-
Notifications
You must be signed in to change notification settings - Fork 1
/
ccex_mp.py
343 lines (299 loc) · 14 KB
/
ccex_mp.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
import os
import sys
import time
import multiprocessing
from rdflib import URIRef, Literal
from rdflib.namespace import RDF
from settings import *
# functions
def check_permission(input_dir, output_dir):
"""
checks if input is readable and output is writeable
"""
if not(os.access(output_dir, os.W_OK)):
os.chmod(output_dir, int(0777))
if not(os.access(input_dir, os.R_OK)):
os.chmod(input_dir, int(0744))
print ("Input: %s\t is OK\nOutput: %s\tis OK" %(input_dir, output_dir))
def explode(c_vals, ref_ids, c, c_parent):
i = 0
for r in ref_ids:
exploded_cross_refs = et.Element(cross_ref_tag_name, refid=r, nsmap=NS_MAP)
exploded_cross_refs.set("connected", str(i))
exploded_cross_refs.text = "[" + c_vals[i] + "]"
c.addprevious(exploded_cross_refs)
i += 1
remove_cross_refs(c, c_parent)
def remove_preserve_tail(element):
if element.tail:
prev = element.getprevious()
parent = element.getparent()
if prev is not None:
prev.tail = (prev.tail or '') + element.tail
else:
parent.text = (parent.text or '') + element.tail
parent.remove(element)
def remove_cross_refs(element, element_parent):
global count_remove_all, count_remove_preserve
count_remove_all += 1
try:
et.strip_elements(element_parent, cross_refs_tag_name, with_tail=False)
except TypeError:
count_remove_preserve += 1
remove_preserve_tail(element)
def build_textual_marker(p_number, ref_id):
# output : [xxxcitxxx[[_'6'_][_'1'_]]xxxcitxxx]
return "[xxxcitxxx[[_'" + str(p_number) + "'_][_'" + ref_id + "'_]]xxxcitxxx]"
def xml_to_rdf(file):
global total_time, number_of_papers, count_remove_preserve, \
count_remove_all, papers_with_block_detect_error, papers_with_no_crossrefs
f_name, f_extension = os.path.splitext(file)
if (f_name[-5:] == "-full"):
eid = f_name[0:-5]
else:
eid = f_name
file_path = os.path.join(input_dir, file)
print("Processing of %s started\n" % eid)
start_time_ind = time.time()
tree = et.parse(file_path)
"""
STEP 1
Expand cross-refs as multiple adjacent <cross-ref refid=''> elements
- finds 'bib1' in bibliography, sets its positionNumber'1' ->
finds all ce:cross-ref with refid = 'bib1' , set its positionNumberofBiblioref same as positionnumber'1'
"""
xpath = "//ce:cross-refs"
cross_refs = tree.xpath(xpath, namespaces=NS_MAP)
for c in cross_refs:
c_parent = c.getparent()
c_val = c.text.strip("[]")
if c_val:
c_vals = re.split(',|and', c_val)
ref_ids = c.attrib['refid'].strip().split()
if (len(c_vals) == len(ref_ids)):
explode(c_vals, ref_ids, c, c_parent)
else:
new_c_vals = []
for element in c_vals:
if element.isdigit():
new_c_vals.append(str(element))
else:
toexpand = NON_DECIMAL.sub('f', element)
toexpand = toexpand.split("f")
try:
for i in range(int(toexpand[0]), int(toexpand[1]) + 1, 1):
new_c_vals.append(str((i)))
except ValueError:
pass
# print "Problem!!"
if (len(new_c_vals) == len(ref_ids)):
explode(new_c_vals, ref_ids, c, c_parent)
else:
# pass in production, print for debugging
pass
# print "# refids NE new_c_vals.\n refids: %s \n c_vals: %s \n" %(ref_ids, new_c_vals)
"""
STEP 2
Count bib-reference(s)
Add an attribute to each cross-ref with the number of the bibliographic reference it denotes.
"""
xpath_bib_references = "//ce:bib-reference"
bib_refrences = tree.xpath(xpath_bib_references, namespaces={'ce': 'http://www.elsevier.com/xml/common/dtd'})
bib_ref_pos = 1
for b in bib_refrences:
b.set("positionNumber", str(bib_ref_pos))
b_ref_id = b.attrib['id'].split()[0]
xpath_cross_ref_bib_pointers = "//ce:cross-ref[@refid='{0}']".format(b_ref_id)
cross_ref_bib_pointers = tree.xpath(xpath_cross_ref_bib_pointers,
namespaces={'ce': 'http://www.elsevier.com/xml/common/dtd'})
for c in cross_ref_bib_pointers:
c.set("positionNumberOfBibliographicReference", str(bib_ref_pos))
bib_ref_pos += 1
"""
STEP 3
Identify InTextPointer (by checking @positionNumberOfBibliographicReference attribute)
- Add position attribute
- Normalize cross-ref content, by substituting content with a marker
- invokes buildTextualMarker()
"""
xpath = "//ce:cross-ref[@positionNumberOfBibliographicReference]"
cross_refs = tree.xpath(xpath, namespaces={'ce': 'http://www.elsevier.com/xml/common/dtd'})
if len(cross_refs) == 0:
no_cross_refs = os.path.join(output_dir, NO_CROSS_REFS_LIST)
with open(no_cross_refs, 'a') as ncf:
ncf.write(f+'\n')
ncf.close()
current_cross_ref_pos = 1
for c in cross_refs:
c.set("positionNumber", str(current_cross_ref_pos))
c_textual_marker = build_textual_marker(current_cross_ref_pos,
c.attrib['positionNumberOfBibliographicReference'])
try:
c.text = c.text + c_textual_marker
except TypeError:
c.text = "" + c_textual_marker
papers_with_block_detect_error.append(f)
tree.write("/Users/sheshkovsky/dev/env/ccex/erroro/OUTPUT_FILE.XML", pretty_print=True)
current_cross_ref_pos += 1
"""
STEP 4
Extract citation contexts and build info array
"""
c_ref_info = []
xpath = "//ce:cross-ref[@positionNumberOfBibliographicReference]"
cross_refs = tree.xpath(xpath, namespaces={'ce': 'http://www.elsevier.com/xml/common/dtd'})
count = 0
for c in cross_refs:
current_ref_id = c.attrib['positionNumberOfBibliographicReference']
c_ref_info_being_added = {}
c_ref_info_being_added['positionNumber'] = c.attrib['positionNumber']
c_ref_info_being_added['positionNumberOfBibliographicReference'] = current_ref_id
c_textual_marker_current = build_textual_marker(c_ref_info_being_added['positionNumber'], current_ref_id)
# print c_textual_marker_current
xpath_block = "//*[self::ce:para or self::ce:entry or self::ce:note-para or self::ce:simple-para or self::ce:textref or self::xocs:item-toc-section-title or self::entry or self::ce:source or self::ce:section-title][descendant::ce:cross-ref[@positionNumberOfBibliographicReference and @positionNumber='{0}']]".format(
c_ref_info_being_added['positionNumber'])
connected = c.get("connected")
block_containing_cross_ref = tree.xpath(xpath_block, namespaces=NMSPCS)
if not connected:
try:
block_content = et.tostring(block_containing_cross_ref[0], method="text", encoding="unicode")
except IndexError:
block_content = ""
else:
if connected == "0":
try:
block_content = et.tostring(block_containing_cross_ref[0], method="text", encoding="unicode")
except IndexError:
block_content = ""
candidate_sentences = sent_detector.tokenize(block_content.strip())
marker_regexp = "\[xxxcitxxx\[\[_'(?P<pos>.*?)'_\]\[_'.*?'_\]\]xxxcitxxx\]"
for i in range(len(candidate_sentences)):
if c_textual_marker_current in candidate_sentences[i]:
citation_context = candidate_sentences[i]
ref_pointers = re.findall("\[_'.*?'_\]", citation_context)
first_ref_pointer = ref_pointers[0].strip("[]_'")
c_ref_info_being_added['sentenceid'] = "sentence-with-in-text-reference-pointer-" + first_ref_pointer
citation_context = re.sub(marker_regexp, "", citation_context)
c_ref_info_being_added['citation_context'] = citation_context
c_ref_info_being_added['DEBUG-blockContent'] = block_content
c_ref_info.append(c_ref_info_being_added)
# print c_ref_info
# tree.write("/Users/sheshkovsky/dev/env/ccex/errori/1-s2.0-S1570826811000473-full_tree_ccxx.XML", pretty_print=True)
"""
STEP 5
Convert to RDF
"""
graph_of_citation_contexts = Graph()
graph_of_citation_contexts.namespace_manager = ns_mgr
work_uri = SEMLANCET_NS + eid # http://www.semanticlancet.eu/resource/1-s2.0-S157082680300009X
exp_uri = work_uri + "/version-of-record" # http://www.semanticlancet.eu/resource/1-s2.0-S157082680300009X/version-of-record
exp_resource = URIRef(exp_uri)
for c in c_ref_info:
# http://www.semanticlancet.eu/resource/1-s2.0-S157082680300009X/version-of-record/reference-list/NUMBER/reference
ref_uri = URIRef(exp_uri + "/reference-list/" + c['positionNumberOfBibliographicReference'] + "/reference")
# http://www.semanticlancet.eu/resource/1-s2.0-S157082680300009X/version-of-record/in-text-reference-pointer/positionNumber
in_text_pointer_uri = URIRef(exp_uri + "/in-text-reference-pointer-" + c['positionNumber'])
# http://www.semanticlancet.eu/resource/1-s2.0-S157082680300009X/version-of-record/sentenceid
citation_sentence_uri = URIRef(exp_uri + "/" + c['sentenceid'])
graph_of_citation_contexts.add((in_text_pointer_uri, RDF.type, c4o.InTextReferencePointer))
graph_of_citation_contexts.add(
(in_text_pointer_uri, c4o.hasContent, Literal("[" + c['positionNumberOfBibliographicReference'] + "]")))
graph_of_citation_contexts.add((in_text_pointer_uri, c4o.denotes, ref_uri))
graph_of_citation_contexts.add((citation_sentence_uri, RDF.type, doco.Sentence))
graph_of_citation_contexts.add((citation_sentence_uri, c4o.hasContent, Literal(c['citation_context'])))
graph_of_citation_contexts.add((citation_sentence_uri, frbr.partOf, exp_resource))
graph_of_citation_contexts.add((citation_sentence_uri, frbr.part, in_text_pointer_uri))
graph_of_citation_contexts.add((exp_resource, frbr.part, citation_sentence_uri))
graph_of_citation_contexts.add((in_text_pointer_uri, c4o.hasContext, citation_sentence_uri))
graph_of_citation_contexts.add((in_text_pointer_uri, frbr.partOf, citation_sentence_uri))
"""
STEP 6
Serialize in a file
"""
g_citation_filename = os.path.join(output_dir, eid + "." + RDF_EXTENSION)
graph_of_citation_contexts.serialize(destination=g_citation_filename, format='turtle')
number_of_papers += 1
exec_t = time.time() - start_time_ind
print("--- %s processed in %s seconds ---\n" %(eid, exec_t))
""""
Summary
TODO : add more information
"""
summary_file = os.path.join(output_dir, SUMMARY_FILENAME)
with open(summary_file, 'a') as sf:
sf.write("\n" + f + "\n")
for c in c_ref_info:
citation_contexts_summary = c['positionNumber'] + " | " + c[
'positionNumberOfBibliographicReference'] + " | " + c['sentenceid'] + " | " + c[
"citation_context"] + " | " + c['DEBUG-blockContent'] + "\n"
citation_contexts_summary = citation_contexts_summary.encode('ascii', 'ignore')
# print citation_contexts_summary
sf.write(citation_contexts_summary)
sf.close()
"""
In case we need to write the tree on a file
"""
# tree.write("OUTPUT_FILE.XML", pretty_print=True)
if __name__ == "__main__":
# set/check input & output directories
try:
arg1 = sys.argv[1]
arg2 = sys.argv[2]
except IndexError:
print "Usage: \tpython ccex.py <input_directory_name> <output_directory_name>"
sys.exit(1)
input_dir = os.path.join(os.path.dirname(os.path.abspath(__file__)), arg1)
output_dir = os.path.join(os.path.dirname(os.path.abspath(__file__)), arg2)
if not os.path.isdir(input_dir):
print(
"Invalid input directory: '%s' is not a valid direcotry, please insert a valid directory name" % input_dir)
sys.exit(1)
if not os.path.exists(output_dir):
print "Making directory %s as output" % output_dir
os.makedirs(output_dir)
check_permission(input_dir, output_dir)
all_files = os.listdir(input_dir)
files = []
# keep only XML files
for f in all_files:
f_name, f_extension = os.path.splitext(f)
if f_extension == '.xml':
files.append(f)
# counters
total_time = 0
number_of_papers = 0
count_remove_preserve = 0
count_remove_all = 0
papers_with_block_detect_error = []
papers_with_no_crossrefs = []
try:
if sys.argv[3] == '-mp':
multiprocess = True
try:
mp = int(sys.argv[4])
except:
mp = 2
except:
multiprocess = False
start_time = time.time()
if not multiprocess:
print "PROCESS STARTED IN ~~~SINGLE PROCESS~~~ MODE\n\n"
for f in files:
xml_to_rdf(f)
total_time = time.time() - start_time
else:
print "PROCESS STARTED IN ~~~MULTIPROCESSING :mp = %d~~~ MODE\n\n" %mp
pool = multiprocessing.Pool(mp)
pool.map(xml_to_rdf, files)
pool.close()
total_time = time.time() - start_time
print "Total execution time: %s seconds" %total_time
print "# Papers have been processed: %d" %number_of_papers
print "# Papers with no cross-ref: %d\n" %len(papers_with_no_crossrefs)
for p in papers_with_no_crossrefs:
print "\t", p
if papers_with_block_detect_error:
print "Block detection erros happened %d times" %len(papers_with_block_detect_error)
for p in papers_with_block_detect_error:
print "\t", p
print "# Cross-refs Rremoved: %d \n# Preserve function used: %d times" %(count_remove_all, count_remove_preserve)