[Bio] / FigKernelPackages / P2P.pm Repository:
ViewVC logotype

Annotation of /FigKernelPackages/P2P.pm

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1.27 - (view) (download) (as text)

1 : olson 1.1 #
2 :     # This module contains the code for the P2P update protocol.
3 :     #
4 :     # Package P2P contains the namespace declarations, and possibly toplevel utility
5 :     # routines. (get_relay ?)
6 :     #
7 :     # Package P2P::Relay contains methods for contacting the P2P relay service. The actual
8 :     # implementation of the relay service is not contained here - it is a standalone module
9 :     # that can be installed on a web server that does not have a full SEED installed.
10 :     #
11 :     # Package P2P::Requestor contains the requestor-side code for the update protocol.
12 :     #
13 :     # Package P2P::Service contains the web service implementation routines for the
14 :     # protocol.
15 :     #
16 :    
17 :     package P2P;
18 :    
19 :     use FIG_Config;
20 :    
21 : olson 1.24 use AnyDBM_File;
22 : olson 1.19 use Fcntl;
23 :    
24 : olson 1.1 use strict;
25 :     use Exporter;
26 :     use base qw(Exporter);
27 :    
28 : olson 1.15 use Time::HiRes qw( usleep ualarm gettimeofday tv_interval );
29 :    
30 : olson 1.1 use Data::Dumper;
31 :    
32 :     use vars qw(@EXPORT @EXPORT_OK);
33 :     @EXPORT = ();
34 :     @EXPORT_OK = qw($ns_p2p $ns_relay);
35 :    
36 :     our $ns_p2p = "http://thefig.info/schemas/p2p_update";
37 :     our $ns_relay = "http://thefig.info/schemas/p2p_relay";
38 :    
39 : olson 1.21 my $peg_batch_size = 1000;
40 :     my $anno_batch_size = 1000;
41 :     my $assign_batch_size = 1000;
42 : olson 1.23 my $fin_batch_size = 1000;
43 : olson 1.21
44 : olson 1.22 my $log_fh;
45 : olson 1.23 my $html_fh;
46 : olson 1.22
47 : olson 1.1 =pod
48 :    
49 : olson 1.25 =head1 perform_update($peer, $last_update, $skip_tough_search, $update_thru, $log_file, $html_file, $assignment_policy))
50 : olson 1.1
51 :     Perform a peer-to-peer update with the given peer. $peer is an instance of
52 :     P2P::Requestor which can connect to the peer. It is expected that the
53 :     SEED infrastructure will create this requestor appropriately for the
54 :     particular circumstance (direct connection, thru relay, etc).
55 :    
56 :     This code executes the high-level protocol, maintaining state between
57 :     calls to the peer to exchange the actual information.
58 :    
59 : olson 1.25 $last_update: Search for updates since this time.
60 :     $skip_tough_search: Do not use the time-consuming $fig->tough_search method as a last resort for peg mapping.
61 :     $update_thru: Search for updates until this time. Undef means to search for all since $last_update.
62 :     $log_file: Write logging information to this file.
63 :     $html_file: Write a HTML summary to this file.
64 :     $assignment_policy: If a list reference, contains the list of users from which we will accept assignments. If a code ref, a predicate that is passed ($peg, $timestamp, $author, $function) and returns true if the assignment should be made.
65 :    
66 : olson 1.1 =cut
67 :    
68 :     sub perform_update
69 :     {
70 : olson 1.25 my($fig, $peer, $last_update, $skip_tough_search, $update_thru, $log_file, $html_file,
71 :     $assignment_policy) = @_;
72 :    
73 :     my $allow_assignment;
74 : olson 1.22
75 :     $log_file = "/dev/null" unless $log_file ne "";
76 :     open($log_fh, ">>$log_file") or die "Cannot open logfile $log_file: $!\n";
77 :     $log_fh->autoflush(1);
78 :    
79 : olson 1.23 $html_file = "/dev/null" unless $html_file ne "";
80 :     open($html_fh, ">$html_file") or die "Cannot open htmlfile $html_file: $!\n";
81 :     $html_fh->autoflush(1);
82 :    
83 : olson 1.27 if (!defined($assignment_policy))
84 :     {
85 :     $allow_assignment = sub { 1;};
86 :     }
87 :     elsif (ref($assignment_policy) eq "CODE")
88 : olson 1.25 {
89 :     $allow_assignment = $assignment_policy;
90 :     }
91 :     elsif (ref($assignment_policy) eq "ARRAY")
92 :     {
93 :     my $ahash = {};
94 :     map { $ahash->{$_}++; } @$assignment_policy;
95 :     $allow_assignment = sub {
96 :     return $ahash->{$_[2]};
97 :     };
98 :     }
99 :     elsif (ref($assignment_policy) eq "HASH")
100 :     {
101 :     $allow_assignment = sub {
102 :     return $assignment_policy->{$_[2]};
103 :     };
104 :     }
105 :     else
106 :     {
107 :     print $log_fh "Invalid assignment policy $assignment_policy\n";
108 :     die "Invalid assignment policy $assignment_policy\n";
109 :     }
110 :    
111 : olson 1.23 my $now = localtime();
112 :     my $last_str = localtime($last_update);
113 :     print $html_fh <<END;
114 :     <h1>P2P Update at $now</h1>
115 :     Peer URL $peer->{url}<br>
116 :     Update from: $last_str<br>
117 :     END
118 :    
119 :     print $log_fh "Beginning P2P update at $now\n";
120 : olson 1.22 print $log_fh " Peer URL: $peer->{url}\n";
121 : olson 1.23 print $log_fh " Update from: $last_str\n";
122 : olson 1.22 print $log_fh "\n";
123 : olson 1.1
124 : olson 1.20 my $ret = $peer->request_update($last_update, $update_thru);
125 : olson 1.1
126 :     if (!$ret or ref($ret) ne "ARRAY")
127 :     {
128 : olson 1.18 die "perform_update: request_update failed\n";
129 : olson 1.1 }
130 :    
131 : olson 1.15 my($session, $target_release, $num_assignments, $num_annos, $num_pegs, $num_genomes,
132 : olson 1.1 $target_time, $compatible) = @$ret;
133 :    
134 : olson 1.18 print "perform_update: session=$session target=@$target_release num_annos=$num_annos\n";
135 : olson 1.1 print " num_pegs=$num_pegs num_genomes=$num_genomes target_time=$target_time compat=$compatible\n";
136 :    
137 : olson 1.22 my @my_release = $fig->get_release_info();
138 :    
139 :     print $log_fh "Session id = $session\n";
140 :     print $log_fh "Target release information: \n\t", join("\n\t", @$target_release), "\n";
141 :     print $log_fh "My release information: \n\t", join("\n\t", @my_release), "\n";
142 :     print $log_fh "$num_annos annotations\n";
143 :     print $log_fh "$num_assignments assignments\n";
144 :     print $log_fh "$num_pegs pegs\n";
145 :    
146 : olson 1.23 print $html_fh "Session id = $session<br>\n";
147 :     print $html_fh "Target release information: <br>\n\t", join("<br>\n\t", @$target_release), "<br>\n";
148 :     print $html_fh "My release information: <br>\n\t", join("<br>\n\t", @my_release), "<br>\n";
149 :     print $html_fh "$num_annos annotations<br>\n";
150 :     print $html_fh "$num_assignments assignments<br>\n";
151 :     print $html_fh "$num_pegs pegs<br>\n";
152 :    
153 : olson 1.1 #
154 : olson 1.19 # We now know the data release for our peer.
155 :     #
156 : olson 1.24 # Open up the peg translation cache database (a AnyDBM_File) tied
157 : olson 1.19 # to %peg_cache. We needn't worry about keeping it in a directory
158 :     # based on our current release, as it the cache directory is kept *in*
159 :     # the current data release directory.
160 :     #
161 :    
162 :     my $cache_handle;
163 :     my %peg_cache;
164 :     if ($target_release->[1] ne "")
165 :     {
166 :     my $cache_file = "pegcache.$target_release->[1].db";
167 :     my $cache_dir = "$FIG_Config::data/P2PQueue";
168 :     $fig->verify_dir($cache_dir);
169 :    
170 : olson 1.24 $cache_handle = tie(%peg_cache, "AnyDBM_File", "$cache_dir/$cache_file",
171 :     O_CREAT | O_RDWR, 0666);
172 : olson 1.19 $cache_handle or warn "Could not tie peg_cache to $cache_dir/$cache_file: $!\n";
173 :     }
174 :    
175 :     #
176 : olson 1.21 # peg_mapping is the local mapping from remote->local peg. This might
177 :     # be replacable by peg_cache from above.
178 : olson 1.1 #
179 : olson 1.21 my %peg_mapping;
180 : olson 1.1
181 : olson 1.21
182 : olson 1.1 #
183 : olson 1.21 # We have the information now to begin the update process. Retrieve the pegs.
184 : olson 1.1 #
185 :    
186 : olson 1.21 _compute_peg_mapping($fig, $peer, $session, $num_pegs, \%peg_mapping, \%peg_cache, $cache_handle,
187 :     $skip_tough_search);
188 : olson 1.7
189 : olson 1.24 eval { $cache_handle->sync();};
190 : olson 1.19 untie %peg_cache;
191 :    
192 :     #
193 : olson 1.15 # Create a list of locally-mapped annotations on a per-genome
194 :     # basis.
195 :     #
196 :    
197 :     my %genome_annos;
198 : olson 1.19
199 :     #
200 :     # %genome_assignments is a hash mapping from genome to a hashref
201 :     # that maps peg to function (since assignments are unique).
202 :     #
203 :     # (Hm. Unless two remote pegs map to the same local peg; unclear what to do
204 :     # then. Punt for now).
205 :     #
206 :     my %genome_assignments;
207 : olson 1.15
208 : olson 1.21 #
209 :     # Retrieve the annotations, and generate a list of mapped annotations.
210 :     #
211 :    
212 :     for (my $anno_start = 0; $anno_start < $num_annos; $anno_start += $anno_batch_size)
213 : olson 1.15 {
214 : olson 1.21 my $anno_req_len = $num_annos - $anno_start;
215 :     $anno_req_len = $anno_batch_size if $anno_req_len > $anno_batch_size;
216 : olson 1.15
217 : olson 1.21 print "Retrieve $anno_req_len annos at $anno_start\n";
218 : olson 1.22 print $log_fh "Retrieve $anno_req_len annos at $anno_start\n";
219 : olson 1.21
220 :     my $annos = $peer->get_annotations($session, $anno_start, $anno_req_len);
221 : olson 1.15
222 : olson 1.21 for my $anno (@$annos)
223 :     {
224 :     my($his_id, $ts, $author, $anno) = @$anno;
225 :    
226 :     my $my_id = $peg_mapping{$his_id};
227 :     next unless $my_id;
228 : olson 1.15
229 : olson 1.21 my $genome = $fig->genome_of($my_id);
230 :    
231 :     push(@{$genome_annos{$genome}}, [$my_id, $ts, $author, $anno]);
232 :     }
233 : olson 1.15 }
234 :    
235 : olson 1.19 #
236 :     # Do the same for the assignments
237 :     #
238 :    
239 : olson 1.20 # print Dumper($assignments);
240 :    
241 : olson 1.21
242 :     for (my $assign_start = 0; $assign_start < $num_assignments; $assign_start += $assign_batch_size)
243 : olson 1.19 {
244 : olson 1.21 my $assign_req_len = $num_assignments - $assign_start;
245 :     $assign_req_len = $assign_batch_size if $assign_req_len > $assign_batch_size;
246 : olson 1.19
247 : olson 1.21 print "Retrieve $assign_req_len assigns at $assign_start\n";
248 : olson 1.22 print $log_fh "Retrieve $assign_req_len assigns at $assign_start\n";
249 : olson 1.21
250 :     my $assignments = $peer->get_assignments($session, $assign_start, $assign_req_len);
251 :    
252 :     for my $assign (@$assignments)
253 :     {
254 :     my($his_id, $ts, $author, $func) = @$assign;
255 : olson 1.19
256 : olson 1.21 my $my_id = $peg_mapping{$his_id};
257 :     next unless $my_id;
258 : olson 1.19
259 : olson 1.21 my $genome = $fig->genome_of($my_id);
260 : olson 1.19
261 : olson 1.21 $genome_assignments{$genome}->{$my_id} = [$my_id, $ts, $author, $func];
262 :     }
263 : olson 1.19 }
264 :    
265 :     # print Dumper(\%genome_annos);
266 : olson 1.15
267 :     #
268 :     # Now install annotations.
269 :     #
270 :    
271 :     for my $genome (keys(%genome_annos))
272 :     {
273 : olson 1.19 #
274 :     # Plan: Apply the merge_annotations.pl logic. Read the annotations
275 :     # from the per-org annotations file, add the new ones here, sort, and remove duplicates.
276 :     # Write the results to the annotations file.
277 :     #
278 :     # When we are all done, rerun the index_annotations script.
279 :     #
280 :     # Why not do that incrementally? Partly because the annotation_seeks table doesn't
281 :     # have a column for the genome id, so a removal of old data would require a
282 :     # string-match query; since a complete reindex of the annotations is pretty
283 :     # fast (60 sec on a G4 laptop on a firewire disk), it's not clear whether the incremental
284 :     # update would actually be a win.
285 :     #
286 :    
287 :     my @annos = @{$genome_annos{$genome}};
288 :     my $assignments = $genome_assignments{$genome};
289 :     #
290 :     # %assignment_annos is a hash from peg to the list
291 :     # of annotations for that peg.
292 :     #
293 :     my %assignment_annos;
294 :    
295 :     my $dir = "$FIG_Config::organisms/$genome";
296 :     my $anno_file = "$dir/annotations";
297 :     my $anno_bak = "$dir/annotations." . time;
298 :    
299 :     my $new_count = @annos;
300 :    
301 :     #
302 :     # Rename the annotations file to a new name based on the current time.
303 :     #
304 :    
305 : olson 1.23 my $gs = $fig->genus_species($genome);
306 :     print $html_fh "<h1>Updates for $genome ($gs)</h1>\n";
307 :    
308 : olson 1.19 if (-f $anno_file)
309 :     {
310 :     rename($anno_file, $anno_bak) or die "Cannot rename $anno_file to $anno_bak: $!";
311 : olson 1.22 print $log_fh "Moved annotations file $anno_file to backup $anno_bak\n";
312 : olson 1.19 }
313 :    
314 :     if (open(my $fh, "<$anno_bak"))
315 :     {
316 :     #
317 :     # While we are scanning here, we look for the latest local assignment
318 :     # for any peg for which we are installing an assignment.
319 :     #
320 :     local($/) = "\n//\n";
321 :    
322 :     my($chunk, $peg, $ts, $author, $anno);
323 :    
324 :     while (defined($chunk = <$fh>))
325 :     {
326 :     chomp $chunk;
327 :     ($peg, $ts, $author, $anno) = split(/\n/, $chunk, 4);
328 :    
329 :     if ($peg =~ /^fig\|/ and $ts =~ /^\d+$/)
330 :     {
331 : olson 1.26 #
332 :     # The last field marks this as an "old" annotation, so we don't
333 :     # log its installation later.
334 :     #
335 :     my $ent = [$peg, $ts, $author, $anno, 1];
336 : olson 1.19 push(@annos, $ent);
337 :    
338 :     if (defined($assignments->{$peg}))
339 :     {
340 :     #
341 :     # We have an incoming assignment for this peg.
342 :     # Don't parse anything yet, but push the annotation
343 :     # on a list so we can sort by date.
344 :     #
345 :     push(@{$assignment_annos{$peg}}, $ent);
346 :     }
347 :     }
348 :     }
349 :     close($fh);
350 :     }
351 :    
352 :     #
353 :     # Determine if we are going to install an assignment.
354 :     #
355 :    
356 : olson 1.23 my $cgi_url = &FIG::cgi_url();
357 :     print $html_fh "<h2>Assignments made</h2>\n";
358 : olson 1.25 print $html_fh "<table border=\"1\">\n";
359 : olson 1.23 print $html_fh "<tr><th>PEG</th><th>Old assignment</th><th>New assignment</th><tr>\n";
360 :    
361 : olson 1.19 for my $peg (keys %$assignments)
362 :     {
363 : olson 1.20 my(undef, $ts, $author, $func) = @{$assignments->{$peg}};
364 :    
365 :     #
366 :     # Sort the existing annotations for this peg by date.
367 :     #
368 : olson 1.26 # Recall that this list has entries [$peg, $timestamp, $author, $anno, $old_flag]
369 : olson 1.20 #
370 :    
371 :     my @eannos;
372 :     if (ref($assignment_annos{$peg}))
373 :     {
374 :     @eannos = sort { $b->[1] <=> $a->[1] } @{$assignment_annos{$peg}};
375 :     }
376 :     else
377 :     {
378 :     #
379 :     # No assignment annotations found.
380 :     #
381 :     @eannos = ();
382 :     }
383 :    
384 :     # print "Assignment annos for $peg: ", Dumper(\@eannos);
385 : olson 1.19
386 :     #
387 : olson 1.20 # Filter out just the master assignments that are newer than
388 :     # the one we are contemplating putting in place.
389 : olson 1.19 #
390 :    
391 : olson 1.20 my @cand = grep {
392 :     ($_->[1] > $ts) and ($_->[3] =~ /Set master function to/)
393 :     } @eannos;
394 :    
395 :     if (@cand > 0)
396 :     {
397 :     #
398 :     # Here is were some policy needs to be put in place --
399 :     # we have a more recent annotation on the current system.
400 :     #
401 :     # For now, we will not install an assignment if there is any
402 :     # newer assignment in place.
403 :     #
404 :    
405 :     warn "Skipping assignment for $peg $func due to more recent assignment $cand[0]->[3]\n";
406 : olson 1.22 print $log_fh "Skipping assignment for $peg $func due to more recent assignment $cand[0]->[3]\n";
407 : olson 1.20 }
408 :     else
409 :     {
410 :     #
411 :     # Nothing is blocking us. While we are testing, just slam this assignment in.
412 :     #
413 : olson 1.19
414 : olson 1.20 my $old = $fig->function_of($peg, 'master');
415 :    
416 : olson 1.25 if ($old ne $func and &$allow_assignment($peg, $ts, $author, $func))
417 : olson 1.21 {
418 : olson 1.23 my $l = "$cgi_url/protein.cgi?prot=$peg";
419 :     print $html_fh "<tr><td><a href=\"$l\">$peg</a></td><td>$old</td><td>$func</td></tr>\n";
420 :    
421 : olson 1.21 print "Assign $peg $func\n";
422 : olson 1.22 print $log_fh "Assign $peg $func\n";
423 :     print $log_fh " was $old\n";
424 : olson 1.21 $fig->assign_function($peg, 'master', $func);
425 : olson 1.23
426 : olson 1.21 }
427 : olson 1.20 }
428 : olson 1.19 }
429 :    
430 : olson 1.23 print $html_fh "</table>\n";
431 :    
432 :     print $html_fh "<h2>Annotations added</h2>\n";
433 : olson 1.25 print $html_fh "<table border=\"1\">\n";
434 : olson 1.23 print $html_fh "<tr><th>PEG</th><th>Time</th><th>Author</th><th>Annotation</th></tr>\n";
435 :    
436 : olson 1.19 open(my $outfh, ">$anno_file") or die "Cannot open new annotation file $anno_file: $!\n";
437 :    
438 :     my $last;
439 :     my @sorted = sort { ($a->[0] cmp $b->[0]) or ($a->[1] <=> $b->[1]) } @annos;
440 :     my $inst = 0;
441 :     my $dup = 0;
442 :     foreach my $ann (@sorted)
443 :     {
444 :     my $txt = join("\n", @$ann);
445 :     #
446 :     # Drop the trailing \n if there is one; we will add it back when we print and
447 :     # want to ensure the file format remains sane.
448 :     #
449 :     chomp $txt;
450 :     if ($txt ne $last)
451 :     {
452 : olson 1.23 my $peg = $ann->[0];
453 :     my $l = "$cgi_url/protein.cgi?prot=$peg";
454 : olson 1.26 if (!$ann->[4])
455 :     {
456 :     print $html_fh "<tr>" . join("\n", map { "<td>$_</td>" }
457 :     "<a href=\"$l\">$peg</a>",
458 :     scalar(localtime($ann->[1])), $ann->[2], $ann->[3])
459 :     . "</tr>\n";
460 :     }
461 : olson 1.23
462 : olson 1.19 print $outfh "$txt\n//\n";
463 :     $last = $txt;
464 : olson 1.21 # print "Inst $ann->[0] $ann->[1] $ann->[2]\n";
465 : olson 1.19 $inst++;
466 :     }
467 :     else
468 :     {
469 : olson 1.21 # print "Dup $ann->[0] $ann->[1] $ann->[2]\n";
470 : olson 1.19 $dup++;
471 :     }
472 :     }
473 : olson 1.23 print $html_fh "</table>\n";
474 : olson 1.19 close($outfh);
475 :     chmod(0666, $anno_file) or warn "Cannot chmod 0666 $anno_file: $!\n";
476 :     print "Wrote $anno_file. $new_count new annos, $inst installed, $dup duplicates\n";
477 : olson 1.22 print $log_fh "Wrote $anno_file. $new_count new annos, $inst installed, $dup duplicates\n";
478 : olson 1.15 }
479 : olson 1.23 close($html_fh);
480 : olson 1.1 }
481 :    
482 : olson 1.21 #
483 :     # Compute the peg mapping for a session.
484 :     #
485 :     # $fig Active FIG instance
486 :     # $peer P2P peer for this session.
487 :     # $session P2P session ID
488 :     # $peg_mapping Hash ref for the remote -> local PEG mapping
489 :     # $peg_cache Hash ref for the persistent remote -> local PEG mapping cache db.
490 : olson 1.24 # $cache_handle AnyDBM_File handle corresponding to $peg_cache.
491 : olson 1.21 #
492 :     sub _compute_peg_mapping
493 :     {
494 :     my($fig, $peer, $session, $num_pegs, $peg_mapping, $peg_cache, $cache_handle, $skip_tough_search) = @_;
495 :    
496 :     #
497 :     # genome_map is a hash mapping from target genome id to a list of
498 :     # pegs on the target. This is used to construct a finalize_pegs request after
499 :     # the first phase of peg mapping.
500 :     #
501 :    
502 :     my %genome_map;
503 :    
504 :     #
505 :     # target_genome_info is a hash mapping from target genome
506 :     # identifier to the target-side information on the genome -
507 :     # number of contigs, number of nucleotides, checksum.
508 :     #
509 :     # We accumulate it here across possibly multiple batches of
510 :     # peg retrievals in order to create a single finalization
511 :     # list.
512 :     #
513 :    
514 :     my %target_genome_info;
515 :    
516 :     #
517 :     # For very large transfers, we need to batch the peg processing.
518 :     #
519 :    
520 :     for (my $peg_start = 0; $peg_start < $num_pegs; $peg_start += $peg_batch_size)
521 :     {
522 :     my $peg_req_len = $num_pegs - $peg_start;
523 :     $peg_req_len = $peg_batch_size if $peg_req_len > $peg_batch_size;
524 :    
525 :     print "Getting $peg_req_len pegs at $peg_start\n";
526 : olson 1.22 print $log_fh "Getting $peg_req_len pegs at $peg_start\n";
527 : olson 1.21 my $ret = $peer->get_pegs($session, $peg_start, $peg_req_len);
528 :    
529 :     if (!$ret or ref($ret) ne "ARRAY")
530 :     {
531 :     die "perform_update: get_pegs failed\n";
532 :     }
533 :    
534 :     my($peg_list, $genome_list) = @$ret;
535 :    
536 :     for my $gent (@$genome_list)
537 :     {
538 :     $target_genome_info{$gent->[0]} = $gent;
539 :     }
540 :    
541 :     _compute_peg_mapping_batch($fig, $peer, $session, $peg_mapping, $peg_cache, $cache_handle,
542 :     $peg_list, \%genome_map);
543 :     }
544 :    
545 :     #
546 :     # We have finished first pass. Now go over the per-genome mappings that need to be made.
547 :     #
548 :     # $genome_map{$genome_id} is a list of pegs that reside on that genome.
549 :     # The pegs and genome id are both target-based identifiers.
550 :     #
551 :     # %target_genome_info defines the list of genome information we have on the remote
552 :     # side.
553 :     #
554 :     # We build a request to be passed to finalize_pegs. Each entry in the request is either
555 :     # ['peg_genome', $peg] which means that we have a genome that corresponds to the
556 :     # genome the peg is in. We can attempt to map via contig locations.
557 :     #
558 :     # If that is not the case, we pass a request entry of ['peg_unknown', $peg]
559 :     # which will result in the sequence data being returned.
560 :     #
561 :    
562 :     my @finalize_req = ();
563 :    
564 :     #
565 :     # local_genome maps a target peg identifier to the local genome id it translates to.
566 :     #
567 :     my %local_genome;
568 :    
569 :     for my $genome (keys(%target_genome_info))
570 :     {
571 :     my($tg, $n_contigs, $n_nucs, $cksum) = @{$target_genome_info{$genome}};
572 :    
573 :     $tg eq $genome or die "Invalid entry in target_genome_info for $genome => $tg, $n_contigs, $n_nucs, $cksum";
574 :    
575 :     #
576 :     # Don't bother unless we have any pegs to look up.
577 :     #
578 :     next unless defined($genome_map{$genome});
579 :    
580 :     #
581 :     # Determine if we have a local genome installed that matches precisely the
582 :     # genome on the target side.
583 :     #
584 :     my $my_genome = $fig->find_genome_by_content($genome, $n_contigs, $n_nucs, $cksum);
585 :    
586 :     my $pegs = $genome_map{$genome};
587 :    
588 :     if ($my_genome)
589 :     {
590 :     #
591 :     # We do have such a local genome. Generate a peg_genome request to
592 :     # get the location information from the target side.
593 :     #
594 :     # Also remember the local genome mapping for this peg.
595 :     #
596 :    
597 :     print "$genome mapped to $my_genome\n";
598 : olson 1.22 print $log_fh "$genome mapped to $my_genome\n";
599 : olson 1.21 for my $peg (@$pegs)
600 :     {
601 :     push(@finalize_req, ['peg_genome', $peg]);
602 :     $local_genome{$peg} = $my_genome;
603 :     }
604 :    
605 :     }
606 :     else
607 :     {
608 :     #
609 :     # We don't have such a genome. We need to retrieve the
610 :     # sequence data in order to finish mapping.
611 :     #
612 :     push(@finalize_req, map { ['peg_unknown', $_] } @$pegs);
613 :     }
614 :     }
615 :    
616 :     #
617 :     # We've built our finalization request. Handle it (possibly with batching here too).
618 :     #
619 :    
620 :     _process_finalization_request($fig, $peer, $session, $peg_mapping, $peg_cache, $cache_handle,
621 :     \%local_genome, \@finalize_req, $skip_tough_search);
622 :    
623 :     }
624 :    
625 :     #
626 :     # Process one batch of PEGs.
627 :     #
628 :     # Same args as _compute_peg_mapping, with the addition of:
629 :     #
630 :     # $peg_list List of pegs to be processed
631 :     # $genome_map Hash maintaining list of genomes with their pegs.
632 :     # $target_genome_info Hash maintaining overall list of target-side genome information.
633 :     #
634 :     sub _compute_peg_mapping_batch
635 :     {
636 :     my($fig, $peer, $session, $peg_mapping, $peg_cache, $cache_handle,
637 :     $peg_list, $genome_map, $target_genome_info) = @_;
638 :    
639 :     #
640 :     # Walk the list of pegs as returned from get_pegs() and determine what has to
641 :     # be done.
642 :     #
643 :     # If the entry is ['peg', $peg], we can use the peg ID as is.
644 :     #
645 :     # If the entry is ['peg_info', $peg, $alias_list, $genome], the peg
646 :     # has the given aliases, and is in the given genome.
647 :     #
648 :     for my $peg_info (@$peg_list)
649 :     {
650 :     my($key, $peg, @rest) = @$peg_info;
651 :    
652 :     if ($key eq 'peg')
653 :     {
654 :     #
655 :     # Peg id is directly usable.
656 :     #
657 :     $peg_mapping->{$peg} = $peg;
658 :     }
659 :     elsif ($key eq 'peg_info')
660 :     {
661 :     #
662 :     # Peg id not directly usable. See if we have it in the cache.
663 :     #
664 :    
665 :     if ((my $cached = $peg_cache->{$peg}) ne "")
666 :     {
667 :     #
668 :     # Cool, we've cached the result. Use it.
669 :     #
670 :    
671 :     $peg_mapping->{$peg} = $cached;
672 :     # warn "Found cached mapping $peg => $cached\n";
673 :     next;
674 :     }
675 :    
676 :     #
677 :     # It is not cached. Attempt to resolve by means of alias IDs.
678 :     #
679 :    
680 :     my($alias_list, $genome_id) = @rest;
681 :    
682 :     for my $alias (@$alias_list)
683 :     {
684 :     my $mapped = $fig->by_alias($alias);
685 :     if ($mapped)
686 :     {
687 :     print "$peg maps to $mapped via $alias\n";
688 : olson 1.22 print $log_fh "$peg maps to $mapped via $alias\n";
689 : olson 1.21 $peg_mapping->{$peg}= $mapped;
690 :     $peg_cache->{$peg} = $mapped;
691 :     last;
692 :     }
693 :     }
694 :    
695 :     #
696 :     # If we weren't able to resolve by ID,
697 :     # add to %genome_map as a PEG that will need
698 :     # to be resolved by means of contig location.
699 :     #
700 :    
701 :     if (!defined($peg_mapping->{$peg}))
702 :     {
703 :     push(@{$genome_map->{$genome_id}}, $peg);
704 :     print "$peg did not map on first pass\n";
705 : olson 1.22 print $log_fh "$peg did not map on first pass\n";
706 : olson 1.21 }
707 :     }
708 :     }
709 :    
710 :     #
711 :     # Flush the cache to write out any computed mappings.
712 :     #
713 : olson 1.24 eval { $cache_handle->sync();};
714 : olson 1.21
715 :     }
716 :    
717 :     sub _process_finalization_request
718 :     {
719 :     my($fig, $peer, $session, $peg_mapping, $peg_cache, $cache_handle,
720 :     $local_genome, $finalize_req, $skip_tough_search) = @_;
721 :    
722 :     #
723 :     # Immediately return unless there's something to do.
724 :     #
725 :     return unless ref($finalize_req) and @$finalize_req > 0;
726 : olson 1.1
727 : olson 1.21 while (@$finalize_req > 0)
728 :     {
729 :     my @req = splice(@$finalize_req, 0, $fin_batch_size);
730 :    
731 :     print "Invoking finalize_pegs on ", int(@req), " pegs\n";
732 : olson 1.22 print $log_fh "Invoking finalize_pegs on ", int(@req), " pegs\n";
733 : olson 1.21 my $ret = $peer->finalize_pegs($session, \@req);
734 :    
735 :     if (!$ret or ref($ret) ne "ARRAY")
736 :     {
737 :     die "perform_update: finalize_pegs failed\n";
738 :     }
739 :    
740 :     #
741 :     # The return is a list of either location entries or
742 :     # sequence data. Attempt to finish up the mapping.
743 :     #
744 :    
745 :     my(%sought, %sought_seq);
746 :    
747 :    
748 :     my $dbh = $fig->db_handle();
749 :     for my $entry (@$ret)
750 :     {
751 :     my($what, $peg, @rest) = @$entry;
752 :    
753 :     if ($what eq "peg_loc")
754 :     {
755 :     my($strand, $start, $end, $cksum, $seq) = @rest;
756 :    
757 :     #
758 :     # We have a contig location. Try to find a matching contig
759 :     # here, and see if it maps to something.
760 :     #
761 :    
762 :     my $my_genome = $local_genome->{$peg};
763 :     my $local_contig = $fig->find_contig_with_checksum($my_genome, $cksum);
764 :     if ($local_contig)
765 :     {
766 :     #
767 :     # Now look up the local peg. We match on the end location; depending on the strand
768 :     # the feature is on, we want to look at either minloc or maxloc.
769 :     #
770 :    
771 : olson 1.27 my($start_loc, $end_loc);
772 :    
773 :     if ($strand eq '-')
774 :     {
775 :     $start_loc = 'maxloc';
776 :     $end_loc = 'minloc';
777 :     }
778 :     else
779 :     {
780 :     $start_loc = 'minloc';
781 :     $end_loc = 'maxloc';
782 :     }
783 : olson 1.21
784 : olson 1.27 my $res = $dbh->SQL(qq!SELECT id, $start_loc from features
785 :     WHERE $end_loc = $end and genome = '$my_genome' and
786 : olson 1.21 contig = '$local_contig'
787 :     !);
788 :    
789 :     if ($res and @$res > 0)
790 :     {
791 : olson 1.27 my $id;
792 :     if (@$res == 1)
793 :     {
794 :     #
795 :     # Found a unique mapping.
796 :     #
797 :     $id = $res->[0]->[0];
798 :     }
799 :     else
800 :     {
801 :     #
802 :     # Multiple mappings found. See if one matches the
803 :     # start location. If it doesn't, pick the one that
804 :     # is closest in length.
805 :     #
806 :    
807 :     my @lens;
808 :    
809 :     for my $res_ent (@$res)
810 :     {
811 :     my($rid, $rloc) = @$res_ent;
812 :    
813 :     push(@lens, [$rid, abs($rloc - $end - ($start - $end))]);
814 :     warn "Matching $rid $rloc to $start\n";
815 :     if ($rloc == $start)
816 :     {
817 :     $id = $rid;
818 :     warn "Matched $rid\n";
819 :     last;
820 :     }
821 :     }
822 :    
823 :     if (!$id)
824 :     {
825 :     my @slens = sort { $a->[1] <=> $b->[1]} @lens;
826 :     my $len;
827 :     ($id, $len) = @{$slens[0]};
828 :     warn "No unique match found, picking closest match $id (len=$len)\n";
829 :     }
830 :     }
831 :    
832 : olson 1.21 $peg_mapping->{$peg} = $id;
833 :     $peg_cache->{$peg} = $id;
834 :     print "Mapped $peg to $id via contigs\n";
835 :     }
836 :     else
837 :     {
838 :     print "failed: $peg $my_genome and contig $local_contig start=$start end=$end strand=$strand\n";
839 : olson 1.22 print $log_fh "failed: $peg $my_genome and contig $local_contig start=$start end=$end strand=$strand\n";
840 : olson 1.23 print $html_fh "Contig match failed: $peg $my_genome contig $local_contig start $start end $end strand $strand<br>\n";
841 : olson 1.21 $sought{$peg}++;
842 :     $sought_seq{$peg} = $seq;
843 :     }
844 :     }
845 :     else
846 :     {
847 :     print "Mapping failed for $my_genome checksum $cksum\n";
848 : olson 1.22 print $log_fh "Mapping failed for $my_genome checksum $cksum\n";
849 : olson 1.23 print $html_fh "Mapping failed for $my_genome checksum $cksum<br>\n";
850 : olson 1.21 $sought{$peg}++;
851 :     $sought_seq{$peg} = $seq;
852 :     }
853 :     }
854 :     elsif ($what eq "peg_seq")
855 :     {
856 :     my($seq) = @rest;
857 :    
858 :     $sought{$peg}++;
859 :     $sought_seq{$peg} = $seq;
860 :     }
861 :     }
862 :    
863 :     #
864 :     # Now see if we need to do a tough search.
865 :     #
866 :    
867 :     if (keys(%sought) > 0 and !$skip_tough_search)
868 :     {
869 :     my %trans;
870 :    
871 :     print "Starting tough search\n";
872 : olson 1.22 print $log_fh "Starting tough search\n";
873 : olson 1.21
874 :     $fig->tough_search(undef, \%sought_seq, \%trans, \%sought);
875 :     print "Tough search translated: \n";
876 : olson 1.22 print $log_fh "Tough search translated: \n";
877 : olson 1.21 while (my($tpeg, $ttrans) = each(%trans))
878 :     {
879 :     print " $tpeg -> $ttrans\n";
880 : olson 1.22 print $log_fh " $tpeg -> $ttrans\n";
881 : olson 1.21 $peg_mapping->{$tpeg} = $ttrans;
882 :     $peg_cache->{$tpeg} = $ttrans;
883 :     }
884 :     }
885 :     }
886 :     }
887 : olson 1.15
888 : olson 1.1 #############
889 :     #
890 :     # P2P Relay
891 :     #
892 :     #############
893 :    
894 :    
895 :     package P2P::Relay;
896 :     use strict;
897 :    
898 :     use Data::Dumper;
899 :     use SOAP::Lite;
900 :    
901 :     use P2P;
902 :    
903 :     sub new
904 :     {
905 :     my($class, $url) = @_;
906 :    
907 : olson 1.25 my $creds = [];
908 : olson 1.24
909 :     my $proxy = SOAP::Lite->uri($P2P::ns_relay)->proxy([$url,
910 :     credentials => $creds]);
911 : olson 1.1
912 :     my $self = {
913 :     url => $url,
914 :     proxy => $proxy,
915 :     };
916 :     return bless($self, $class);
917 :     }
918 :    
919 :     sub enumerate_annotation_systems
920 :     {
921 :     my($self) = @_;
922 :    
923 :     return $self->{proxy}->enumerate_annotation_systems()->result;
924 :     }
925 :    
926 :     sub fetch_queries
927 :     {
928 :     my($self, $id) = @_;
929 :    
930 :     my $reply = $self->{proxy}->fetch_queries($id);
931 :    
932 :     if ($reply->fault)
933 :     {
934 :     print "Failed to fetch queries: ", $reply->faultcode, " ", $reply->faultstring, "\n";
935 :     return undef;
936 :     }
937 :    
938 :     return $reply->result;
939 :     }
940 :    
941 :     sub deposit_answer
942 :     {
943 :     my($self, $id, $key, $answer) = @_;
944 :    
945 :     my $reply = $self->{proxy}->deposit_answer($id, $key,
946 :     SOAP::Data->type('base64')->value($answer));
947 :    
948 :     if ($reply->fault)
949 :     {
950 :     print "deposit_answer got fault: ", $reply->faultcode, " ", $reply->faultstring, "\n";
951 :     return undef;
952 :     }
953 :    
954 :     return $reply;
955 :     }
956 :    
957 :     =pod
958 :    
959 :     =head1 await_result
960 :    
961 :     Await the result from a possibly-asynchronous soap request.
962 :    
963 :     Look at the reply that we have. If it's a deferred reply, loop polling
964 :     the relay for the actual result.
965 :    
966 :     We determine if the reply is a deferred reply by examining the namespace
967 :     URI of the response. A response will be generated from the relay's namespace,
968 :     rather than that of the application itself.
969 :    
970 :     =cut
971 :    
972 :     sub await_result
973 :     {
974 :     my($self, $reply) = @_;
975 :    
976 :     while (1)
977 :     {
978 :     #
979 :     # Retrieve the namespace of the response, which is the first
980 :     # element in the body of the message.
981 :     #
982 :     my $ns = $reply->namespaceuriof('/Envelope/Body/[1]');
983 : olson 1.20 # print "Reply ns=$ns want $P2P::ns_relay\n";
984 : olson 1.1
985 :     if ($ns eq $P2P::ns_relay)
986 :     {
987 :     my $val = $reply->result;
988 : olson 1.20 # print "got val=", Dumper($val);
989 : olson 1.1 if ($val->[0] eq 'deferred')
990 :     {
991 :     #
992 :     # Sleep a little, then try to retrieve the response.
993 :     #
994 :    
995 :     sleep(1);
996 :     my $id = $val->[1];
997 :    
998 :     print "Retrieving reply\n";
999 :     $reply = $self->{proxy}->call_completed($id);
1000 :     }
1001 :     else
1002 :     {
1003 :     #
1004 :     # We're not sure what to do here..
1005 :     #
1006 :     return undef;
1007 :     }
1008 :     }
1009 :     else
1010 :     {
1011 :     #
1012 :     # We got an actual response. Return it.
1013 :     #
1014 :    
1015 :     return $reply;
1016 :     }
1017 :     }
1018 :     }
1019 :    
1020 :     #############
1021 :     #
1022 :     # P2P Requestor
1023 :     #
1024 :     #############
1025 :    
1026 :     package P2P::Requestor;
1027 :     use strict;
1028 :    
1029 :     use Data::Dumper;
1030 : olson 1.15 use Time::HiRes qw( usleep ualarm gettimeofday tv_interval );
1031 : olson 1.1
1032 :     use SOAP::Lite;
1033 : olson 1.15
1034 :     #use SOAP::Lite +trace => [qw(transport dispatch result debug)];
1035 : olson 1.1 use P2P;
1036 :    
1037 :     #
1038 :     # Create a new Requestor. It contains a reference to the FIG instance
1039 :     # so that we can run the protocol completely from in here.
1040 :     #
1041 :    
1042 :     sub new
1043 :     {
1044 : olson 1.24 my($class, $fig, $url, $peer_id, $relay, $credentials) = @_;
1045 : olson 1.1
1046 : olson 1.24 $credentials = [] unless ref($credentials);
1047 :    
1048 : olson 1.27 my $proxy = SOAP::Lite->uri($ns_p2p)->proxy($url, timeout => 3600);
1049 :    
1050 :     for my $cred (@$credentials)
1051 :     {
1052 :     $proxy->transport->credentials(@$cred);
1053 :     }
1054 : olson 1.1
1055 :     my $self = {
1056 :     fig => $fig,
1057 :     url => $url,
1058 :     peer_id => $peer_id,
1059 :     proxy => $proxy,
1060 :     relay => $relay,
1061 :     };
1062 :     return bless($self, $class);
1063 :     }
1064 :    
1065 :     #
1066 :     # First step: Request an update.
1067 :     #
1068 :     # We need to determine some notion of what our release is, since we are not
1069 :     # currently tagging them explicitly. Until we delve into this more,
1070 :     # I am going to return a null release, which means the same-release
1071 :     # optimization won't be able to kick in.
1072 :     #
1073 :     # We also need to determine the last time we got an update from this
1074 :     # system.
1075 :     #
1076 :    
1077 :     sub request_update
1078 :     {
1079 : olson 1.20 my($self, $last_update, $update_thru) = @_;
1080 : olson 1.1
1081 : olson 1.18 my $rel = [$self->{fig}->get_release_info()];
1082 : olson 1.1
1083 :     if (!defined($last_update))
1084 :     {
1085 :     $last_update = $self->{fig}->get_peer_last_update($self->{peer_id});
1086 :     }
1087 : olson 1.17
1088 :     print "Requesting update via $self->{proxy}\n";
1089 : olson 1.20 my $reply = $self->{proxy}->request_update($rel, $last_update, $update_thru);
1090 :     # print "Got reply ", Dumper($reply);
1091 : olson 1.1
1092 :     if ($self->{relay})
1093 :     {
1094 :     $reply = $self->{relay}->await_result($reply);
1095 :     }
1096 :    
1097 :     if ($reply->fault)
1098 :     {
1099 :     print "request_update triggered fault: ", $reply->faultcode, " ", $reply->faultstring, "\n";
1100 :     return undef;
1101 :     }
1102 :    
1103 :     return $reply->result;
1104 :     }
1105 :    
1106 :     =pod
1107 :    
1108 :     =head1 get_pegs($session_id, $start, $length)
1109 :    
1110 :    
1111 :     =cut
1112 :    
1113 :     sub get_pegs
1114 :     {
1115 :     my($self, $session_id, $start, $length) = @_;
1116 :    
1117 :     return $self->call("get_pegs", $session_id, $start, $length);
1118 :     }
1119 :    
1120 : olson 1.6 sub finalize_pegs
1121 :     {
1122 :     my($self, $session_id, $request) = @_;
1123 :    
1124 :     return $self->call("finalize_pegs", $session_id, $request);
1125 :     }
1126 :    
1127 : olson 1.15 sub get_annotations
1128 :     {
1129 :     my($self, $session_id, $start, $length) = @_;
1130 :    
1131 :     return $self->call("get_annotations", $session_id, $start, $length);
1132 :     }
1133 :    
1134 : olson 1.19 sub get_assignments
1135 :     {
1136 :     my($self, $session_id, $start, $length) = @_;
1137 :    
1138 :     return $self->call("get_assignments", $session_id, $start, $length);
1139 :     }
1140 :    
1141 : olson 1.1 sub call
1142 :     {
1143 :     my($self, $func, @args) = @_;
1144 : olson 1.15
1145 :     my $t0 = [gettimeofday()];
1146 :     print "Calling $func\n";
1147 : olson 1.1 my $reply = $self->{proxy}->$func(@args);
1148 : olson 1.15 my $t1 = [gettimeofday()];
1149 :    
1150 :     my $elap = tv_interval($t0, $t1);
1151 :     print "Call to $func took $elap\n";
1152 : olson 1.1
1153 :     if ($self->{relay})
1154 :     {
1155 :     $reply = $self->{relay}->await_result($reply);
1156 :     }
1157 :    
1158 :     if ($reply->fault)
1159 :     {
1160 :     print "$func triggered fault: ", $reply->faultcode, " ", $reply->faultstring, "\n";
1161 :     return undef;
1162 :     }
1163 :    
1164 :     return $reply->result;
1165 :     }
1166 :    
1167 :    
1168 :     #############
1169 :     #
1170 :     # P2P Service
1171 :     #
1172 :     # Code in this module is invoked on the target on behalf of a requestor.
1173 :     #
1174 :     #############
1175 :    
1176 :     package P2P::Service;
1177 :    
1178 :     use Data::Dumper;
1179 :    
1180 :     use FIG;
1181 :     use FIG_Config;
1182 :     use strict;
1183 :    
1184 :     use File::Temp qw(tempdir);
1185 :     use File::Basename;
1186 :    
1187 :     sub request_update
1188 :     {
1189 : olson 1.20 my($class, $his_release, $last_update, $update_thru)= @_;
1190 : olson 1.1
1191 :     #
1192 :     # Verify input.
1193 :     #
1194 :    
1195 :     if ($last_update !~ /^\d+$/)
1196 :     {
1197 :     die "request_update: last_update must be a number (not '$last_update')\n";
1198 :     }
1199 :    
1200 : olson 1.20 if ($update_thru eq "")
1201 :     {
1202 :     $update_thru = time + 10000;
1203 :     }
1204 :    
1205 : olson 1.1 #
1206 :     # Create a new session id and a spool directory to use for storage
1207 :     # of information about it. This can go in the tempdir since it is
1208 :     # not persistent.
1209 :     #
1210 :    
1211 :     &FIG::verify_dir("$FIG_Config::temp/p2p_spool");
1212 : olson 1.22 my $spool_dir = tempdir(DIR => "$FIG_Config::temp/p2p_spool");
1213 : olson 1.1
1214 : olson 1.22 #my $spool_dir = "$FIG_Config::temp/p2p_spool/test";
1215 : olson 1.1 &FIG::verify_dir($spool_dir);
1216 :    
1217 :     my $session_id = basename($spool_dir);
1218 :     my $now = time;
1219 :    
1220 :     #
1221 :     # Gather the list of pegs and annotations for the update.
1222 :     #
1223 :    
1224 :     my $fig = new FIG;
1225 :    
1226 :     my $all_genomes = [$fig->genomes];
1227 :    
1228 :     my %all_genomes = map { $_ => 1 } @$all_genomes;
1229 :    
1230 :     my %pegs;
1231 : olson 1.15
1232 :     #
1233 :     # We keep track of usernames that have been seen, so that
1234 :     # we can both update our local user database and
1235 :     # we can report them to our peer.
1236 :     #
1237 :    
1238 :     my %users;
1239 : olson 1.1
1240 :     my $num_annos = 0;
1241 :     my $num_genomes = 0;
1242 :     my $num_pegs = 0;
1243 : olson 1.15 my $num_assignments = 0;
1244 : olson 1.1
1245 :     my $anno_fh;
1246 :     open($anno_fh, ">$spool_dir/annos");
1247 :    
1248 :     my $peg_fh;
1249 :     open($peg_fh, ">$spool_dir/pegs");
1250 :    
1251 :     my $genome_fh;
1252 :     open($genome_fh, ">$spool_dir/genomes");
1253 :    
1254 : olson 1.15 my $assign_fh;
1255 :     open($assign_fh, ">$spool_dir/assignments");
1256 :    
1257 : olson 1.25 #
1258 :     # We originally used a query to get the PEGs that needed to have annotations
1259 :     # sent. Unfortunately, this performed very poorly due to all of the resultant
1260 :     # seeking around in the annotations files.
1261 :     #
1262 :     # The code below just runs through all of the anno files looking for annos.
1263 :     #
1264 :     # A better way to do this would be to do a query to retrieve the genome id's for
1265 :     # genomes that have updates. The problem here is that the annotation_seeks
1266 :     # table doesn't have an explicit genome field.
1267 :     #
1268 :     # Surprisingly, to me anyway, the following query appers to run quickly, in both
1269 :     # postgres and mysql:
1270 :     #
1271 :     # SELECT distinct(substring(fid from 5 for position('.peg.' in fid) - 5))
1272 :     # FROM annotation_seeks
1273 :     # WHERE dateof > some-date.
1274 :     #
1275 :     # The output of that can be parsed to get the genome id and just those
1276 :     # annotations files searched.
1277 :     #
1278 :    
1279 : olson 1.1 for my $genome (@$all_genomes)
1280 :     {
1281 :     my $num_annos_for_genome = 0;
1282 : olson 1.15 my %assignment;
1283 : olson 1.1
1284 :     my $genome_dir = "$FIG_Config::organisms/$genome";
1285 :     next unless -d $genome_dir;
1286 :    
1287 :     my $afh;
1288 :     if (open($afh, "$genome_dir/annotations"))
1289 :     {
1290 :     my($fid, $anno_time, $who, $anno_text);
1291 :     local($/);
1292 :     $/ = "//\n";
1293 :     while (my $ann = <$afh>)
1294 :     {
1295 :     chomp $ann;
1296 :    
1297 :     if ((($fid, $anno_time, $who, $anno_text) =
1298 :     ($ann =~ /^(fig\|\d+\.\d+\.peg\.\d+)\n(\d+)\n(\S+)\n(.*\S)/s)) and
1299 : olson 1.20 $anno_time > $last_update and
1300 :     $anno_time < $update_thru)
1301 : olson 1.1
1302 :     {
1303 :     #
1304 : olson 1.15 # Update users list.
1305 :     #
1306 :    
1307 :     $users{$who}++;
1308 :    
1309 :     #
1310 : olson 1.1 # Look up aliases if we haven't seen this fid before.
1311 :     #
1312 :    
1313 :     if (!defined($pegs{$fid}))
1314 :     {
1315 :     my @aliases = $fig->feature_aliases($fid);
1316 :    
1317 :     print $peg_fh join("\t", $fid, $genome, @aliases), "\n";
1318 :     $num_pegs++;
1319 :     }
1320 :    
1321 :     print $anno_fh "$ann//\n";
1322 :    
1323 :     $pegs{$fid}++;
1324 :    
1325 :     $num_annos_for_genome++;
1326 :     $num_annos++;
1327 : olson 1.15
1328 :     #
1329 :     # While we're here, see if this is an assignment. We check in the
1330 :     # %assignment hash, which is keyed on fid, to see if we already
1331 :     # saw an assignment for this fid. If we have, we keep this one only if
1332 :     # the assignment time on it is later than the one we saw already.
1333 :     #
1334 :     # We are only looking at master assignments for now. We will need
1335 :     # to return to this issue and reexamine it, but in order to move
1336 :     # forward I am only matching master assignments.
1337 :     #
1338 :    
1339 :     if ($anno_text =~ /Set master function to\n(\S[^\n]+\S)/)
1340 :     {
1341 :     my $func = $1;
1342 :    
1343 :     my $other = $assignment{$fid};
1344 :    
1345 :     #
1346 :     # If we haven't seen an assignment for this fid,
1347 :     # or if it the other assignment has a timestamp that
1348 :     # is earlier than this one, set the assignment.
1349 :     #
1350 :    
1351 :     if (!defined($other) or
1352 :     ($other->[1] < $anno_time))
1353 :     {
1354 :     $assignment{$fid} = [$fid, $anno_time, $who, $func];
1355 :     }
1356 :     }
1357 : olson 1.1 }
1358 :     }
1359 :     close($afh);
1360 : olson 1.15
1361 :     #
1362 :     # Write out the assignments that remain.
1363 :     #
1364 :    
1365 :     for my $fid (sort keys(%assignment))
1366 :     {
1367 :     print $assign_fh join("\t", @{$assignment{$fid}}), "\n";
1368 :     $num_assignments++;
1369 :     }
1370 : olson 1.1 }
1371 : olson 1.15
1372 : olson 1.1
1373 :     #
1374 :     # Determine genome information if we have annotations for this one.
1375 :     #
1376 :    
1377 :     if ($num_annos_for_genome > 0)
1378 :     {
1379 :     $num_genomes++;
1380 :     if (open(my $cfh, "<$genome_dir/COUNTS"))
1381 :     {
1382 :     if ($_ = <$cfh>)
1383 :     {
1384 :     chomp;
1385 :     my($cgenome, $n_contigs, $total_nucs, $cksum) = split(/\t/, $_);
1386 :     if ($cgenome ne $genome)
1387 :     {
1388 :     warn "Hm, $genome has a COUNTS file with genome=$cgenome that does not match\n";
1389 :     }
1390 :     else
1391 :     {
1392 :     print $genome_fh join("\t",
1393 :     $genome, $num_annos_for_genome, $n_contigs,
1394 :     $total_nucs, $cksum), "\n";
1395 :     }
1396 :     }
1397 :     }
1398 :     }
1399 :    
1400 :     }
1401 :     close($anno_fh);
1402 :     close($peg_fh);
1403 :     close($genome_fh);
1404 : olson 1.15 close($assign_fh);
1405 : olson 1.1
1406 :     print "Pegs: $num_pegs\n";
1407 :     print "Genomes: $num_genomes\n";
1408 :     print "Annos: $num_annos\n";
1409 :    
1410 :     #
1411 :     # Check compatibility.
1412 :     #
1413 :    
1414 : olson 1.18 my $my_release = [$fig->get_release_info()];
1415 :    
1416 :     #
1417 :     # Release id is $my_release->[1].
1418 :     #
1419 :    
1420 :     my $compatible;
1421 :     if ($my_release->[1] ne "" and $his_release->[1] ne "")
1422 :     {
1423 :     #
1424 :     # Both releases must be defined for them to be compatible.
1425 :     #
1426 :     # At some point we need to consider the derived-release issue.
1427 :     #
1428 :    
1429 :     $compatible = $my_release->[1] eq $his_release->[1];
1430 :     }
1431 :     else
1432 :     {
1433 :     $compatible = 0;
1434 :     }
1435 : olson 1.1
1436 :     open(my $fh, ">$spool_dir/INFO");
1437 :     print $fh "requestor_release\t$his_release\n";
1438 :     print $fh "last_update\t$last_update\n";
1439 : olson 1.20 print $fh "update_thru\t$update_thru\n";
1440 : olson 1.1 print $fh "cur_update\t$now\n";
1441 :     print $fh "target_release\t$my_release\n";
1442 :     print $fh "compatible\t$compatible\n";
1443 :     print $fh "num_pegs\t$num_pegs\n";
1444 :     print $fh "num_genomes\t$num_genomes\n";
1445 :     print $fh "num_annos\t$num_annos\n";
1446 : olson 1.15 print $fh "num_assignments\t$num_assignments\n";
1447 : olson 1.1 close($fh);
1448 :    
1449 : olson 1.15 #
1450 :     # Construct list of users, and pdate local user database.
1451 :     #
1452 :    
1453 :     my @users = keys(%users);
1454 : olson 1.17 # $fig->ensure_users(\@users);
1455 : olson 1.15
1456 :     return [$session_id, $my_release, $num_assignments, $num_annos, $num_pegs, $num_genomes,
1457 : olson 1.16 $now, $compatible, \@users];
1458 : olson 1.1 }
1459 :    
1460 :    
1461 :     sub get_pegs
1462 :     {
1463 :     my($self, $session_id, $start, $len) = @_;
1464 :     my(%session_info);
1465 :    
1466 :     my $spool_dir = "$FIG_Config::temp/p2p_spool/$session_id";
1467 :    
1468 :     -d $spool_dir or die "Invalid session id $session_id";
1469 :    
1470 :     #
1471 :     # Read in the cached information for this session.
1472 :     #
1473 :    
1474 :     open(my $info_fh, "<$spool_dir/INFO") or die "Cannot open INFO file: $!";
1475 :     while (<$info_fh>)
1476 :     {
1477 :     chomp;
1478 :     my($var, $val) = split(/\t/, $_, 2);
1479 :     $session_info{$var} = $val;
1480 :     }
1481 :     close($info_fh);
1482 :    
1483 :     #
1484 :     # Sanity check start and length.
1485 :     #
1486 :    
1487 :     if ($start < 0 or $start >= $session_info{num_pegs})
1488 :     {
1489 :     die "Invalid start position $start";
1490 :     }
1491 :    
1492 :     if ($len < 0 or ($start + $len - 1) >= $session_info{num_pegs})
1493 :     {
1494 :     die "Invalid length $len";
1495 :     }
1496 :    
1497 :     #
1498 :     # Open file, spin to the starting line, then start reading.
1499 :     #
1500 :    
1501 :     open(my $peg_fh, "<$spool_dir/pegs") or die "Cannot open pegs file: $!";
1502 :    
1503 :     my $peg_output = [];
1504 :     my $genome_output = [];
1505 :    
1506 :     my $peg_num = 0;
1507 :     my $genomes_to_show = [];
1508 :     my %genomes_to_show;
1509 :    
1510 :     my($fid, $genome, @aliases);
1511 :    
1512 :     while (<$peg_fh>)
1513 :     {
1514 :     next if ($peg_num < $start);
1515 :    
1516 :     last if ($peg_num > ($start + $len));
1517 :    
1518 :     chomp;
1519 :    
1520 :     #
1521 :     # OK, this is a peg to process.
1522 :     # It's easy if we're compatible.
1523 :     #
1524 :    
1525 :     ($fid, $genome, @aliases) = split(/\t/, $_);
1526 :    
1527 :     if ($session_info{compatible})
1528 :     {
1529 :     push(@$peg_output, ['peg', $fid]);
1530 :     }
1531 :     else
1532 :     {
1533 :     if (!$genomes_to_show{$genome})
1534 :     {
1535 :     push(@$genomes_to_show, $genome);
1536 :     $genomes_to_show{$genome}++;
1537 :     }
1538 :     push(@$peg_output, ['peg_info', $fid, [@aliases], $genome]);
1539 :     }
1540 :     }
1541 :     continue
1542 :     {
1543 :     $peg_num++;
1544 :     }
1545 :    
1546 :     #
1547 :     # Read the genomes file, returning information about genomes referenced
1548 :     # in the pegs returned.
1549 :     #
1550 :    
1551 :     my $n_left = @$genomes_to_show;
1552 :    
1553 :     open(my $gfh, "<$spool_dir/genomes") or die "Cannot open genomes file: $!";
1554 :     while ($n_left > 0 and $_ = <$gfh>)
1555 :     {
1556 :     chomp;
1557 :    
1558 :     my($genome, $n_annos, $n_contigs, $n_nucs, $cksum) = split(/\t/);
1559 :    
1560 :     if ($genomes_to_show{$genome})
1561 :     {
1562 :     push(@$genome_output, [$genome, $n_contigs, $n_nucs, $cksum]);
1563 :     $n_left--;
1564 :     }
1565 :     }
1566 :     close($gfh);
1567 :    
1568 :     return [$peg_output, $genome_output];
1569 :     }
1570 : olson 1.6
1571 :     sub finalize_pegs
1572 :     {
1573 :     my($self, $session, $request) = @_;
1574 :     my($out);
1575 :    
1576 :     my $fig = new FIG;
1577 :    
1578 :     #
1579 :     # Walk the request handling appropriately. This is fairly easy, as it
1580 :     # is just a matter of pulling either sequence or location/contig data.
1581 :     #
1582 :    
1583 :     for my $item (@$request)
1584 :     {
1585 :     my($what, $peg) = @$item;
1586 :    
1587 :     if ($what eq "peg_genome")
1588 :     {
1589 :     #
1590 :     # Return the location and contig checksum for this peg.
1591 :     #
1592 : olson 1.13 # We also include the sequence in case the contig mapping doesn't work.
1593 :     #
1594 : olson 1.6
1595 :     my $loc = $fig->feature_location($peg);
1596 :     my $contig = $fig->contig_of($loc);
1597 : olson 1.7 my $cksum = $fig->contig_checksum($fig->genome_of($peg), $contig);
1598 : olson 1.13 my $seq = $fig->get_translation($peg);
1599 : olson 1.6
1600 :     push(@$out, ['peg_loc', $peg,
1601 : olson 1.13 $fig->strand_of($peg),
1602 : olson 1.6 $fig->beg_of($loc), $fig->end_of($loc),
1603 : olson 1.13 $cksum, $seq]);
1604 : olson 1.6
1605 :     }
1606 : olson 1.7 elsif ($what eq "peg_unknown")
1607 : olson 1.6 {
1608 :     my $seq = $fig->get_translation($peg);
1609 :     push(@$out, ['peg_seq', $peg, $seq]);
1610 :     }
1611 :     }
1612 :     return $out;
1613 :     }
1614 :    
1615 : olson 1.15
1616 :     sub get_annotations
1617 :     {
1618 :     my($self, $session_id, $start, $len) = @_;
1619 :    
1620 :     #
1621 :     # This is now easy; just run thru the saved annotations and return.
1622 :     #
1623 :    
1624 :     my(%session_info);
1625 :    
1626 :     my $spool_dir = "$FIG_Config::temp/p2p_spool/$session_id";
1627 :    
1628 :     -d $spool_dir or die "Invalid session id $session_id";
1629 :    
1630 :     #
1631 :     # Read in the cached information for this session.
1632 :     #
1633 :    
1634 :     open(my $info_fh, "<$spool_dir/INFO") or die "Cannot open INFO file: $!";
1635 :     while (<$info_fh>)
1636 :     {
1637 :     chomp;
1638 :     my($var, $val) = split(/\t/, $_, 2);
1639 :     $session_info{$var} = $val;
1640 :     }
1641 :     close($info_fh);
1642 :    
1643 :     #
1644 :     # Sanity check start and length.
1645 :     #
1646 :    
1647 :     if ($start < 0 or $start >= $session_info{num_annos})
1648 :     {
1649 :     die "Invalid start position $start";
1650 :     }
1651 :    
1652 :     if ($len < 0 or ($start + $len - 1) >= $session_info{num_annos})
1653 :     {
1654 :     die "Invalid length $len";
1655 :     }
1656 :    
1657 :     #
1658 :     # Open file, spin to the starting line, then start reading.
1659 :     #
1660 :    
1661 :     open(my $anno_fh, "<$spool_dir/annos") or die "Cannot open annos file: $!";
1662 :    
1663 :     my $anno_output = [];
1664 :    
1665 :     my $anno_num = 0;
1666 :    
1667 :     local $/ = "//\n";
1668 :     while (<$anno_fh>)
1669 :     {
1670 :     next if ($anno_num < $start);
1671 :    
1672 :     last if ($anno_num > ($start + $len));
1673 :    
1674 :     chomp;
1675 :    
1676 :     my($id, $date, $author, $anno) = split(/\n/, $_, 4);
1677 :    
1678 :     push(@$anno_output, [$id, $date, $author, $anno]);
1679 :     }
1680 :     continue
1681 :     {
1682 :     $anno_num++;
1683 :     }
1684 :    
1685 :     return $anno_output;
1686 :     }
1687 : olson 1.19
1688 :     sub get_assignments
1689 :     {
1690 :     my($self, $session_id, $start, $len) = @_;
1691 :    
1692 :     #
1693 :     # This is now easy; just run thru the saved assignments and return.
1694 :     #
1695 :    
1696 :     my(%session_info);
1697 :    
1698 :     my $spool_dir = "$FIG_Config::temp/p2p_spool/$session_id";
1699 :    
1700 :     -d $spool_dir or die "Invalid session id $session_id";
1701 :    
1702 :     #
1703 :     # Read in the cached information for this session.
1704 :     #
1705 :    
1706 :     open(my $info_fh, "<$spool_dir/INFO") or die "Cannot open INFO file: $!";
1707 :     while (<$info_fh>)
1708 :     {
1709 :     chomp;
1710 :     my($var, $val) = split(/\t/, $_, 2);
1711 :     $session_info{$var} = $val;
1712 :     }
1713 :     close($info_fh);
1714 :    
1715 :     #
1716 :     # Sanity check start and length.
1717 :     #
1718 :    
1719 :     if ($start < 0 or $start >= $session_info{num_assignments})
1720 :     {
1721 :     die "Invalid start position $start";
1722 :     }
1723 :    
1724 :     if ($len < 0 or ($start + $len - 1) >= $session_info{num_assignments})
1725 :     {
1726 :     die "Invalid length $len";
1727 :     }
1728 :    
1729 :     #
1730 :     # Open file, spin to the starting line, then start reading.
1731 :     #
1732 :    
1733 :     open(my $assign_fh, "<$spool_dir/assignments") or die "Cannot open assignments file: $!";
1734 :    
1735 :     my $assign_output = [];
1736 :    
1737 :     my $assign_num = 0;
1738 :    
1739 :     while (<$assign_fh>)
1740 :     {
1741 :     next if ($assign_num < $start);
1742 :    
1743 :     last if ($assign_num > ($start + $len));
1744 :    
1745 :     chomp;
1746 :    
1747 :     my($id, $date, $author, $func) = split(/\t/, $_, 4);
1748 :    
1749 :     push(@$assign_output, [$id, $date, $author, $func]);
1750 :     }
1751 :     continue
1752 :     {
1753 :     $assign_num++;
1754 :     }
1755 :    
1756 :     return $assign_output;
1757 :     }
1758 :    
1759 : olson 1.22 1;

MCS Webmaster
ViewVC Help
Powered by ViewVC 1.0.3