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

Annotation of /FigKernelPackages/P2P.pm

Parent Directory Parent Directory | Revision Log Revision Log


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

MCS Webmaster
ViewVC Help
Powered by ViewVC 1.0.3