3 from BeautifulSoup import BeautifulSoup
4 from datetime import datetime
6 from operator import attrgetter
7 from optparse import OptionParser, OptionGroup
19 from htsworkflow.util import api
20 from htsworkflow.util.rdfhelp import \
32 libraryNS = RDF.NS("http://jumpgate.caltech.edu/library/")
35 from htsworkflow.submission.ucsc import submission_view_url, UCSCEncodePipeline
36 download_ddf = urlparse.urljoin(UCSCEncodePipeline, "download_ddf#", allow_fragments=True)
37 ddfNS = RDF.NS(download_ddf)
39 DBDIR = os.path.expanduser("~diane/proj/submission")
41 logger = logging.getLogger("encode_find")
43 LOGIN_URL = 'http://encodesubmit.ucsc.edu/account/login'
44 USER_URL = 'http://encodesubmit.ucsc.edu/pipeline/show_user'
49 def main(cmdline=None):
50 parser = make_parser()
51 opts, args = parser.parse_args(cmdline)
54 logging.basicConfig(level=logging.INFO)
56 htsw_authdata = api.make_auth_from_opts(opts, parser)
57 htswapi = api.HtswApi(opts.host, htsw_authdata)
60 model = get_model(opts.load_model, DBDIR)
62 if opts.load_rdf is not None:
63 ns_uri = submitOntology[''].uri
64 load_into_model(model, opts.rdf_parser_name, opts.load_rdf, ns_uri)
67 cookie = login(cookie=cookie)
68 load_my_submissions(model, cookie=cookie)
69 load_encode_libraries(model, htswapi)
71 if opts.sparql is not None:
72 sparql_query(model, opts.sparql)
74 if opts.find_submission_with_no_library:
75 missing = find_submissions_with_no_library(model)
78 serializer = get_serializer(name=opts.rdf_parser_name)
79 print serializer.serialize_model_to_string(model)
83 parser = OptionParser()
84 commands = OptionGroup(parser, "Commands")
85 commands.add_option('--load-model', default=None,
86 help="Load model database")
87 commands.add_option('--load-rdf', default=None,
88 help="load rdf statements into model")
89 commands.add_option('--print-rdf', action="store_true", default=False,
90 help="print ending model state")
91 commands.add_option('--update', action="store_true", default=False,
92 help="Query remote data sources and update our database")
93 #commands.add_option('--update-ucsc-status', default=None,
94 # help="download status from ucsc, requires filename for extra rules")
95 #commands.add_option('--update-ddfs', action="store_true", default=False,
96 # help="download ddf information for known submission")
97 #commands.add_option('--update-library', default=None,
98 # help="download library info from htsw, requires filename for extra rules")
99 parser.add_option_group(commands)
101 queries = OptionGroup(parser, "Queries")
102 queries.add_option('--sparql', default=None,
103 help="execute arbitrary sparql query")
104 queries.add_option('--find-submission-with-no-library', default=False,
106 help="find submissions with no library ID")
107 parser.add_option_group(queries)
109 options = OptionGroup(parser, "Options")
110 options.add_option("--rdf-parser-name", default="turtle",
111 help="set rdf file parser type")
112 options.add_option("-v", "--verbose", action="store_true", default=False)
113 parser.add_option_group(options)
115 api.add_auth_options(parser)
119 def load_my_submissions(model, cookie=None):
123 soup = get_url_as_soup(USER_URL, 'GET', cookie)
124 p = soup.find('table', attrs={'id':'projects'})
125 tr = p.findNext('tr')
126 # first record is header
128 TypeN = rdfsNS['type']
129 NameN = submitOntology['name']
130 SpeciesN = submitOntology['species']
131 LibraryURN = submitOntology['library_urn']
133 while tr is not None:
134 td = tr.findAll('td')
135 if td is not None and len(td) > 1:
136 subUrnText = td[0].contents[0].contents[0].encode(CHARSET)
137 subUrn = RDF.Uri(submission_view_url(subUrnText))
139 add_stmt(model, subUrn, TypeN, submitOntology['Submission'])
141 name = get_contents(td[4])
142 add_stmt(model, subUrn, NameN, name)
144 species = get_contents(td[2])
145 if species is not None:
146 add_stmt(model, subUrn, SpeciesN, species)
148 library_id = get_library_id(name)
149 if library_id is not None:
150 add_submission_to_library_urn(model,
155 add_submission_creation_date(model, subUrn, cookie)
157 # grab changing atttributes
158 status = get_contents(td[6]).strip()
159 last_mod_datetime = get_date_contents(td[8])
160 last_mod = last_mod_datetime.isoformat()
162 update_submission_detail(model, subUrn, status, last_mod, cookie=cookie)
164 logging.info("Processed {0}".format( subUrn))
166 tr = tr.findNext('tr')
169 def add_submission_to_library_urn(model, submissionUrn, predicate, library_id):
170 """Add a link from a UCSC submission to woldlab library if needed
172 libraryUrn = libraryNS[library_id]
173 query = RDF.Statement(submissionUrn, predicate, libraryUrn)
174 if not model.contains_statement(query):
175 link = RDF.Statement(submissionUrn, predicate, libraryNS[library_id])
176 logger.info("Adding Sub -> Lib link: {0}".format(link))
177 model.add_statement(link)
179 logger.debug("Found: {0}".format(str(query)))
182 def find_submissions_with_no_library(model):
183 missing_lib_query = RDF.SPARQLQuery("""
184 PREFIX submissionOntology:<{submissionOntology}>
189 ?subid submissionOntology:name ?name
190 OPTIONAL {{ ?subid submissionOntology:library_urn ?libid }}
191 FILTER (!bound(?libid))
192 }}""".format(submissionOntology=submitOntology[''].uri)
195 results = missing_lib_query.execute(model)
199 print "# {0}".format(name)
200 print "<{0}>".format(subid.uri)
201 print " encodeSubmit:library_urn <http://jumpgate.caltech.edu/library/> ."
205 def add_submission_creation_date(model, subUrn, cookie):
206 # in theory the submission page might have more information on it.
207 creationDateN = libraryOntology['date']
208 dateTimeType = xsdNS['dateTime']
209 query = RDF.Statement(subUrn, creationDateN, None)
210 creation_dates = list(model.find_statements(query))
211 if len(creation_dates) == 0:
212 logger.info("Getting creation date for: {0}".format(str(subUrn)))
213 soup = get_url_as_soup(str(subUrn.uri), 'GET', cookie)
214 created_label = soup.find(text="Created: ")
216 created_date = get_date_contents(created_label.next)
217 created_date_node = RDF.Node(literal=created_date.isoformat(),
218 datatype=dateTimeType.uri)
219 add_stmt(model, subUrn, creationDateN, created_date_node)
221 logger.debug("Found creation date for: {0}".format(str(subUrn)))
223 def update_submission_detail(model, subUrn, status, recent_update, cookie):
224 HasStatusN = submitOntology['has_status']
225 StatusN = submitOntology['status']
226 LastModifyN = submitOntology['last_modify_date']
228 status_nodes_query = RDF.Statement(subUrn, HasStatusN, None)
229 status_nodes = list(model.find_statements(status_nodes_query))
231 if len(status_nodes) == 0:
232 # has no status node, add one
233 logging.info("Adding status node to {0}".format(subUrn))
234 status_blank = RDF.Node()
235 add_stmt(model, subUrn, HasStatusN, status_blank)
236 add_stmt(model, status_blank, rdfsNS['type'], StatusN)
237 add_stmt(model, status_blank, StatusN, status)
238 add_stmt(model, status_blank, LastModifyN, recent_update)
239 update_ddf(model, subUrn, status_blank, cookie=cookie)
241 logging.info("Found {0} status blanks".format(len(status_nodes)))
242 for status_statement in status_nodes:
243 status_blank = status_statement.object
244 last_modified_query = RDF.Statement(status_blank, LastModifyN, None)
245 last_mod_nodes = model.find_statements(last_modified_query)
246 for last_mod_statement in last_mod_nodes:
247 last_mod_date = str(last_mod_statement.object)
248 if recent_update == str(last_mod_date):
249 update_ddf(model, subUrn, status_blank, cookie=cookie)
254 def update_ddf(model, subUrn, statusNode, cookie):
255 TypeN = rdfsNS['type']
257 download_ddf_url = str(subUrn).replace('show', 'download_ddf')
258 ddfUrn = RDF.Uri(download_ddf_url)
260 status_is_ddf = RDF.Statement(statusNode, TypeN, ddfNS['ddf'])
261 if not model.contains_statement(status_is_ddf):
262 logging.info('Adding ddf to {0}, {1}'.format(subUrn, statusNode))
263 ddf_text = get_url_as_text(download_ddf_url, 'GET', cookie)
264 add_ddf_statements(model, statusNode, ddf_text)
265 model.add_statement(status_is_ddf)
268 def add_ddf_statements(model, statusNode, ddf_string):
269 """Convert a ddf text file into RDF Statements
271 ddf_lines = ddf_string.split('\n')
272 # first line is header
273 header = ddf_lines[0].split()
274 attributes = [ ddfNS[x] for x in header ]
277 for ddf_line in ddf_lines[1:]:
278 ddf_line = ddf_line.strip()
279 if len(ddf_line) == 0:
281 if ddf_line.startswith("#"):
284 ddf_record = ddf_line.split('\t')
285 files = ddf_record[0].split(',')
286 file_attributes = ddf_record[1:]
289 fileNode = RDF.Node()
290 add_stmt(model, statusNode, submitOntology['has_file'], fileNode)
291 add_stmt(model, fileNode, rdfsNS['type'], ddfNS['file'])
292 add_stmt(model, fileNode, ddfNS['filename'], f)
294 for predicate, object in zip( attributes[1:], file_attributes):
295 add_stmt(model, fileNode, predicate, object)
298 def load_encode_libraries(model, htswapi):
299 """Get libraries associated with encode.
301 encodeUrl = os.path.join(htswapi.root_url + "/library/?affiliations__id__exact=44")
302 rdfaParser = RDF.Parser(name='rdfa')
304 rdfaParser.parse_into_model(model, encodeUrl)
305 query = RDF.Statement(None, libraryOntology['library_id'], None)
306 libraries = model.find_statements(query)
307 for statement in libraries:
308 libraryUrn = statement.subject
309 load_library_detail(model, libraryUrn)
312 def load_library_detail(model, libraryUrn):
313 """Grab detail information from library page
315 rdfaParser = RDF.Parser(name='rdfa')
316 query = RDF.Statement(libraryUrn, libraryOntology['date'], None)
317 results = list(model.find_statements(query))
318 if len(results) == 0:
319 logger.info("Loading {0}".format(str(libraryUrn)))
320 rdfaParser.parse_into_model(model, libraryUrn.uri)
321 elif len(results) == 1:
322 pass # Assuming that a loaded dataset has one record
324 logging.warning("Many dates for {0}".format(libraryUrn))
326 def get_library_id(name):
327 """Guess library ID from library name
329 match = re.search(r"[ -](?P<id>([\d]{5})|(SL[\d]{4}))", name)
331 if match is not None:
332 library_id = match.group('id')
336 def get_contents(element):
337 """Return contents or none.
339 if len(element.contents) == 0:
342 a = element.find('a')
344 return a.contents[0].encode(CHARSET)
346 return element.contents[0].encode(CHARSET)
349 def get_date_contents(element):
350 data = get_contents(element)
352 return datetime.strptime(data, "%Y-%m-%d %H:%M")
356 def sparql_query(model, query_filename):
357 """Execute sparql query from file
359 query_body = open(query_filename,'r').read()
360 query = RDF.SPARQLQuery(query_body)
361 results = query.execute(model)
364 for k,v in row.items()[::-1]:
365 print "{0}: {1}".format(k,v)
369 def load_into_model(model, parser_name, filename):
370 if not os.path.exists(filename):
371 raise IOError("Can't find {0}".format(filename))
373 data = open(filename, 'r').read()
374 rdf_parser = RDF.Parser(name=parser_name)
375 rdf_parser.parse_string_into_model(model, data, ns_uri)
377 def add_stmt(model, subject, predicate, object):
378 """Convienence create RDF Statement and add to a model
380 return model.add_statement(
381 RDF.Statement(subject, predicate, object)
384 def login(cookie=None):
385 """Login if we don't have a cookie
387 if cookie is not None:
390 keys = keyring.get_keyring()
391 password = keys.get_password(LOGIN_URL, USERNAME)
392 credentials = {'login': USERNAME,
393 'password': password}
394 headers = {'Content-type': 'application/x-www-form-urlencoded'}
395 http = httplib2.Http()
396 response, content = http.request(LOGIN_URL,
399 body=urllib.urlencode(credentials))
400 logging.debug("Login to {0}, status {1}".format(LOGIN_URL,
403 cookie = response.get('set-cookie', None)
405 raise RuntimeError("Wasn't able to log into: {0}".format(LOGIN_URL))
409 def get_url_as_soup(url, method, cookie=None):
410 http = httplib2.Http()
412 if cookie is not None:
413 headers['Cookie'] = cookie
414 response, content = http.request(url, method, headers=headers)
415 if response['status'] == '200':
416 soup = BeautifulSoup(content,
417 fromEncoding="utf-8", # should read from header
418 convertEntities=BeautifulSoup.HTML_ENTITIES
422 msg = "error accessing {0}, status {1}"
423 msg = msg.format(url, response['status'])
424 e = httplib2.HttpLib2ErrorWithResponse(msg, response, content)
426 def get_url_as_text(url, method, cookie=None):
427 http = httplib2.Http()
429 if cookie is not None:
430 headers['Cookie'] = cookie
431 response, content = http.request(url, method, headers=headers)
432 if response['status'] == '200':
435 msg = "error accessing {0}, status {1}"
436 msg = msg.format(url, response['status'])
437 e = httplib2.HttpLib2ErrorWithResponse(msg, response, content)
441 SUBMISSIONS_LACKING_LIBID = [
442 ('1x75-Directional-HeLa-Rep1', '11208'),
443 ('1x75-Directional-HeLa-Rep2', '11207'),
444 ('1x75-Directional-HepG2-Rep1', '11210'),
445 ('1x75-Directional-HepG2-Rep2', '11209'),
446 ('1x75-Directional-H1-hESC-Rep1', '10947'),
447 ('1x75-Directional-H1-hESC-Rep2', '11009'),
448 ('1x75-Directional-HUVEC-Rep1', '11206'),
449 ('1x75-Directional-HUVEC-Rep2', '11205'),
450 ('1x75-Directional-K562-Rep1', '11008'),
451 ('1x75-Directional-K562-Rep2', '11007'),
452 ('1x75-Directional-NHEK-Rep1', '11204'),
453 ('1x75-Directional-GM12878-Rep1', '11011'),
454 ('1x75-Directional-GM12878-Rep2', '11010'),
459 def select_by_library_id(submission_list):
460 subl = [ (x.library_id, x) for x in submission_list if x.library_id ]
462 for lib_id, subobj in subl:
463 libraries.setdefault(lib_id, []).append(subobj)
465 for submission in libraries.values():
466 submission.sort(key=attrgetter('date'), reverse=True)
470 def library_to_freeze(selected_libraries):
471 freezes = ['2010-Jan', '2010-Jul', '2011-Jan']
472 lib_ids = sorted(selected_libraries.keys())
473 report = ['<html><table border="1">']
476 <style type="text/css">
477 td {border-width:0 0 1px 1px; border-style:solid;}
483 report.append('<thead>')
484 report.append('<tr><td>Library ID</td><td>Name</td>')
486 report.append('<td>{0}</td>'.format(f))
487 report.append('</tr>')
488 report.append('</thead>')
489 report.append('<tbody>')
490 for lib_id in lib_ids:
491 report.append('<tr>')
492 lib_url = libraryNS[lib_id].uri
493 report.append('<td><a href="{0}">{1}</a></td>'.format(lib_url, lib_id))
494 submissions = selected_libraries[lib_id]
495 report.append('<td>{0}</td>'.format(submissions[0].name))
497 for sub in submissions:
498 date = date_to_freeze(sub.date)
499 batched.setdefault(date, []).append(sub)
500 print lib_id, batched
502 report.append('<td>')
503 for s in batched.get(d, []):
504 show_url = submission_view_url(s.subid)
505 subid = '<a href="{0}">{1}</a>'.format(show_url, s.subid)
506 report.append("{0}:{1}".format(subid, s.status))
507 report.append('</td>')
509 report.append('<td></td>')
510 report.append("</tr>")
511 report.append('</tbody>')
512 report.append("</table></html>")
513 return "\n".join(report)
516 def date_to_freeze(d):
517 freezes = [ (datetime(2010, 1, 30), '2010-Jan'),
518 (datetime(2010, 7, 30), '2010-Jul'),
519 (datetime(2011, 1, 30), '2011-Jan'),
521 for end, name in freezes:
527 if __name__ == "__main__":