Index: trunk/wikistats/squids/SquidCountArchiveReadInput.pm |
— | — | @@ -0,0 +1,353 @@ |
| 2 | + #!/usr/bin/perl |
| 3 | + |
| 4 | +sub CollectFilesToProcess |
| 5 | +{ |
| 6 | + trace CollectFilesToProcess ; |
| 7 | + |
| 8 | + if (! $job_runs_on_production_server) |
| 9 | + { |
| 10 | + push @files, $file_test ; |
| 11 | + return $true ; |
| 12 | + } |
| 13 | + |
| 14 | + my ($days_ago, $date_collect_files, $time_to_start, $time_to_stop, $path_out, $path_out_month) = @_ ; |
| 15 | + |
| 16 | + print "Collect files for date $date_collect_files: files with timestamps between $time_to_start and $time_to_stop\n\n" ; |
| 17 | + |
| 18 | + my $all_files_found = $true ; |
| 19 | + |
| 20 | + my ($date_archived) ; |
| 21 | + |
| 22 | + $dir_in = "/a/squid/archive" ; |
| 23 | + |
| 24 | + $some_files_found = $false ; |
| 25 | + $full_range_found = $false ; |
| 26 | + |
| 27 | + $path_head_tail = "$path_out_month/$file_head_tail" ; |
| 28 | + |
| 29 | + # file naming scheme on server: sampled-1000.log-yyyymmdd, does not mean on that day file sampled-1000.log was archived |
| 30 | + # file can contain data for days(s) before and day (days?) after yyyymmdd, see e.g. sampled-10000.log-20090802 (days 0801-0803) |
| 31 | + # this is confusing so start a few days earlier and check for each day: |
| 32 | + # whether a file exists and whether it's 'head' and or 'tail' time (first last record) fall within range |
| 33 | + |
| 34 | + # find first and last file to process, meaning all files that comprise log records within date range |
| 35 | + |
| 36 | + $head_found = $false ; |
| 37 | + $tail_found = $false ; |
| 38 | + |
| 39 | + for ($days_ago_inspect = $days_ago + 2 ; $days_ago_inspect >= $days_ago - 5 ; $days_ago_inspect--) |
| 40 | + { |
| 41 | + next if $days_ago_inspect < 0 ; # days ago can't be negative |
| 42 | + |
| 43 | + ($sec,$min,$hour,$day,$month,$year) = localtime ($time_start - $days_ago_inspect * 24 * 3600) ; |
| 44 | + $date_archived = sprintf ("%4d%02d%02d", $year+1900, $month+1, $day) ; |
| 45 | + print "\n- Inspect file saved $days_ago_inspect days ago: sampled-1000.log-$date_archived.gz\n" ; |
| 46 | + |
| 47 | + my $file = "$dir_in/sampled-1000.log-$date_archived.gz" ; |
| 48 | + |
| 49 | + if (! -e $file) |
| 50 | + { print "- File not found: $file\n" ; } |
| 51 | + else |
| 52 | + { |
| 53 | + ($timehead,$timetail) = &GetLogRange ($file, $path_head_tail) ; |
| 54 | + |
| 55 | + if (($timetail ge $time_to_start) && ($timehead le $time_to_stop)) |
| 56 | + { |
| 57 | + print "- Include this file\n" ; |
| 58 | + |
| 59 | + $some_files_found = $true ; |
| 60 | + push @files, $file ; |
| 61 | + if ($timehead le $time_to_start) { $head_found = $true ; print "- Head found\n" ; } |
| 62 | + if ($timetail ge $time_to_stop) { $tail_found = $true ; print "- Tail found\n" ; } |
| 63 | + } |
| 64 | + |
| 65 | + # assuming only one file is archived per day ! |
| 66 | + if ($head_found && $tail_found) |
| 67 | + { |
| 68 | + $full_range_found = $true ; |
| 69 | + last ; |
| 70 | + } |
| 71 | + } |
| 72 | + } |
| 73 | + |
| 74 | + if (! $some_files_found) |
| 75 | + { print "Not any file was found which contains log records for $days_ago days ago. Skip processing for $date_collect_files.\n\n" ; return $false ; } |
| 76 | + if (! $full_range_found) |
| 77 | + { print "Not all files were found which contain log records for $days_ago days ago. Skip processing for $date_collect_files.\n\n" ; return $false ; } |
| 78 | + |
| 79 | + print "\n" ; |
| 80 | + foreach $file (sort @files) |
| 81 | + { print "Process $file\n" ; } |
| 82 | + |
| 83 | + return $true ; |
| 84 | +} |
| 85 | + |
| 86 | +sub ReadIpFrequencies |
| 87 | +{ |
| 88 | + trace ReadIpFrequencies ; |
| 89 | + |
| 90 | + my $path_out = shift ; |
| 91 | + |
| 92 | + my $data_read = $false ; |
| 93 | + |
| 94 | + if ($job_runs_on_production_server) |
| 95 | + { |
| 96 | + if (! -e "$path_out/$file_ip_frequencies_bz2") |
| 97 | + { print "$path_out/$file_ip_frequencies_bz2 not found. Abort processing for this day." ; return $false ; } |
| 98 | + |
| 99 | + open CSV_ADDRESSES, "-|", "bzip2 -dc \"$path_out/$file_ip_frequencies_bz2\"" || abort ("Input file $path_out/$file_ip_frequencies_bz2 could not be opened.") ; |
| 100 | + } |
| 101 | + else |
| 102 | + { |
| 103 | + if (! -e "$path_out/$file_ip_frequencies") |
| 104 | + { print "$path_out/$file_ip_frequencies not found. Abort processing for this day." ; return $false ; } |
| 105 | + |
| 106 | + open CSV_ADDRESSES, '<', "$path_out/$file_ip_frequencies" || abort ("Input file $path_out/$file_ip_frequencies could not be opened.") ; |
| 107 | + } |
| 108 | + |
| 109 | + while ($line = <CSV_ADDRESSES>) |
| 110 | + { |
| 111 | + $data_read = $true ; |
| 112 | + |
| 113 | + if ($line =~ /^#/o) { next ; } |
| 114 | + chomp ($line) ; |
| 115 | + ($frequency, $address) = split (',', $line) ; |
| 116 | + $ip_frequencies {$address} = $frequency ; |
| 117 | + $addresses_stored++ ; |
| 118 | + } |
| 119 | + |
| 120 | + print "\n$addresses_stored addresses stored that occur more than once\n\n" ; |
| 121 | + |
| 122 | + return $data_read ; |
| 123 | +} |
| 124 | + |
| 125 | +sub ReadSquidLogFiles |
| 126 | +{ |
| 127 | + trace ReadSquidLogFiles ; |
| 128 | + |
| 129 | + my $data_read = $false ; |
| 130 | + |
| 131 | + my ($path_out, $time_to_start, $time_to_stop, @files) = @_ ; |
| 132 | + |
| 133 | + if ($#files == -1) |
| 134 | + { print "ReadInput: No files to process.\n\n" ; } |
| 135 | + |
| 136 | + print "Read log records in range $time_to_start till $time_to_stop\n\n" ; |
| 137 | + |
| 138 | + if ($job_runs_on_production_server && $scan_all_fields) |
| 139 | + { open FILE_EDITS_SAVES, '>', "$path_out/$file_edits_saves" ; } |
| 140 | + |
| 141 | + my $lines = 0 ; |
| 142 | + while ($#files > -1) |
| 143 | + { |
| 144 | + $file_in = shift (@files) ; |
| 145 | + |
| 146 | + print "Process $file_in\n" ; |
| 147 | + if (! -e $file_in) |
| 148 | + { print "ReadInput: File not found: $file_in. Aborting...\n\n" ; exit ; } |
| 149 | + |
| 150 | + if ($job_runs_on_production_server) |
| 151 | + { |
| 152 | + if ($file_in =~ /\.gz$/o) |
| 153 | + { open IN, "-|", "gzip -dc $file_in | /usr/local/bin/geoiplogtag 5" ; } # http://perldoc.perl.org/functions/open.html |
| 154 | + else |
| 155 | + { open IN, "-|", "cat $file_in | /usr/local/bin/geoiplogtag 5" ; } # http://perldoc.perl.org/functions/open.html |
| 156 | + $fields_expected = 14 ; |
| 157 | + } |
| 158 | + else |
| 159 | + { |
| 160 | + open IN, '<', $file_in ; |
| 161 | + # $fields_expected = 14 ; |
| 162 | + $fields_expected = 13 ; |
| 163 | + } |
| 164 | + |
| 165 | + $line = "" ; |
| 166 | + while ($line = <IN>) |
| 167 | + { |
| 168 | + $lines_in_file ++ ; |
| 169 | + |
| 170 | + # if ($line =~ /fy\.wikipedia\.org/o) # test/debug |
| 171 | + # { |
| 172 | + # print FILTER_FY $line ; |
| 173 | + # print $line ; |
| 174 | + # } |
| 175 | + |
| 176 | + @fields = split (' ', $line) ; |
| 177 | + if ($#fields < $fields_expected) { $fields_too_few ++ ; next ; } |
| 178 | + if ($#fields > $fields_expected) { $fields_too_many ++ ; next ; } |
| 179 | + |
| 180 | + $time = $fields [2] ; |
| 181 | + |
| 182 | + if (($oldest_time_read eq "") || ($time lt $oldest_time_read)) |
| 183 | + { $oldest_time_read = $time ; } |
| 184 | + if (($newest_time_read eq "") || ($time gt $newest_time_read)) |
| 185 | + { $newest_time_read = $time ; } |
| 186 | + |
| 187 | + if ($oldest_time_read ge $time_to_stop) |
| 188 | + { last ; } |
| 189 | + |
| 190 | + if ($time lt $time_to_start) |
| 191 | + { |
| 192 | + if (++ $times % 100000 == 0) |
| 193 | + { print "[$time]\n" ; } |
| 194 | + next ; |
| 195 | + } |
| 196 | + |
| 197 | + if ($time ge $time_to_stop) |
| 198 | + { last ; } |
| 199 | + |
| 200 | + $date = substr ($time,0,10) ; |
| 201 | + if ($date lt $date_prev) { next ; } # occasionally one record for previous day arrives late |
| 202 | + |
| 203 | + $data_read = $true ; |
| 204 | + if ($date ne $date_prev) |
| 205 | + { |
| 206 | + print &ddhhmmss (time - $time_start) . " $date\n" ; |
| 207 | + if ($date_prev ne "") |
| 208 | + { |
| 209 | + print "$date_prev: $lines_this_day\n" ; |
| 210 | + $lines_read {$date_prev} = $lines_this_day ; |
| 211 | + } |
| 212 | + $lines_this_day = 0 ; |
| 213 | + $date_prev = $date ; |
| 214 | + } |
| 215 | + $lines_this_day++ ; |
| 216 | + |
| 217 | + if ($job_runs_on_production_server) |
| 218 | + { |
| 219 | + if (($line =~ /action=edit/o) || ($line =~ /action=submit/o)) |
| 220 | + { print FILE_EDITS_SAVES $line ; } |
| 221 | + } |
| 222 | + |
| 223 | + $lines++ ; |
| 224 | + |
| 225 | +#next if $line !~ /http:\/\/\w+\.m\./ ; |
| 226 | +#print "$line\n" ; |
| 227 | + &ProcessLine ($line) ; |
| 228 | + if (++ $lines_processed % 10000 == 0) |
| 229 | + { |
| 230 | + if ($banner_requests_ignored == 0) |
| 231 | + { print "$time $lines_processed\n" ; } |
| 232 | + else |
| 233 | + { print "$time $lines_processed ($banner_requests_ignored banner requests ignored)\n" ; } |
| 234 | + } |
| 235 | + if ($test and $lines_processed >= $test_maxlines) |
| 236 | + { last ; } |
| 237 | + } |
| 238 | + close IN ; |
| 239 | + } |
| 240 | + |
| 241 | + if ($scan_ip_frequencies) |
| 242 | + { return ($data_read) ; } |
| 243 | + |
| 244 | + if ($job_runs_on_production_server) |
| 245 | + { close FILE_EDITS_SAVES ; } |
| 246 | + |
| 247 | + $lines_read {$date_prev} = $lines_this_day ; |
| 248 | + |
| 249 | + if ($lines == 0) |
| 250 | + { |
| 251 | + $data_read = $false ; |
| 252 | + print "No data found for $time_to_start - $time_to_stop\n" ; |
| 253 | + } |
| 254 | + else |
| 255 | + { print "$lines_this_day out $lines_in_file processed\n" ; } |
| 256 | + |
| 257 | + return ($data_read) ; |
| 258 | +} |
| 259 | + |
| 260 | +sub ReadInputEditsSavesFile |
| 261 | +{ |
| 262 | + trace ReadInputEditsSavesFile ; |
| 263 | + |
| 264 | + my $file_txt = shift ; |
| 265 | + |
| 266 | + print "Process $file_txt\n" ; |
| 267 | + |
| 268 | + open IN, "-|", "bzip2 -dc \"$file_txt\"" || abort ("Input file '" . $file_txt . "' could not be opened.") ; |
| 269 | +# open IN, '<', "2010-04/SquidDataEditsSaves2010-04-01.txt" || abort ("Input file '" . $file_txt . "' could not be opened.") ; # test |
| 270 | + |
| 271 | + while ($line = <IN>) |
| 272 | + { |
| 273 | + if ($line =~ /index\.php/o) |
| 274 | + { &ProcessLine ($line) ; } |
| 275 | + } |
| 276 | + close IN ; |
| 277 | +} |
| 278 | + |
| 279 | +sub GetLogRange # finding first and last timestamp ('head' and 'tail') in compressed file is costly, cache results for reuse |
| 280 | +{ |
| 281 | + my ($file,$path_head_tail) = @_ ; |
| 282 | + |
| 283 | + if (-e $path_head_tail) |
| 284 | + { |
| 285 | + open CSV_HEAD_TAIL, '<', $path_head_tail ; |
| 286 | + while ($line = <CSV_HEAD_TAIL>) |
| 287 | + { |
| 288 | + chomp $line ; |
| 289 | + my ($logfile,$head,$tail) = split (',', $line) ; |
| 290 | + $timeheads {$logfile} = $head ; |
| 291 | + $timetails {$logfile} = $tail ; |
| 292 | + } |
| 293 | + close CSV_HEAD_TAIL ; |
| 294 | + } |
| 295 | + |
| 296 | + $timehead = $timeheads {$file} ; |
| 297 | + $timetail = $timetails {$file} ; |
| 298 | + |
| 299 | + if (($timehead ne '') && ($timetail ne '')) |
| 300 | + { |
| 301 | + print "- HEAD $timehead TAIL $timetail (from head-tail cache)\n" ; |
| 302 | + return ($timehead, $timetail) ; |
| 303 | + } |
| 304 | + |
| 305 | + my ($line, @fields, $timehead, $timetail) ; |
| 306 | + print "$file: " ; |
| 307 | + if (! -e $file) |
| 308 | + { |
| 309 | + print "- GetLogRange error: File not found: $file\n" ; |
| 310 | + exit ; |
| 311 | + } |
| 312 | + |
| 313 | + if ($file =~ /\.gz$/o) |
| 314 | + { $line = `gzip -dc $file | head -n 1 ` ; } |
| 315 | + else |
| 316 | + { $line = `head -n 1 $file` ; } |
| 317 | + # print "HEAD $line\n" ; |
| 318 | + @fields = split (' ', $line) ; |
| 319 | + # $timehead = substr ($fields [2],0,10) ; |
| 320 | + $timehead = $fields [2] ; |
| 321 | + |
| 322 | + if ($file =~ /\.gz$/o) |
| 323 | + { $line = `gzip -dc $file | tail -n 1 ` ; } |
| 324 | + else |
| 325 | + { $line = `tail -n 1 $file` ; } |
| 326 | + |
| 327 | + # print "TAIL $line\n" ; |
| 328 | + @fields = split (' ', $line) ; |
| 329 | + # $timetail = substr ($fields [2],0,10) ; |
| 330 | + $timetail = $fields [2] ; |
| 331 | + |
| 332 | + print "- HEAD $timehead TAIL $timetail\n" ; |
| 333 | + |
| 334 | + open CSV_HEAD_TAIL, '>>', $path_head_tail ; |
| 335 | + print CSV_HEAD_TAIL "$file,$timehead,$timetail\n" ; |
| 336 | + close CSV_HEAD_TAIL ; |
| 337 | + |
| 338 | + return ($timehead, $timetail) ; |
| 339 | +} |
| 340 | + |
| 341 | +sub GetTimeIso8601 |
| 342 | +{ |
| 343 | + my $time = shift ; |
| 344 | + my $year = substr ($time,0,4) ; |
| 345 | + my $mon = substr ($time,5,2) ; |
| 346 | + my $mday = substr ($time,8,2) ; |
| 347 | + my $hour = substr ($time,11,2) ; |
| 348 | + my $min = substr ($time,14,2) ; |
| 349 | + my $sec = substr ($time,17,2) ; |
| 350 | + $time = timelocal($sec,$min,$hour,$mday,$mon-1,$year-1900); |
| 351 | + return ($time) ; |
| 352 | +} |
| 353 | + |
| 354 | +1; |
Index: trunk/wikistats/squids/SquidReportArchive.pl |
— | — | @@ -5,11 +5,9 @@ |
6 | 6 | $trace_on_exit = $true ; |
7 | 7 | ez_lib_version (2) ; |
8 | 8 | |
9 | | -# $quarter_only = '2010 Q3' ; # if not empty filter process for this quarter only |
10 | | - |
11 | 9 | # set defaults mainly for tests on local machine |
12 | 10 | # default_argv "-m 201010 " ; |
13 | | - default_argv "-c " ; |
| 11 | + default_argv "-c -q 2010Q4" ; |
14 | 12 | |
15 | 13 | # to do: add text from http://wiki.squid-cache.org/SquidFaq/SquidLogs |
16 | 14 | # ReportOrigin how to handle '!error <-> other |
— | — | @@ -30,7 +28,7 @@ |
31 | 29 | $ratio_sqrt = $true ; |
32 | 30 | $ratio_linear = $false ; |
33 | 31 | |
34 | | - getopt ("dm", \%options) ; |
| 32 | + getopt ("dmq", \%options) ; |
35 | 33 | |
36 | 34 | if (-d "/a/squid") |
37 | 35 | { |
— | — | @@ -63,6 +61,15 @@ |
64 | 62 | if (defined ($options {"c"})) |
65 | 63 | { $reportcountries = $true ; } |
66 | 64 | |
| 65 | + if (defined ($options {"q"})) |
| 66 | + { |
| 67 | + $quarter_only = $options {"q"} ; # process for this quarter only |
| 68 | + if ($quarter_only !~ /^2\d\d\dQ\d$/) |
| 69 | + { abort ("Specify run for one single quarter as -q yyyyQ[1-4], e.g. -q 2011Q3, not '$quarter_only'\n") ; } |
| 70 | + $quarter_only =~ s/^(\d\d\d\d)(Q\d)$/$1 $2/ ; |
| 71 | + print "QUARTER ONLY $quarter_only\n" ; |
| 72 | + } |
| 73 | + |
67 | 74 | # date range used to be read from csv file with ReadDate, now there are daily csv files |
68 | 75 | # if earlier methods still is useful it needs to be tweaked |
69 | 76 | # if (($reportmonth ne "") && ($reportmonth !~ /^\d{6}$/)) |
— | — | @@ -83,7 +90,7 @@ |
84 | 91 | &CollectRegionCounts ; |
85 | 92 | |
86 | 93 | &ReportCountries ('Saves'); |
87 | | -# &ReportCountries ('Views'); |
| 94 | + &ReportCountries ('Views'); |
88 | 95 | |
89 | 96 | exit ; |
90 | 97 | } |
— | — | @@ -1257,7 +1264,8 @@ |
1258 | 1265 | $path_csv_country_codes = "$path_in/$file_csv_country_codes" ; |
1259 | 1266 | if (! -e $path_csv_country_codes) { abort ("Input file $path_csv_country_codes not found!") ; } |
1260 | 1267 | |
1261 | | - open CSV_COUNTRY_CODES, '<', $path_csv_country_codes ; |
| 1268 | + open CSV_COUNTRY_CODES, '<', $path_csv_country_codes ; |
| 1269 | + binmode CSV_COUNTRY_CODES ; |
1262 | 1270 | $country_names {"--"} = "Unknown" ; |
1263 | 1271 | while ($line = <CSV_COUNTRY_CODES>) |
1264 | 1272 | { |
— | — | @@ -1265,6 +1273,7 @@ |
1266 | 1274 | |
1267 | 1275 | next if $line =~ /^#/ ; |
1268 | 1276 | |
| 1277 | + $line =~ s/[\x00-\x1f]//g ; |
1269 | 1278 | $line =~ s/C..?te d'Ivoire/Côte d'Ivoire/g ; |
1270 | 1279 | |
1271 | 1280 | ($country_code,$region_code,$north_south_code,$country_name) = split (',', $line,4) ; |
— | — | @@ -1295,11 +1304,13 @@ |
1296 | 1305 | { |
1297 | 1306 | # http://en.wikipedia.org/wiki/List_of_countries_by_population |
1298 | 1307 | # http://en.wikipedia.org/wiki/List_of_countries_by_number_of_Internet_users |
1299 | | - open COUNTRY_META_INFO, '<', "$path_in/$file_csv_country_meta_info" ; |
| 1308 | + open COUNTRY_META_INFO, '<', "$path_in/$file_csv_country_meta_info" ; |
| 1309 | + binmode COUNTRY_META_INFO ; |
1300 | 1310 | while ($line = <COUNTRY_META_INFO>) |
1301 | 1311 | { |
1302 | 1312 | chomp $line ; |
1303 | 1313 | |
| 1314 | + $line =~ s/[\x00-\x1f]//g ; |
1304 | 1315 | $line =~ s/C..?te d'Ivoire/Côte d'Ivoire/g ; |
1305 | 1316 | |
1306 | 1317 | ($country,$link,$population,$connected,$icon) = split ',', $line ; |
— | — | @@ -1363,6 +1374,9 @@ |
1364 | 1375 | |
1365 | 1376 | # print "CODE $country_code NAME $country_name POP $population, $CONN $connected REGION $region_code NS $north_south_code PPR ${population_per_region {$region_code}}\n" ; |
1366 | 1377 | } |
| 1378 | + |
| 1379 | + if ($population_tot == 0) |
| 1380 | + { abort ("No valid data found: population_tot = 0 !") ; } |
1367 | 1381 | } |
1368 | 1382 | |
1369 | 1383 | sub ReadInputCountriesMonthly |
— | — | @@ -1424,7 +1438,6 @@ |
1425 | 1439 | |
1426 | 1440 | $year = substr ($yyyymm,0,4) ; |
1427 | 1441 | $month = substr ($yyyymm,5,2) ; |
1428 | | - # print "year $year report_year month $month $report_year $report_month\n" ; |
1429 | 1442 | |
1430 | 1443 | $recently = $false ; |
1431 | 1444 | |
— | — | @@ -1622,7 +1635,6 @@ |
1623 | 1636 | $new = &CorrectForMissingDays ($week , ${$requests_per_week_per_country_code {$week }} {$country_code}) ; |
1624 | 1637 | $old = &CorrectForMissingDays ($week-1, ${$requests_per_week_per_country_code {$week-1}} {$country_code}) ; |
1625 | 1638 | |
1626 | | - # print "country_code $country_code\n" ; |
1627 | 1639 | if ($old == 0) |
1628 | 1640 | { |
1629 | 1641 | if ($new > 0) |
— | — | @@ -4432,7 +4444,10 @@ |
4433 | 4445 | $population2 = &i2KM2 ($population) ; |
4434 | 4446 | $connected2 = &i2KM2 ($connected) ; |
4435 | 4447 | $requests_this_country2 = &i2KM2 ($requests_this_country2) ; |
4436 | | - $perc_population = &Percentage ($population / $population_tot) ; |
| 4448 | + |
| 4449 | + if ($population_tot > 0) |
| 4450 | + { $perc_population = &Percentage ($population / $population_tot) ; } |
| 4451 | + |
4437 | 4452 | if ($perc_population =~ /\.0\d/) |
4438 | 4453 | { $perc_population = "<small>$perc_population</small>" ; } |
4439 | 4454 | |