# -*- coding: utf-8 -*-
#
# This file is part of INSPIRE.
# Copyright (C) 2014-2017 CERN.
#
# INSPIRE is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# INSPIRE is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with INSPIRE. If not, see <http://www.gnu.org/licenses/>.
#
# In applying this license, CERN does not waive the privileges and immunities
# granted to it by virtue of its status as an Intergovernmental Organization
# or submit itself to any jurisdiction.
"""Functions to parse an authorlist."""
from __future__ import absolute_import, division, print_function
import re
import six
re_emptyline = re.compile(r'\n\s*\n', re.UNICODE)
re_hyphens = re.compile(
r'(\\255|\u02D7|\u0335|\u0336|\u2212|\u2013|\u002D|\uFE63|\uFF0D)', re.UNICODE)
re_multiple_space = re.compile(r'\s{2,}', re.UNICODE)
re_potential_key = re.compile(r"^(?:\d|[^\w.'-])+$", re.UNICODE)
re_trailing_nonword = re.compile(r"((?:\d|[^\w,.'-])+ )", re.UNICODE)
re_symbols = re.compile(r'[^\w ]', re.UNICODE)
[docs]def split_id(word):
"""
Separate potential aff-ids .
E.g.: '*12%$' -> ['*', '12' '%', '$']
"""
aff_ids = []
symbols = re_symbols.findall(word)
if symbols:
aff_ids += symbols
for rest in re_symbols.split(word):
if rest:
aff_ids.append(rest)
else:
aff_ids.append(word)
return aff_ids
[docs]def parse_authors(text, affiliations):
"""
Parse author names and convert to Lastname, Firstnames.
Can be separated by ',', newline or affiliation tag.
Returns:
List of tuples: (author_fullname, [author_affiliations])
List of strings: warnings
"""
import copy
authors = []
warnings = []
text = text.replace(',', ' , ')
text = text.replace('\n', ' , ')
text = text.replace(' and ', ' , ') + ' '
text = re_trailing_nonword.sub(r' \1', text)
text = re_multiple_space.sub(' ', text)
aff_keys = affiliations.keys()
unused_aff_keys = copy.deepcopy(aff_keys)
key_type = ''
if aff_keys:
if aff_keys[0].isalpha():
key_type = 'alpha'
elif aff_keys[0].isdigit():
key_type = 'digit'
else:
key_type = 'symbol'
warnings.append('CAUTION! Using symbols (# and stuff) as aff-IDs.')
else:
warnings.append('Found no affiliations (empty line needed)')
author_names = []
author_affs = []
fullname = ''
list_of_words = text.split(' ')
for nw, word in enumerate(list_of_words):
if word in aff_keys or word == ',' or re_potential_key.search(word):
# author name stops here, add affiliations
if author_names:
fullname = ' '.join(author_names)
if len(author_names) == 1:
warnings.append('Author without firstname: %s' % fullname)
author_names = []
if word in aff_keys:
author_affs.append(affiliations[word])
if word in unused_aff_keys:
unused_aff_keys.remove(word)
elif not word == ',':
# something left over or not separated
for aff_key in split_id(word):
if aff_key in aff_keys:
author_affs.append(affiliations[aff_key])
if aff_key in unused_aff_keys:
unused_aff_keys.remove(aff_key)
else:
warnings.append(
'Unresolved aff-ID or stray footnote symbol. '
'Problematic author and aff-id: %s %s' %
(fullname, aff_key)
)
else:
# (part of) (next) author name, process previous author
if key_type == 'alpha' and word.islower() and word.isalpha():
three_words = \
list_of_words[max(nw - 2, 0):min(nw + 3, len(list_of_words)) - 1]
warnings.append(
'Is this part of a name or missing aff-id? "%s" in %s' %
(word, ' '.join(three_words))
)
if fullname:
if affiliations and not author_affs:
# there should be affiliations
warnings.append(
'Author without affiliation-id. '
'Problematic author: %s' % fullname
)
authors.append((fullname, author_affs))
author_affs = []
fullname = ''
if word:
author_names.append(word)
if author_names:
fullname = ' '.join(author_names)
if len(author_names) == 1:
warnings.append('Author without firstname: %s' % fullname)
author_affs = []
if fullname:
authors.append((fullname, author_affs))
if unused_aff_keys:
warnings.append('Unused affiliation-IDs: %s' % unused_aff_keys)
return authors, warnings
[docs]def determine_aff_type_character(char_list):
"""
Guess whether affiliation are by number, letter or symbols (e.g. dagger).
Numbers and letters should not be mixed.
"""
aff_type = None
for char in char_list:
if aff_type:
if aff_type == 'alpha':
if not char.isalpha():
return None
elif aff_type == 'digit':
if not char.isdigit():
return None
else:
if char.isalpha():
aff_type = 'alpha'
elif char.isdigit():
aff_type = 'digit'
else:
aff_type = 'symbol'
break
return aff_type
[docs]def determine_aff_type(text):
"""
Guess format for affiliations.
Return corresponding search pattern.
"""
line_pattern_single = {'alpha': re.compile(r'^([a-z]+)\.*$', re.UNICODE),
'digit': re.compile(r'^(\d+)\.*$', re.UNICODE),
'symbol': re.compile(r'^(.)\.*$', re.UNICODE)}
line_pattern_line = {'alpha': re.compile(r'^([a-z]+)[ .]+(.*)', re.UNICODE),
'digit': re.compile(r'^(\d+)[ .]*(.*)', re.UNICODE),
'symbol': re.compile(r'^(.)[ .]+(.*)', re.UNICODE)}
single_char = []
first_char = []
for line in text.split('\n'):
line = line.strip(' .')
if len(line) == 1:
single_char.append(line)
elif line:
first_char.append(line[0])
if single_char:
aff_type = determine_aff_type_character(single_char)
if aff_type:
aff_pattern = line_pattern_single[aff_type]
else:
raise ValueError('Cannot identify type of affiliation, '
'found IDs: %s' % single_char)
else:
aff_type = determine_aff_type_character(first_char)
if aff_type:
aff_pattern = line_pattern_line[aff_type]
else:
raise ValueError('Cannot identify type of affiliations, '
'found IDs: %s' % first_char)
return aff_pattern
[docs]def parse_affiliations(text):
"""
Determine how affiliations are formatted.
Return hash of id:affiliation
Allowed formats:
don't mix letters and numbers, lower-case letters only
1
CERN, Switzerland
2
DESY,
Germany
1 CERN, Switzerland
2DESY, Germany
a CERN, Switzerland
bb DESY, Germany
*
CERN, Switzerland
#
DESY, Germany
"""
affiliations = {}
aff_pattern = determine_aff_type(text)
aff_id = None
this_aff = []
for line in text.split('\n'):
line = line.strip()
get_affiliation = aff_pattern.search(line)
if get_affiliation:
if len(get_affiliation.groups()) == 2:
affiliations[get_affiliation.group(1)] = \
get_affiliation.group(2).strip()
else:
if aff_id and this_aff:
affiliations[aff_id] = ' '.join(this_aff).strip()
aff_id = None
this_aff = []
aff_id = get_affiliation.group(1)
elif aff_id:
this_aff.append(line)
elif line:
raise ValueError('Something is wrong with the affiliation list')
if aff_id and this_aff:
affiliations[aff_id] = ' '.join(this_aff).strip()
return affiliations
[docs]def create_authors(text):
"""
Split text in (useful) blocks, sepatated by empty lines.
1 block: no affiliations
2 blocks: authors and affiliations
more blocks: authors grouped by affiliation (not implemented yet)
Returns:
dict: with two keys: ``authors`` of the form ``(author_fullname,
[author_affiliations])`` and ``warnings`` which is a list of strings.
"""
if not text:
return {}
if not isinstance(text, six.text_type):
text = text.decode('utf-8')
text = text.replace('\r', '') # Input from the form contains unwanted \r's
text = re_hyphens.sub('-', text)
empty_blocks = []
text_blocks = re_emptyline.split(text)
for num, block in enumerate(text_blocks):
if not re.search(r'\w', block):
empty_blocks.append(num)
empty_blocks.reverse()
for num in empty_blocks:
text_blocks.pop(num)
if len(text_blocks) == 0:
authors, warnings = [], []
elif len(text_blocks) == 1:
authors, warnings = parse_authors(text_blocks[0], {})
elif len(text_blocks) == 2:
affiliations = parse_affiliations(text_blocks[1])
authors, warnings = parse_authors(text_blocks[0], affiliations)
else:
# authors = parse_blocks(text_blocks)
raise ValueError('Authors grouped by affiliation? - Comming soon.'
'Or too many empty lines.')
if warnings:
return {'authors': authors, 'warnings': warnings}
else:
return {'authors': authors}
