
704 705 706 707
end

local deletelogfile = function(source, cookiesfile, logfile)
	if source.method == "http" or source.method == "https" then
708
		modelfunctions.run_executable({"wget", "-O", "-", "--no-check-certificate", "--load-cookies", cookiesfile, "--post-data", "submit=true&name="..logfile.."&filename="..logfile, source.method.."://"..source.source.."/cgi-bin/acf/alpine-baselayout/logfiles/delete"})
709 710
	elseif source.method == "local" then
		os.remove(logfile)
Ted Trask's avatar
Ted Trask committed
711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734
	end
end

-- ################################################################################
-- PUBLIC FUNCTIONS

function getsourcelist()
	local retval = cfe({ type="list", value={}, label="Weblog Source List" })
	local res, err = pcall(function()
		databaseconnect(DatabaseUser)
		retval.value = listsourceentries()
		databasedisconnect()
	end)
	if not res then
		retval.errtxt = err
	end

	return retval
end

function getsource(sourcename)
	local sourcedata = getnewsource()
	sourcedata.value.sourcename.value = sourcename
	sourcedata.value.sourcename.errtxt = "Source name does not exist"
735
	sourcedata.value.sourcename.readonly = true
Ted Trask's avatar
Ted Trask committed
736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761

	local res, err = pcall(function()
		databaseconnect(DatabaseUser)
		local sourcelist = listsourceentries()
		databasedisconnect()
		for i,source in ipairs(sourcelist) do
			if source.sourcename == sourcename then
				sourcedata.value.sourcename.errtxt = nil
				for name,val in pairs(source) do
					if sourcedata.value[name] then
						sourcedata.value[name].value = val
					end
				end
				break
			end
		end
	end)
	if not res then
		sourcedata.errtxt = err
	end

	return sourcedata
end

local validatesource = function(sourcedata)
	local success = modelfunctions.validateselect(sourcedata.value.method)
762 763 764 765 766 767
	local test = {"sourcename", "source"}
	if sourcedata.value.method.value ~= "local" then
		test[#test+1] = "userid"
		test[#test+1] = "passwd"
	end
	for i,name in ipairs(test) do
Ted Trask's avatar
Ted Trask committed
768 769 770 771 772 773 774 775 776
		if sourcedata.value[name].value == "" then
			sourcedata.value[name].errtxt = "Cannot be empty"
			success = false
		end
	end

	return success
end

777
function updatesource(self, sourcedata)
Ted Trask's avatar
Ted Trask committed
778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801
	local success = validatesource(sourcedata)
	sourcedata.errtxt = "Failed to update source"
	if success then
		local source = {}
		for name,val in pairs(sourcedata.value) do
			source[name] = val.value
		end

		local res, err = pcall(function()
			databaseconnect(DatabaseUser)
			sourcedata.descr = updatesourceentry(source)
			databasedisconnect()
			sourcedata.errtxt = nil
		end)
		if not res and err then
			sourcedata.errtxt = sourcedata.errtxt .. "\n" .. err
		end
	end

	return sourcedata
end

function getnewsource()
	local source = {}
802 803 804 805 806 807 808
	source.sourcename = cfe({ label="Source Name", seq=0 })
	source.method = cfe({ type="select", value="local", label="Method", option={"http", "https", "local"}, seq=3 })
	source.userid = cfe({ label="UserID", seq=4 })
	source.passwd = cfe({ type="password", label="Password", seq=5 })
	source.source = cfe({ value="/var/log", label="Source Location / Address", seq=2 })
	source.tzislocal = cfe({ type="boolean", value=false, label="Using local timezone", seq=6 })
	source.enabled = cfe({ type="boolean", value=false, label="Enabled", seq=1 })
Ted Trask's avatar
Ted Trask committed
809 810 811
	return cfe({ type="group", value=source, label="Source" })
end

812
function createsource(self, sourcedata)
Ted Trask's avatar
Ted Trask committed
813 814 815 816 817 818 819
	local success = validatesource(sourcedata)
	sourcedata.errtxt = "Failed to create source"
	if success then
		local source = {}
		for name,val in pairs(sourcedata.value) do
			source[name] = val.value
		end
820 821
		-- remove spaces from sourcename
		source.sourcename = string.gsub(source.sourcename, "%s+$", "")
Ted Trask's avatar
Ted Trask committed
822 823 824 825 826 827 828 829 830 831 832 833 834 835 836

		local res, err = pcall(function()
			databaseconnect(DatabaseUser)
			sourcedata.descr = importsourceentry(source)
			databasedisconnect()
			sourcedata.errtxt = nil
		end)
		if not res and err then
			sourcedata.errtxt = sourcedata.errtxt .. "\n" .. err
		end
	end

	return sourcedata
end

837 838 839 840 841 842 843 844
function getdeletesource(self, clientdata)
	local retval = {}
	retval.sourcename = cfe({ value=clientdata.sourcename or "", label="Source Name" })
	return cfe({ type="group", value=retval, label="Delete Source" })
end

function deletesource(self, delsource)
	delsource.errtxt="Failed to delete source"
Ted Trask's avatar
Ted Trask committed
845 846
	local res, err = pcall(function()
		databaseconnect(DatabaseUser)
847
		local number = deletesourceentry(delsource.value.sourcename.value)
Ted Trask's avatar
Ted Trask committed
848 849
		databasedisconnect()
		if number > 0 then
850 851 852
			delsource.errtxt = nil
		else
			delsource.value.sourcename.errtxt = "Failed to find source"
Ted Trask's avatar
Ted Trask committed
853 854 855
		end
	end)
	if not res and err then
856
		delsource.errtxt = delsource.errtxt .. "\n" .. err
Ted Trask's avatar
Ted Trask committed
857
	end
858
	return delsource
Ted Trask's avatar
Ted Trask committed
859 860
end

861 862 863 864 865
function gettestsource(self, clientdata)
	local retval = {}
	retval.sourcename = cfe({ value=clientdata.sourcename or "", label="Source Name" })
	return cfe({ type="group", value=retval, label="Test Source" })
end
Ted Trask's avatar
Ted Trask committed
866

867
function testsource(self, test)
Ted Trask's avatar
Ted Trask committed
868
	-- temporary override of logme function to capture messages to result.value
869
	test.descr = {}
Ted Trask's avatar
Ted Trask committed
870
	local temp = logme
Ted Trask's avatar
Ted Trask committed
871
	logme = function(message) table.insert(test.descr, message) end
872 873

	test.errtxt = "Test Failed"
Ted Trask's avatar
Ted Trask committed
874 875 876 877

	local cookiesfile = "/tmp/cookies-"..tostring(os.time())
	local res, err = pcall(function()
		databaseconnect(DatabaseUser)
878
		local sources = listsourceentries(test.value.sourcename.value)
Ted Trask's avatar
Ted Trask committed
879 880
		databasedisconnect()
		if #sources < 1 then
881
			test.value.sourcename.errtxt = "Failed to find source"
Ted Trask's avatar
Ted Trask committed
882 883 884 885 886
		else
			local source = sources[1]
			-- run the test
			if connecttosource(source, cookiesfile) then
				local files = getlogcandidates(source, cookiesfile)
887
				test.errtxt = nil
Ted Trask's avatar
Ted Trask committed
888 889 890 891 892 893 894 895 896 897
				if #files == 0 then
					logme("No log files found")
				else
					for i,file in ipairs(files) do
						logme("Found log file "..file)
					end
				end
			end
		end
	end)
898 899
	if err and not res then
		test.errtxt = test.errtxt .. "\n" .. err
Ted Trask's avatar
Ted Trask committed
900 901 902 903
	end
	os.remove(cookiesfile)

	-- fix the result
904
	test.descr = table.concat(test.descr, "\n") or ""
Ted Trask's avatar
Ted Trask committed
905 906
	logme = temp

907
	return test
Ted Trask's avatar
Ted Trask committed
908 909
end

910 911
-- import a logfile and delete logfile after
local function importlogfile(source, cookiesfile, file, parselog_func)
Ted Trask's avatar
Ted Trask committed
912
	logme("Getting " .. file )
913 914 915 916 917
	local loghandle, loghandle2 = openlogfile(source, cookiesfile, file)
	if not loghandle then
		logme("Failed to get " .. file )
		return
	end
Ted Trask's avatar
Ted Trask committed
918
	logme("Processing " .. file )
919 920 921
	local res, err = pcall(function()
		con:execute("START TRANSACTION")
		for line in loghandle:lines() do
922 923 924
			assert(con:execute("SAVEPOINT before_line"))
			local res2, err2 = pcall(function()
				local logentry = parselog_func(line)
925
				importlogentry(logentry, source.sourcename)
926 927 928 929 930 931
			end)
			if not res2 then
				if (config.stoponerror == "true") then
					pcall(function() con:execute("ROLLBACK") end)
				else
					assert(con:execute("ROLLBACK TO before_line"))
932
					con:execute("COMMIT")
933 934 935
				end
				pcall(function() logme("Exception on line:"..line) end)
				if err2 then
Ted Trask's avatar
Ted Trask committed
936
					pcall(function() logme(err2) end)
937 938 939
				end
				if (config.stoponerror == "true") then
					assert(res2, "Import halted on exception")
940 941
				else
					con:execute("START TRANSACTION")
942
				end
943 944
			else
				assert(con:execute("RELEASE SAVEPOINT before_line"))
945
			end
946 947 948 949 950 951 952 953
		end
		con:execute("COMMIT")
	end)
	if not res then
		pcall(function() con:execute("ROLLBACK") end)
		if err then
			pcall(function() logme(err) end)
		end
954
	end
955
	loghandle:close()
956 957 958
	if loghandle2 then
		loghandle2:close()
	end
959 960 961 962 963
	if res then
		logme("Deleting " .. file )
		deletelogfile(source, cookiesfile, file)
	end
	return res
964 965
end

966 967 968 969 970 971
function getimportlogs(self, clientdata)
	local retval = {}
	return cfe({ type="group", value=retval, label="Import Logs" })
end

function importlogs(self, import)
Ted Trask's avatar
Ted Trask committed
972
	local count = 0
973
	local success = true
Ted Trask's avatar
Ted Trask committed
974 975 976 977 978

	local res, err = pcall(function()
		databaseconnect(DatabaseOwner, config.password)

		-- Download, parse, and import the logs
979 980 981 982
		logme("Executing importlogs")
		logme("Analyzing...")
		local sql = "ANALYZE"
		res = assert (con:execute(sql))
983 984 985 986 987 988

		-- Get the word lists
		goodwords = fs.read_file_as_array(goodwordslist) or {}
		badwords = fs.read_file_as_array(badwordslist) or {}
		ignorewords = fs.read_file_as_array(ignorewordslist) or {}

989 990
		-- Determine sources
		local sources = listsourceentries(sourcename)
Ted Trask's avatar
Ted Trask committed
991

Ted Trask's avatar
Ted Trask committed
992 993 994 995 996 997 998
		for i,source in ipairs(sources) do
			if source.enabled then
				logme("Getting logs from source " .. source.sourcename)
				local cookiesfile = "/tmp/cookies-"..tostring(os.time())
				if connecttosource(source, cookiesfile) then
					local files = getlogcandidates(source, cookiesfile)
					for j,file in ipairs(files) do
999
						if string.match(file, "dansguardian/access%.log[%.%-]") then
Ted Trask's avatar
Ted Trask committed
1000
							count = count + 1
1001
							success = importlogfile(source, cookiesfile, file, parsedglog) and success
Ted Trask's avatar
Ted Trask committed
1002
						elseif string.match(file, "squid/access%.log[%.%-]") then
Ted Trask's avatar
Ted Trask committed
1003
							count = count + 1
1004
							success = importlogfile(source, cookiesfile, file, parsesquidlog) and success
Ted Trask's avatar
Ted Trask committed
1005 1006 1007 1008 1009 1010 1011 1012
						end
					end
				end
				os.remove(cookiesfile)
			end
		end

		-- Process the logs
1013 1014
		if success then
			updateusagestats()
1015
			importpubweblog()
1016
		end
Ted Trask's avatar
Ted Trask committed
1017 1018 1019 1020 1021 1022 1023
		-- Purge old database entries
		groomusagestat()
		groomdbhistlog()
		groompublogs()

		databasedisconnect()
	end)
1024
	if not res or not success then
1025
		import.errtxt = "Import Logs Failure"
Ted Trask's avatar
Ted Trask committed
1026 1027
		if err then
			pcall(function() logme(err) end)
1028
			import.errtxt = import.errtxt .. "\n" .. err
Ted Trask's avatar
Ted Trask committed
1029
		end
1030
		pcall(function() databasedisconnect() end)
1031 1032
	else
		import.descr = "Imported "..tostring(count).." logs"
Ted Trask's avatar
Ted Trask committed
1033 1034
	end

1035
	return import
Ted Trask's avatar
Ted Trask committed
1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051
end

function getactivitylog()
	local retval = cfe({ type="list", value={}, label="Weblog Activity Log" })
	local res, err = pcall(function()
		databaseconnect(DatabaseUser)
		retval.value = listhistorylogentries() or {}
		databasedisconnect()
	end)
	if not res then
		retval.errtxt = err
	end

	return retval
end

Ted Trask's avatar
Ted Trask committed
1052 1053 1054 1055 1056 1057 1058
function geteditselected()
	local result = {}
	result.select = cfe({ type="list", value={}, label="Entries to mark as selected" })
	result.deselect = cfe({ type="list", value={}, label="Entries to mark as selected" })
	return cfe({ type="group", value=result, label="Select / Deselect log entries" })
end

1059
function editselected(self, data)
Ted Trask's avatar
Ted Trask committed
1060 1061 1062 1063
	local res, err = pcall(function()
		databaseconnect(DatabaseOwner)
		con:execute("START TRANSACTION")
		for i,sel in ipairs(data.value.select.value) do
1064
			assert (con:execute("UPDATE pubweblog SET selected = true WHERE id = '"..escape(sel).."'"))
Ted Trask's avatar
Ted Trask committed
1065
		end
Ted Trask's avatar
Ted Trask committed
1066
		for i,sel in ipairs(data.value.deselect.value) do
1067
			assert (con:execute("UPDATE pubweblog SET selected = false WHERE id = '"..escape(sel).."'"))
Ted Trask's avatar
Ted Trask committed
1068 1069 1070 1071 1072 1073
		end
		con:execute("COMMIT")
		databasedisconnect()
	end)
	if not res then
		data.errtxt = err
Ted Trask's avatar
Ted Trask committed
1074
	end
Ted Trask's avatar
Ted Trask committed
1075
	return data
1076 1077
end

1078 1079 1080 1081 1082 1083 1084
function getclearselected(self, clientdata)
	local retval = {}
	return cfe({ type="group", value=retval, label="Clear select fields" })
end

function clearselected(self, clear)
	clear.errtxt = "Failed to clear select fields"
1085 1086 1087 1088 1089
	local res, err = pcall(function()
		sql = "UPDATE pubweblog SET selected = false WHERE selected = true"
		databaseconnect(DatabaseOwner)
		assert (con:execute(sql))
		databasedisconnect()
1090
		clear.errtxt = nil
1091 1092
	end)
	if not res then
1093
		clear.errtxt = clear.errtxt.."\n"..err
1094
	end
1095
	return clear
1096 1097
end

1098 1099 1100 1101 1102 1103
local validateweblogparameters = function(params)
	local success = modelfunctions.validateselect(params.value.activelog)
	success = modelfunctions.validateselect(params.value.sortby) and success
	if params.value.clientip.value ~= "" and string.find(params.value.clientip.value, "[^%d%.]") then
		params.value.clientip.errtxt = "Invalid IP Address"
		success = false
Ted Trask's avatar
Ted Trask committed
1104
	end
1105 1106 1107 1108
	if not validator.is_integer(params.value.score.value) then
		params.value.score.errtxt = "Must be a number"
		success = false
	end
1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130
	local res, err = pcall(function()
		databaseconnect(DatabaseUser)
		local s
		if params.value.starttime.value ~= "" then
			s,params.value.starttime.errtxt,params.value.starttime.value = convertdatabaseentry("TIMESTAMP", params.value.starttime.value)
			success = success and s
		end
		if params.value.endtime.value ~= "" then
			s,params.value.endtime.errtxt,params.value.endtime.value = convertdatabaseentry("TIMESTAMP", params.value.endtime.value)
			success = success and s
		end
		if params.value.focus.value ~= "" then
			s,params.value.focus.errtxt,params.value.focus.value = convertdatabaseentry("TIMESTAMP", params.value.focus.value)
			success = success and s
		end
		databasedisconnect()
	end)
	if not res and err then
		params.value.starttime.errtxt = err
		params.value.endtime.errtxt = err
		params.value.focus.errtxt = err
		success = false
1131
	end
1132 1133
	return success
end
Ted Trask's avatar
Ted Trask committed
1134

1135 1136 1137 1138 1139 1140 1141 1142 1143
function getweblogparameters()
	local c = getconfig()
	local result = {}
	result.activelog = cfe({ type="select", value="pubweblog", option={"pubweblog", "pubweblog_history"}, label="Active Weblog", seq=1 })
	result.starttime = cfe({ value=c.value.auditstart.value, label="Start Time", seq=2 })
	result.endtime = cfe({ value=c.value.auditend.value, label="End Time", seq=3 })
	result.clientuserid = cfe({ value="", label="User ID", seq=4 })
	result.clientip = cfe({ value="", label="Client IP", seq=5 })
	result.urisearch = cfe({ value="", label="URI Contains", descr="Retrieve records where the URI contains this word", seq=6 })
1144
	result.score = cfe({ value=c.value.score.value, label="Minimum Score", descr="Minimum score to search on", seq=7 })
1145 1146 1147 1148 1149 1150 1151 1152
	result.sortby = cfe({ type="select", value=c.value.sortby.value, option=c.value.sortby.option, label="Sort By field", descr="Sort by this field when displaying records", seq=8 })
	result.badyesno = cfe({ type="boolean", value=c.value.badyesno.value, label="Show Suspect Records", descr="Limit search to records marked as suspect", seq=9 })
	result.deniedyesno = cfe({ type="boolean", value=c.value.deniedyesno.value, label="Show Denied Records", descr="Limit search to Denied URIs", seq=10 })
	result.bypassyesno = cfe({ type="boolean", value=c.value.bypassyesno.value, label="Show Bypass Records", descr="Limit search to Bypass attempts", seq=11 })
	result.selected = cfe({ type="boolean", value=false, label="Show Selected Records", descr="Limit search to records that have been selected", seq=12 })
	result.focus = cfe({ value="", label="Focus Time", seq=13 })
	return cfe({ type="group", value=result, label="Weblog Access Log" })
end
Ted Trask's avatar
Ted Trask committed
1153

1154
function getweblog(self, result)
1155 1156 1157 1158
	local success = validateweblogparameters(result)
	result.value.log = cfe({ type="list", value={}, label="Weblog Access Log" })
	result.value.window = cfe({ value=config.window or "5", label="Time Window" })
	local err
1159 1160 1161
	if success then
		local res, err = pcall(function()
			databaseconnect(DatabaseUser)
1162
			result.value.log.value = listlogentries(result.value.activelog.value, result.value.clientuserid.value, result.value.starttime.value, result.value.endtime.value, result.value.clientip.value, result.value.badyesno.value, result.value.deniedyesno.value, result.value.bypassyesno.value, result.value.score.value, result.value.urisearch.value, result.value.sortby.value, result.value.selected.value ) or {}
1163 1164
			databasedisconnect()
		end)
1165 1166 1167
		if not res then
			result.errtxt = err
		end
1168
	else
1169
		result.errtxt = "Invalid search parameters"
1170
	end
1171
	return result
Ted Trask's avatar
Ted Trask committed
1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
end

function getusagestats()
	local retval = cfe({ type="list", value={}, label="Weblog Usage Stats" })
	local res, err = pcall(function()
		databaseconnect(DatabaseUser)
		retval.value = listusagestats() or {}
		databasedisconnect()
	end)
	if not res then
		retval.errtxt = err
	end

	return retval
end

1188 1189 1190 1191 1192 1193 1194 1195 1196
function getauditstats()
	local result = {}
	result.auditstart = cfe({ value=config.auditstart or "", label="Audit Start Time" })
	result.auditend = cfe({ value=config.auditend or "", label="Audit End Time" })
	result.groupby = cfe({ value=config.groupby or "clientuserid", label="Group By" })
	result.stats = cfe({ type="list", value={}, label="Audit Block Statistics" })
	local res, err = pcall(function()
		if config.auditstart ~= "" and config.auditend ~= "" then
			databaseconnect(DatabaseUser)
1197
			result.stats.value = groupflaggedlogentries(config.auditstart, config.auditend, result.groupby.value) or {}
1198 1199 1200 1201 1202 1203
			databasedisconnect()
		end
	end)
	return cfe({ type="group", value=result, errtxt=err, label="Weblog Audit Statistics" })
end

1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214
function getcompleteaudit(self, clientdata)
	local retval = {}
	retval.timestamp = cfe({ value=clientdata.timestamp or "", label="New Audit End Time" })
	return cfe({ type="group", value=retval, label="Complete Audit" })
end

function completeaudit(self, complete)
	if "" == complete.value.timestamp.value then
		local now = os.time()
		complete.value.timestamp.value = os.date("%Y-%m-%d %H:%M:%S", now - now%86400 - 86400)
	end
1215 1216
	local conf = getconfig()
	conf.value.auditstart.value = conf.value.auditend.value
1217
	conf.value.auditend.value = complete.value.timestamp.value
1218
	conf = updateconfig(self, conf)
1219
	if conf.errtxt then
1220
		complete.errtxt = "Failed to complete audit\n"..conf.errtxt.."\n"..conf.value.auditend.errtxt
1221
	end
1222
	return complete
1223
end
Ted Trask's avatar
Ted Trask committed
1224 1225 1226

function getconfig()
	local result = {}
Ted Trask's avatar
Ted Trask committed
1227 1228
	result.auditstart = cfe({ value=config.auditstart or "", label="Audit Start Time", seq=1 })
	result.auditend = cfe({ value=config.auditend or "", label="Audit End Time", seq=2 })
1229
	result.groupby = cfe({ type="select", value=config.groupby or "clientuserid", label="Group results by", option={"clientuserid", "clientip"}, descr="Display audit results based on user ID or IP", seq=3 })
1230
	result.score = cfe({ value=config.score or "0", label="Minimum Score", descr="Default minimum Score to search for", seq=4 })
1231 1232 1233 1234 1235 1236 1237 1238 1239
	result.sortby = cfe({ type="select", value=config.sortby or "logdatetime", label="Sort By field", option={"logdatetime", "logdatetime DESC", "clientuserid", "clientuserid DESC", "clientip", "clientip DESC", "bytes", "bytes DESC", "score", "score DESC", "reason"}, descr="Default sort order", seq=5 })
	result.badyesno = cfe({ type="boolean", value=(config.badyesno == "true"), label="Display Suspect Records", descr="By default, only show records flagged as suspect", seq=6 })
	result.deniedyesno = cfe({ type="boolean", value=(config.deniedyesno == "true"), label="Display Denied Records", descr="By default, only show records with denied URI", seq=7 })
	result.bypassyesno = cfe({ type="boolean", value=(config.bypassyesno == "true"), label="Display Bypass Records", descr="By default, only show records with bypass attempts", seq=8 })
	result.shorturi = cfe({ type="boolean", value=(config.shorturi == "true"), label="Truncate URLs", descr="You can limit the length of displayed URLs by enabling this option", seq=9 })
	result.shortreason = cfe({ type="boolean", value=(config.shortreason == "true"), label="Short Reason", descr="Display a short reason (dansguardian only)", seq=10 })
	result.window = cfe({ value=config.window or "5", label="Time Window", descr="Minutes of activity to display before and after selected block", seq=11 })
	result.purgedays = cfe({ value=config.purgedays or "30", label="Days before Purge", descr="Days to keep full history, regardless of audit", seq=12 })
	result.stoponerror = cfe({ type="boolean", value=(config.stoponerror == "true"), label="Stop on Error", descr="Stop import of logs if an error is encountered", seq=13})
Ted Trask's avatar
Ted Trask committed
1240 1241 1242
	return cfe({ type="group", value=result, label="Weblog Config" })
end

1243
local function validateconfig(newconfig)
Ted Trask's avatar
Ted Trask committed
1244 1245
	local success = modelfunctions.validateselect(newconfig.value.groupby)
	success = modelfunctions.validateselect(newconfig.value.sortby) and success
1246 1247 1248 1249
	if not validator.is_integer(newconfig.value.score.value) then
		newconfig.value.score.errtxt = "Must be a number"
		success = false
	end
Ted Trask's avatar
Ted Trask committed
1250 1251 1252 1253 1254 1255
	if newconfig.value.window.value == "" then
		newconfig.value.window.errtxt = "Cannot be blank"
		success = false
	elseif not validator.is_integer(newconfig.value.window.value) then
		newconfig.value.window.errtxt = "Must be a number"
		success = false
Timo Teräs's avatar
Timo Teräs committed
1256
	end
Ted Trask's avatar
Ted Trask committed
1257 1258 1259
	if not validator.is_integer(newconfig.value.purgedays.value) then
		newconfig.value.purgedays.errtxt = "Must be a number"
		success = false
Timo Teräs's avatar
Timo Teräs committed
1260
	end
1261 1262 1263 1264
	local res, err = pcall(function()
		databaseconnect(DatabaseUser)
		local s
		if newconfig.value.auditstart.value ~= "" then
1265
			s,newconfig.value.auditstart.errtxt,newconfig.value.auditstart.value = convertdatabaseentry("TIMESTAMP", newconfig.value.auditstart.value)
1266 1267 1268
			success = success and s
		end
		if newconfig.value.auditend.value ~= "" then
1269
			s,newconfig.value.auditend.errtxt,newconfig.value.auditend.value = convertdatabaseentry("TIMESTAMP", newconfig.value.auditend.value)
1270 1271 1272 1273 1274 1275 1276 1277 1278
			success = success and s
		end
		databasedisconnect()
	end)
	if not res and err then
		newconfig.value.auditstart.errtxt = err
		newconfig.value.auditend.errtxt = err
		success = false
	end
1279
	return success, newconfig
1280 1281
end

1282
function updateconfig(self, newconfig)
1283
	local success = validateconfig(newconfig)
Ted Trask's avatar
Ted Trask committed
1284 1285