|
1
|
|
|
"""Utility functions for Annif""" |
|
2
|
|
|
|
|
3
|
|
|
import glob |
|
4
|
|
|
import os |
|
5
|
|
|
import os.path |
|
6
|
|
|
import tempfile |
|
7
|
|
|
import numpy as np |
|
8
|
|
|
import cld3 |
|
9
|
|
|
from annif import logger |
|
10
|
|
|
from annif.suggestion import VectorSuggestionResult |
|
11
|
|
|
|
|
12
|
|
|
|
|
13
|
|
|
def atomic_save(obj, dirname, filename, method=None): |
|
14
|
|
|
"""Save the given object (which must have a .save() method, unless the |
|
15
|
|
|
method parameter is given) into the given directory with the given |
|
16
|
|
|
filename, using a temporary file and renaming the temporary file to the |
|
17
|
|
|
final name.""" |
|
18
|
|
|
|
|
19
|
|
|
prefix, suffix = os.path.splitext(filename) |
|
20
|
|
|
tempfd, tempfilename = tempfile.mkstemp( |
|
21
|
|
|
prefix=prefix, suffix=suffix, dir=dirname) |
|
22
|
|
|
os.close(tempfd) |
|
23
|
|
|
logger.debug('saving %s to temporary file %s', str(obj)[:90], tempfilename) |
|
24
|
|
|
if method is not None: |
|
25
|
|
|
method(obj, tempfilename) |
|
26
|
|
|
else: |
|
27
|
|
|
obj.save(tempfilename) |
|
28
|
|
|
for fn in glob.glob(tempfilename + '*'): |
|
29
|
|
|
newname = fn.replace(tempfilename, os.path.join(dirname, filename)) |
|
30
|
|
|
logger.debug('renaming temporary file %s to %s', fn, newname) |
|
31
|
|
|
os.rename(fn, newname) |
|
32
|
|
|
|
|
33
|
|
|
|
|
34
|
|
|
def cleanup_uri(uri): |
|
35
|
|
|
"""remove angle brackets from a URI, if any""" |
|
36
|
|
|
if uri.startswith('<') and uri.endswith('>'): |
|
37
|
|
|
return uri[1:-1] |
|
38
|
|
|
return uri |
|
39
|
|
|
|
|
40
|
|
|
|
|
41
|
|
|
def merge_hits(weighted_hits, subject_index): |
|
42
|
|
|
"""Merge hits from multiple sources. Input is a sequence of WeightedSuggestion |
|
43
|
|
|
objects. A SubjectIndex is needed to convert between subject IDs and URIs. |
|
44
|
|
|
Returns an SuggestionResult object.""" |
|
45
|
|
|
|
|
46
|
|
|
weights = [whit.weight for whit in weighted_hits] |
|
47
|
|
|
scores = [whit.hits.as_vector(subject_index) for whit in weighted_hits] |
|
48
|
|
|
result = np.average(scores, axis=0, weights=weights) |
|
49
|
|
|
return VectorSuggestionResult(result) |
|
50
|
|
|
|
|
51
|
|
|
|
|
52
|
|
|
def parse_sources(sourcedef): |
|
53
|
|
|
"""parse a source definition such as 'src1:1.0,src2' into a sequence of |
|
54
|
|
|
tuples (src_id, weight)""" |
|
55
|
|
|
|
|
56
|
|
|
sources = [] |
|
57
|
|
|
totalweight = 0.0 |
|
58
|
|
|
for srcdef in sourcedef.strip().split(','): |
|
59
|
|
|
srcval = srcdef.strip().split(':') |
|
60
|
|
|
src_id = srcval[0] |
|
61
|
|
|
if len(srcval) > 1: |
|
62
|
|
|
weight = float(srcval[1]) |
|
63
|
|
|
else: |
|
64
|
|
|
weight = 1.0 |
|
65
|
|
|
sources.append((src_id, weight)) |
|
66
|
|
|
totalweight += weight |
|
67
|
|
|
return [(srcid, weight / totalweight) for srcid, weight in sources] |
|
68
|
|
|
|
|
69
|
|
|
|
|
70
|
|
|
def parse_args(param_string): |
|
71
|
|
|
"""Parse a string of comma separated arguments such as '42,43,key=abc' into |
|
72
|
|
|
a list of positional args [42, 43] and a dict of keyword args {key: abc}""" |
|
73
|
|
|
|
|
74
|
|
|
if not param_string: |
|
75
|
|
|
return [], {} |
|
76
|
|
|
posargs = [] |
|
77
|
|
|
kwargs = {} |
|
78
|
|
|
param_strings = param_string.split(',') |
|
79
|
|
|
for p_string in param_strings: |
|
80
|
|
|
parts = p_string.split('=') |
|
81
|
|
|
if len(parts) == 1: |
|
82
|
|
|
posargs.append(p_string) |
|
83
|
|
|
elif len(parts) == 2: |
|
84
|
|
|
kwargs[parts[0]] = parts[1] |
|
85
|
|
|
return posargs, kwargs |
|
86
|
|
|
|
|
87
|
|
|
|
|
88
|
|
|
def detect_language(text): |
|
89
|
|
|
"""Tries to detect the language of a text input. Outputs a BCP-47-style |
|
90
|
|
|
language code (e.g. 'en') and a probability for the language in a tuple.""" |
|
91
|
|
|
|
|
92
|
|
|
lan_info = cld3.get_language(text) |
|
93
|
|
|
if lan_info is not None and lan_info.is_reliable: |
|
94
|
|
|
return (lan_info.language, lan_info.probability) |
|
95
|
|
|
else: |
|
96
|
|
|
return (None, None) |
|
97
|
|
|
|
|
98
|
|
|
|
|
99
|
|
|
def boolean(val): |
|
100
|
|
|
"""Convert the given value to a boolean True/False value, if it isn't already. |
|
101
|
|
|
True values are '1', 'yes', 'true', and 'on' (case insensitive), everything |
|
102
|
|
|
else is False.""" |
|
103
|
|
|
|
|
104
|
|
|
return str(val).lower() in ('1', 'yes', 'true', 'on') |
|
105
|
|
|
|
|
106
|
|
|
|
|
107
|
|
|
def identity(x): |
|
108
|
|
|
"""Identity function: return the given argument unchanged""" |
|
109
|
|
|
return x |
|
110
|
|
|
|
NatLibFi /
Annif
