]>
Commit | Line | Data |
---|---|---|
a7fb5630 MT |
1 | #!/usr/bin/perl |
2 | ||
3 | ########################################## | |
4 | ## | |
5 | ## NAME | |
6 | ## | |
7 | ## DESCRIPTION | |
8 | ## | |
9 | ## Which is part of the ADSL-optimizer. | |
10 | ## | |
11 | ## USAGE / FUNCTIONS | |
12 | ## | |
13 | ## | |
14 | ## | |
15 | ## | |
16 | ## | |
17 | ## REQUIRES | |
18 | ## | |
19 | ## | |
20 | ## AUTHOR | |
21 | ## Jesper Dangaard Brouer <hawk@diku.dk>, d.15/4-2004 | |
22 | ## | |
23 | ## CHANGELOG | |
24 | ## 2004-04-15: Initial version. | |
25 | ## 2005-04-18: Remove some warnings. | |
26 | ## | |
27 | ## $Id: parse-func.pl,v 1.15 2005/04/18 12:37:17 hawk Exp $ | |
28 | ########################################## | |
29 | ||
30 | #use Data::Dumper; | |
31 | ||
32 | #our %classes_data; | |
33 | #our %classes_info; | |
34 | #our $tc_command="/sbin/tc"; | |
35 | ||
36 | my @input_htb = (<<"END_OF_HERE_HTB" =~ m/^\s*(.+)/gm); | |
37 | class tbf 4220:1 parent 4220: | |
38 | class htb 1:1 root rate 400Kbit ceil 400Kbit burst 2111b cburst 2111b | |
39 | Sent 12369084336 bytes 80967118 pkts (dropped 0, overlimits 0) | |
40 | rate 45020bps 258pps | |
41 | lended: 23353805 borrowed: 0 giants: 0 | |
42 | tokens: 30210 ctokens: 30210 | |
43 | ||
44 | class htb 1:10 parent 1:1 prio 0 rate 80Kbit ceil 320Kbit burst 1701b cburst 2008b | |
45 | Sent 80640087 bytes 247988 pkts (dropped 0, overlimits 0) | |
46 | backlog 42p | |
47 | lended: 230876 borrowed: 17112 giants: 0 | |
48 | tokens: 127200 ctokens: 37940 | |
49 | ||
50 | class htb 1:20 parent 1:1 leaf 4220: prio 1 rate 100Kbit ceil 200Kbit burst 1727b cburst 1855b | |
51 | Sent 2495181573 bytes 44034303 pkts (dropped 5837, overlimits 0) | |
52 | lended: 43825585 borrowed: 208718 giants: 0 | |
53 | tokens: 103424 ctokens: 55808 | |
54 | ||
55 | class htb 1:30 parent 1:1 leaf 4230: prio 3 rate 80Kbit ceil 400Kbit burst 1701b cburst 2111b | |
56 | Sent 2060213567 bytes 5465574 pkts (dropped 121, overlimits 0) | |
57 | rate 16851bps 35pps | |
58 | lended: 4556992 borrowed: 908582 giants: 0 | |
59 | tokens: -25364 ctokens: 32897 | |
60 | ||
61 | class htb 1:50 parent 1:1 leaf 4250: prio 5 rate 40Kbit ceil 120Kbit burst 1650b cburst 1752b | |
62 | Sent 6071486687 bytes 24448436 pkts (dropped 8086739, overlimits 0) | |
63 | rate 15801bps 85pps backlog 126p | |
64 | lended: 8324530 borrowed: 16123780 giants: 0 | |
65 | tokens: -202717 ctokens: -172499 | |
66 | ||
67 | class htb 1:666 parent 1:1 leaf 666: prio 7 rate 4Kbit ceil 40Kbit burst 1604b cburst 1650b | |
68 | Sent 2148626078 bytes 6771069 pkts (dropped 2078536, overlimits 0) | |
69 | rate 5221bps 17pps backlog 125p | |
70 | lended: 675330 borrowed: 6095613 giants: 0 | |
71 | tokens: -1149121 ctokens: -293386 | |
72 | ||
73 | END_OF_HERE_HTB | |
74 | ||
75 | ||
76 | my @input_hfsc = (<<"END_OF_HERE_HFSC" =~ m/^\s*(.+)/gm); | |
77 | class hfsc 1: root | |
78 | Sent 0 bytes 0 pkts (dropped 0, overlimits 0) | |
79 | period 0 level 2 | |
80 | ||
81 | class hfsc 1:1 parent 1: ls m1 0bps d 0us m2 250Kbit ul m1 0bps d 0us m2 250Kbit | |
82 | Sent 0 bytes 0 pkts (dropped 0, overlimits 0) | |
83 | period 6 work 131770097 bytes level 1 | |
84 | ||
85 | class hfsc 1:10 parent 1:1 rt m1 250Kbit d 30.0ms m2 50Kbit ls m1 250Kbit d 50.0ms m2 50Kbit | |
86 | Sent 1300885 bytes 7052 pkts (dropped 0, overlimits 0) | |
87 | period 6502 work 1300885 bytes rtwork 1245495 bytes level 0 | |
88 | ||
89 | class hfsc 1:20 parent 1: rt m1 0bps d 64.0ms m2 75Kbit ls m1 0bps d 0us m2 250Kbit | |
90 | Sent 19144279 bytes 325503 pkts (dropped 46, overlimits 0) | |
91 | backlog 3p | |
92 | period 20242 work 19143778 bytes level 0 | |
93 | ||
94 | class hfsc 1:30 parent 1:1 leaf 4230: ls m1 0bps d 150.0ms m2 50Kbit | |
95 | Sent 45139930 bytes 74200 pkts (dropped 1664, overlimits 0) | |
96 | backlog 24p | |
97 | period 140 work 44885232 bytes level 0 | |
98 | ||
99 | class hfsc 1:50 parent 1:1 leaf 4250: ls m1 0bps d 235.7ms m2 72Kbit | |
100 | Sent 73910198 bytes 301294 pkts (dropped 104807, overlimits 0) | |
101 | backlog 62p | |
102 | period 115 work 64625490 bytes level 0 | |
103 | ||
104 | class hfsc 1:666 parent 1:1 leaf 666: ls m1 0bps d 1.0s m2 2Kbit | |
105 | Sent 2217104 bytes 17018 pkts (dropped 74526, overlimits 0) | |
106 | backlog 22p | |
107 | period 1 work 1814712 bytes level 0 | |
108 | ||
109 | END_OF_HERE_HFSC | |
110 | ||
111 | sub parse_class($) { | |
112 | my $device = "$_[0]"; | |
113 | my $return_val = 1; | |
114 | ||
115 | my $timestamp = time; | |
116 | my @tc_output = `$tc_command -statistics class show dev $device`; | |
117 | # my @tc_output = @input_hfsc; | |
118 | # my @tc_output = @input_htb; | |
119 | my $result = $?; | |
120 | if ( $result != 0 ) { | |
121 | print "Error executing $tc_command\n"; | |
122 | return $result; | |
123 | } | |
124 | ||
125 | $classes_data{last_update}{$device} = $timestamp; | |
126 | $classes_info{last_update}{$device} = $timestamp; | |
127 | ||
128 | #for my $line (@tc_output) { | |
129 | for my $i (0 .. $#tc_output) { | |
130 | ||
131 | my $line=$tc_output[$i]; | |
132 | # Parsing HTB: | |
133 | # ------------ | |
f1df2c41 | 134 | if ( $line =~ m/class htb (\d+):(\d+)( root| parent )?(\d+:\d+)?( leaf )?(\d+)?:?( prio )?(\d+)? rate (.*) ceil (.*) burst (.*) cburst (.*)/ ) { |
a7fb5630 MT |
135 | my $type = "htb"; |
136 | my $major = $1; | |
137 | my $minor = $2; | |
138 | my $class = "${major}-${minor}"; | |
139 | #my $hash = "${class}_${device}"; | |
140 | my $parent= $4; | |
141 | my $leaf = $6; | |
142 | my $prio = $8; | |
143 | my $rate = $9; | |
144 | my $ceil = $10; | |
145 | my $burst = $11; | |
146 | my $cburst= $12; | |
147 | ||
6f300951 MT |
148 | #print "class: $class\n"."parent: $parent\n"."leaf: $leaf\n"."prio: $prio\n"; |
149 | #print "rate: $rate\n"."ceil: $ceil\n"."burst: $burst\n"."cburst: $cburst\n"; | |
a7fb5630 MT |
150 | |
151 | my ($bytes, $pkts, $dropped, $overlimits); | |
6f300951 | 152 | if ($tc_output[$i + 1] =~ m/Sent (\d+) bytes (\d+) pkt \(dropped (\d+), overlimits (\d+) requeues (\d+)\)/ ) { |
a7fb5630 MT |
153 | $bytes = $1; |
154 | $pkts = $2; | |
155 | $dropped = $3; | |
156 | $overlimits = $4; | |
6f300951 MT |
157 | $requeues = $5; |
158 | #print "bytes: $bytes\n"."pkts: $pkts\n"; | |
159 | #print "dropped: $dropped\n"."overlimits: $overlimits\n"."requeues: $requeues\n"; | |
a7fb5630 | 160 | } else { |
6f300951 MT |
161 | print "$timestamp: ERROR(+1) - Unable to parse (class ${class}_$device): "; |
162 | print "\"$tc_output[$i + 1]\"\n"; | |
a7fb5630 MT |
163 | $return_val=""; |
164 | next; | |
165 | } | |
166 | ||
167 | # Problem: | |
168 | # Sometimes the "rate" line is not shown (when a rate cannot be calculated) | |
169 | # And sometimes only "backlog"... | |
170 | # Use $next_index to specify the next line to parse | |
171 | # | |
172 | my $next_index = 3; | |
173 | my ($backlog); | |
174 | if ($tc_output[$i + 2] =~ m/((rate (\d+\w+) )|backlog )(\d+)?(pps )?(backlog )?(\d+)?p?/ ) { | |
175 | $backlog = $7; | |
176 | #print "backlog: $backlog\n"; | |
177 | } else { | |
178 | # Too verbose: | |
179 | # print "$timestamp: WARNING \"rate\" line missing"; | |
180 | # print " very inactive class ${class}_$device).\n"; | |
181 | $next_index = 2; | |
182 | } | |
183 | ||
184 | my ($lended, $borrowed, $giants); | |
185 | if ($tc_output[$i + $next_index] =~ m/lended: (\d+) borrowed: (\d+) giants: (\d+)/ ) { | |
186 | $lended = $1; | |
187 | $borrowed = $2; | |
188 | $giants = $3; | |
189 | #print "lended: $lended\n"."borrowed: $borrowed\n"."giants: $giants\n"; | |
190 | } else { | |
191 | print "$timestamp: ERROR(+$next_index) - Unable to parse (class ${class}_$device): "; | |
192 | print "\"$tc_output[$i + $next_index]\"\n"; | |
193 | $return_val=""; | |
194 | next; | |
195 | } | |
196 | ||
197 | # Update the hash tables | |
198 | my $hash="${class}_$device"; | |
199 | ||
200 | # Tests if previous data have been updated to file | |
201 | if ( (exists $classes_data{$hash}{last_update}) && | |
202 | (exists $classes_data{$hash}{file_update})) { | |
203 | if ( $classes_data{$hash}{last_update} > | |
204 | $classes_data{$hash}{file_update} ){ | |
205 | print "Warning: old data from $hash has not been updated to file!\n"; | |
206 | } | |
207 | } | |
208 | ||
209 | # Update the statistics data | |
210 | # (need a function call for error checking) | |
211 | $classes_data{$hash}{last_update} = $timestamp; | |
212 | update_counter( $hash, $timestamp, "bytes" , $bytes); | |
213 | #(yes I know its bad/redundant, but it makes in easier elsewhere) | |
214 | update_counter( $hash, $timestamp, "bits" , $bytes*8); | |
215 | update_counter( $hash, $timestamp, "pkts" , $pkts); | |
216 | update_counter( $hash, $timestamp, "dropped" , $dropped); | |
217 | update_counter( $hash, $timestamp, "overlimits", $overlimits); | |
218 | update_counter( $hash, $timestamp, "lended" , $lended); | |
219 | update_counter( $hash, $timestamp, "borrowed" , $borrowed); | |
220 | update_counter( $hash, $timestamp, "giants" , $giants); | |
221 | # Not a counter value... | |
222 | $classes_data{$hash}{backlog} = $backlog; | |
223 | ||
224 | # Update the info data | |
225 | # (remember to update the "type" first) | |
226 | update_info( $hash, $timestamp, "type" , $type); | |
227 | update_info( $hash, $timestamp, "parent", $parent); | |
228 | update_info( $hash, $timestamp, "leaf" , $leaf); | |
229 | update_info( $hash, $timestamp, "prio" , $prio); | |
230 | update_info( $hash, $timestamp, "rate" , $rate); | |
231 | update_info( $hash, $timestamp, "ceil" , $ceil); | |
232 | update_info( $hash, $timestamp, "burst" , $burst); | |
233 | update_info( $hash, $timestamp, "cburst", $cburst); | |
234 | ||
235 | #print "\n"; | |
236 | } | |
237 | ||
238 | # Parsing HFSC: | |
239 | # ------------- | |
240 | if ( $line =~ m/class hfsc (\d+):(\d+)( root| parent )?(\d+:\d?)?( leaf )?(\d+)?:?( rt m1 (\d+\w+?) d (\d+.?\d?\w+) m2 (\d+\w+?))?( ls m1 (\d+\w+?) d (\d+.?\d?\w+) m2 (\d+\w+?))?( ul m1 (\d+\w+?) d (\d+.?\d?\w+) m2 (\d+\w+?))? / ){ | |
241 | ||
242 | my $type = "hfsc"; | |
243 | my $major = $1; | |
244 | my $minor = $2; | |
245 | my $class = "${major}-${minor}"; | |
246 | #my $hash = "${class}_${device}"; | |
247 | my $parent= $4; | |
248 | my $leaf = $6; | |
249 | ||
250 | my $realtime_m1; if (defined $8 && $8 ne '0bps') {$realtime_m1 = $8;} | |
251 | my $realtime_d; if (defined $9 && $9 ne '0us' ) {$realtime_d = $9;} | |
252 | my $realtime_m2 = $10; | |
253 | ||
254 | my $linkshare_m1; if (defined $12 && $12 ne '0bps') { $linkshare_m1 = $12;} | |
255 | my $linkshare_d ; if (defined $13 && $13 ne '0us' ) { $linkshare_d = $13;} | |
256 | my $linkshare_m2 = $14; | |
257 | ||
258 | my $upperlimit_m1; if (defined $16 && $16 ne '0bps') { $upperlimit_m1 = $16;} | |
259 | my $upperlimit_d ; if (defined $17 && $17 ne '0us' ) { $upperlimit_d = $17;} | |
260 | my $upperlimit_m2 = $18; | |
261 | ||
262 | #print "\nType: $type\n"; | |
263 | my ($bytes, $pkts, $dropped, $overlimits); | |
264 | if ($tc_output[$i + 1] =~ m/Sent (\d+) bytes (\d+) pkts \(dropped (\d+), overlimits (\d+)\)/ ) { | |
265 | $bytes = $1; | |
266 | $pkts = $2; | |
267 | $dropped = $3; | |
268 | $overlimits = $4; | |
269 | #print "bytes: $bytes\n"."pkts: $pkts\n"; | |
270 | #print "dropped: $dropped\n"."overlimits: $overlimits\n"; | |
271 | } else { | |
272 | print "$timestamp: ERROR(+1) - Unable to parse (class ${class}_$device): "; | |
273 | print "\"$tc_output[$i + 1]\"\n"; | |
274 | $return_val=""; | |
275 | next; | |
276 | } | |
277 | ||
278 | # Sometimes the "backlog" line is not shown (when there is no backlog...) | |
279 | # Use $next_index to specify the next line to parse | |
280 | # | |
281 | my $next_index = 3; | |
282 | my ($backlog); | |
283 | if ($tc_output[$i + 2] =~ m/backlog (\d+)?p?/ ) { | |
284 | $backlog = $1; | |
285 | #print "backlog: $backlog\n"; | |
286 | } else { | |
287 | $next_index = 2; | |
288 | } | |
289 | ||
290 | my ($period, $work, $rtwork, $level); | |
291 | if ($tc_output[$i + $next_index] =~ m/period (\d+) (work (\d+) bytes )?(rtwork (\d+) bytes )?level (\d+)/ ) { | |
292 | $period = $1; | |
293 | $work = $3; | |
294 | $rtwork = $5; | |
295 | $level = $6 | |
296 | } else { | |
297 | print "$timestamp: ERROR(+$next_index) - Unable to parse (class ${class}_$device): "; | |
298 | print "\"$tc_output[$i + $next_index]\"\n"; | |
299 | $return_val=""; | |
300 | next; | |
301 | } | |
302 | ||
303 | ||
304 | # Update the hash tables | |
305 | my $hash="${class}_$device"; | |
306 | ||
307 | # Tests if previous data have been updated to file | |
308 | if ( (exists $classes_data{$hash}{last_update}) && | |
309 | (exists $classes_data{$hash}{file_update})) { | |
310 | if ( $classes_data{$hash}{last_update} > | |
311 | $classes_data{$hash}{file_update} ){ | |
312 | print "Warning: old data from $hash has not been updated to file!\n"; | |
313 | } | |
314 | } | |
315 | ||
316 | # HFSC - Update the statistics data | |
317 | # (need a function call for error checking) | |
318 | $classes_data{$hash}{last_update} = $timestamp; | |
319 | update_counter( $hash, $timestamp, "bytes" , $bytes); | |
320 | #(yes I know its bad/redundant, but it makes in easier elsewhere) | |
321 | update_counter( $hash, $timestamp, "bits" , $bytes*8); | |
322 | update_counter( $hash, $timestamp, "pkts" , $pkts); | |
323 | update_counter( $hash, $timestamp, "dropped" , $dropped); | |
324 | update_counter( $hash, $timestamp, "overlimits", $overlimits); | |
325 | # Not a counter value... | |
326 | $classes_data{$hash}{backlog} = $backlog; | |
327 | # | |
328 | # Extra HFSC counters | |
329 | $classes_data{$hash}{hfsc_period} = $period; | |
330 | update_counter( $hash, $timestamp, "hfsc_work" , $work); | |
331 | update_counter( $hash, $timestamp, "hfsc_rtwork" , $rtwork); | |
332 | ||
333 | ||
334 | # HFSC - Update the info data | |
335 | # (remember to update the "type" first) | |
336 | update_info( $hash, $timestamp, "type" , $type); | |
337 | update_info( $hash, $timestamp, "parent", $parent); | |
338 | update_info( $hash, $timestamp, "leaf" , $leaf); | |
339 | # | |
340 | # Extra HFSC information | |
341 | update_info( $hash, $timestamp, "level" , $level); | |
342 | update_info( $hash, $timestamp, "realtime_m1", $realtime_m1); | |
343 | update_info( $hash, $timestamp, "realtime_d" , $realtime_d); | |
344 | update_info( $hash, $timestamp, "realtime_m2", $realtime_m2); | |
345 | ||
346 | update_info( $hash, $timestamp, "linkshare_m1", $linkshare_m1); | |
347 | update_info( $hash, $timestamp, "linkshare_d" , $linkshare_d); | |
348 | update_info( $hash, $timestamp, "linkshare_m2", $linkshare_m2); | |
349 | ||
350 | update_info( $hash, $timestamp, "upperlimit_m1", $upperlimit_m1); | |
351 | update_info( $hash, $timestamp, "upperlimit_d" , $upperlimit_d); | |
352 | update_info( $hash, $timestamp, "upperlimit_m2", $upperlimit_m2); | |
353 | ||
354 | ||
355 | } | |
356 | ||
357 | # Parsing XXX: | |
358 | # ------------ | |
359 | if ( $line =~ m/class XXX/ ) { | |
360 | print "Matching class XXX\n"; | |
361 | } | |
362 | ||
363 | } | |
364 | return $return_val; | |
365 | } | |
366 | ||
367 | # The main purpose of this function is to detect counter resets | |
368 | # and avoid parsing them on to RRDtool which interprets them | |
369 | # as counter overflows, thus updating with a very large number. | |
370 | sub update_counter ($$$$) { | |
371 | my $class_hash = "$_[0]"; | |
372 | my $timestamp = "$_[1]"; | |
373 | my $data_key = "$_[2]"; | |
374 | my $new_value; | |
375 | if ( defined $_[3]) { | |
376 | $new_value = "$_[3]"; | |
377 | } | |
378 | # | |
379 | my $max_allowed_wrap_increase = 100000000; | |
380 | my $old_value; | |
381 | if (exists $classes_data{$class_hash}{$data_key}) { | |
382 | $old_value = $classes_data{$class_hash}{$data_key}; | |
383 | #print "old_value: $old_value\n"; | |
384 | } | |
385 | ||
386 | # # If the new and old value is not defined, nothing is done | |
387 | # if ((not defined $new_value) && (not defined $old_value)) { | |
388 | # return ""; | |
389 | # } | |
390 | ||
391 | # Argh... the tc program outputs in unsigned long long (64 bit). | |
392 | # but perls integers should be 32 bit, but some how perl | |
393 | # manages to store numbers larger than 32 bit numbers. | |
394 | my $MAX_VALUE=0xFFFFFFFF; | |
395 | ||
396 | if ((defined $new_value) && (defined $old_value)) { | |
397 | my $delta = $new_value - $old_value; | |
398 | if ( $delta < 0 ) { | |
399 | # Counter wrap around... | |
400 | my $real_delta = $delta + $MAX_VALUE + 1; | |
401 | if ($real_delta < 0) { | |
402 | print "($class_hash:$data_key): Perl-Magic using numbers bigger than 32bit "; | |
403 | print "new:$new_value - old:$old_value = delta:$delta, real_delta:$real_delta.\n"; | |
404 | } | |
405 | print time . " ($class_hash:$data_key) Info: Counter wrap around (real delta:$real_delta)\n"; | |
406 | if ( ($real_delta > $max_allowed_wrap_increase) || | |
407 | ($real_delta < 0)) { | |
408 | # Properly a counter reset and not a wrap around | |
409 | # A counter reset normally a result of a reload of the classes | |
410 | $classes_data{$class_hash}{$data_key} = undef; | |
411 | $classes_info{$class_hash}{counter_reset} = $timestamp; | |
412 | $classes_info{$class_hash}{last_update} = $timestamp; | |
413 | print time . "Warning: Real_delta too big, assuming Counter reset"; | |
414 | print "($class_hash:$data_key)\n"; | |
415 | return "Counter reset"; | |
416 | } | |
417 | } | |
418 | } | |
419 | ||
420 | $classes_data{$class_hash}{$data_key} = $new_value; | |
421 | return 1; | |
422 | } | |
423 | ||
424 | sub update_info ($$$$) { | |
425 | my $class_hash = "$_[0]"; | |
426 | my $timestamp = "$_[1]"; | |
427 | my $info_key = "$_[2]"; | |
428 | my $new_value; | |
429 | if ( defined $_[3]) { | |
430 | $new_value = "$_[3]"; | |
431 | } | |
432 | my $old_value; | |
433 | if (exists $classes_info{$class_hash}{$info_key}) { | |
434 | $old_value = $classes_info{$class_hash}{$info_key}; | |
435 | #print "old_value: $old_value\n"; | |
436 | } | |
437 | ||
438 | # If the new and old value is not defined, nothing is done | |
439 | if ((not defined $new_value) && (not defined $old_value)) { | |
440 | return ""; | |
441 | } | |
442 | ||
443 | # An update is needed | |
444 | # - if the old_value is not defined and new_value is defined | |
445 | # - if the new_value is not defined and old_value is defined | |
446 | # - if the old_value differs from the new, | |
447 | # | |
448 | if ( ((not defined $old_value) and (defined $new_value)) || | |
449 | ((not defined $new_value) and (defined $old_value)) || | |
450 | ("$old_value" ne "$new_value")) { | |
451 | ||
452 | # Special case: If the "type" changes the hash should be cleared | |
453 | if ( "$info_key" eq "type") { | |
454 | #print "Type has changed clearing hash \n"; | |
455 | for my $key ( keys %{ $classes_info{$class_hash} } ) { | |
456 | delete( $classes_info{$class_hash}{$key}); | |
457 | print " Deleting key: $key from: $class_hash \n"; | |
458 | } | |
459 | } | |
460 | ||
461 | if (defined $new_value) { | |
462 | $classes_info{$class_hash}{$info_key} = $new_value; | |
463 | } else { | |
464 | #print "New value undef -> Deleting key: $info_key from: $class_hash\n"; | |
465 | delete($classes_info{$class_hash}{$info_key}); | |
466 | } | |
467 | ||
468 | # Mark the class for an info-file update | |
469 | $classes_info{$class_hash}{last_update} = $timestamp; | |
470 | ||
471 | # Update list/array of "changed" keys | |
472 | push @{ $classes_info{$class_hash}{changed} }, $info_key; | |
473 | ||
474 | # Print debug info | |
475 | #print "Update class:$class_hash $info_key="; | |
476 | #if (defined $new_value) {print "$new_value"}; | |
477 | #print "\n"; | |
478 | return 1; | |
479 | } | |
480 | return ""; | |
481 | } | |
482 | ||
483 | # test | |
484 | #parse_class(eth1); | |
485 | ||
486 | #print Dumper(%classes_data); | |
487 | #print Dumper(%classes_info); | |
488 | ||
489 | return 1; |