|
1
|
|
|
"""Utility functions for Annif""" |
|
2
|
|
|
|
|
3
|
|
|
import glob |
|
4
|
|
|
import logging |
|
5
|
|
|
import os |
|
6
|
|
|
import os.path |
|
7
|
|
|
import tempfile |
|
8
|
|
|
|
|
9
|
|
|
import numpy as np |
|
10
|
|
|
|
|
11
|
|
|
from annif import logger |
|
12
|
|
|
from annif.suggestion import VectorSuggestionResult |
|
13
|
|
|
|
|
14
|
|
|
|
|
15
|
|
|
class DuplicateFilter(logging.Filter): |
|
16
|
|
|
"""Filter out log messages that have already been displayed.""" |
|
17
|
|
|
|
|
18
|
|
|
def __init__(self): |
|
19
|
|
|
self.logged = set() |
|
20
|
|
|
|
|
21
|
|
|
def filter(self, record): |
|
22
|
|
|
current_log = hash((record.module, record.levelno, record.msg, record.args)) |
|
23
|
|
|
if current_log not in self.logged: |
|
24
|
|
|
self.logged.add(current_log) |
|
25
|
|
|
return True |
|
26
|
|
|
return False |
|
27
|
|
|
|
|
28
|
|
|
|
|
29
|
|
|
def atomic_save(obj, dirname, filename, method=None): |
|
30
|
|
|
"""Save the given object (which must have a .save() method, unless the |
|
31
|
|
|
method parameter is given) into the given directory with the given |
|
32
|
|
|
filename, using a temporary file and renaming the temporary file to the |
|
33
|
|
|
final name.""" |
|
34
|
|
|
|
|
35
|
|
|
prefix, suffix = os.path.splitext(filename) |
|
36
|
|
|
tempfd, tempfilename = tempfile.mkstemp(prefix=prefix, suffix=suffix, dir=dirname) |
|
37
|
|
|
os.close(tempfd) |
|
38
|
|
|
logger.debug("saving %s to temporary file %s", str(obj)[:90], tempfilename) |
|
39
|
|
|
if method is not None: |
|
40
|
|
|
method(obj, tempfilename) |
|
41
|
|
|
else: |
|
42
|
|
|
obj.save(tempfilename) |
|
43
|
|
|
for fn in glob.glob(tempfilename + "*"): |
|
44
|
|
|
newname = fn.replace(tempfilename, os.path.join(dirname, filename)) |
|
45
|
|
|
logger.debug("renaming temporary file %s to %s", fn, newname) |
|
46
|
|
|
os.rename(fn, newname) |
|
47
|
|
|
|
|
48
|
|
|
|
|
49
|
|
|
def cleanup_uri(uri): |
|
50
|
|
|
"""remove angle brackets from a URI, if any""" |
|
51
|
|
|
if uri.startswith("<") and uri.endswith(">"): |
|
52
|
|
|
return uri[1:-1] |
|
53
|
|
|
return uri |
|
54
|
|
|
|
|
55
|
|
|
|
|
56
|
|
|
def merge_hits(weighted_hits, size): |
|
57
|
|
|
"""Merge hits from multiple sources. Input is a sequence of WeightedSuggestion |
|
58
|
|
|
objects. The size parameter determines the length of the subject vector. |
|
59
|
|
|
Returns an SuggestionResult object.""" |
|
60
|
|
|
|
|
61
|
|
|
weights = [whit.weight for whit in weighted_hits] |
|
62
|
|
|
scores = [whit.hits.as_vector(size) for whit in weighted_hits] |
|
63
|
|
|
result = np.average(scores, axis=0, weights=weights) |
|
64
|
|
|
return VectorSuggestionResult(result) |
|
65
|
|
|
|
|
66
|
|
|
|
|
67
|
|
|
def parse_sources(sourcedef): |
|
68
|
|
|
"""parse a source definition such as 'src1:1.0,src2' into a sequence of |
|
69
|
|
|
tuples (src_id, weight)""" |
|
70
|
|
|
|
|
71
|
|
|
sources = [] |
|
72
|
|
|
totalweight = 0.0 |
|
73
|
|
|
for srcdef in sourcedef.strip().split(","): |
|
74
|
|
|
srcval = srcdef.strip().split(":") |
|
75
|
|
|
src_id = srcval[0] |
|
76
|
|
|
if len(srcval) > 1: |
|
77
|
|
|
weight = float(srcval[1]) |
|
78
|
|
|
else: |
|
79
|
|
|
weight = 1.0 |
|
80
|
|
|
sources.append((src_id, weight)) |
|
81
|
|
|
totalweight += weight |
|
82
|
|
|
return [(srcid, weight / totalweight) for srcid, weight in sources] |
|
83
|
|
|
|
|
84
|
|
|
|
|
85
|
|
|
def parse_args(param_string): |
|
86
|
|
|
"""Parse a string of comma separated arguments such as '42,43,key=abc' into |
|
87
|
|
|
a list of positional args [42, 43] and a dict of keyword args {key: abc}""" |
|
88
|
|
|
|
|
89
|
|
|
if not param_string: |
|
90
|
|
|
return [], {} |
|
91
|
|
|
posargs = [] |
|
92
|
|
|
kwargs = {} |
|
93
|
|
|
param_strings = param_string.split(",") |
|
94
|
|
|
for p_string in param_strings: |
|
95
|
|
|
parts = p_string.split("=") |
|
96
|
|
|
if len(parts) == 1: |
|
97
|
|
|
posargs.append(p_string) |
|
98
|
|
|
elif len(parts) == 2: |
|
99
|
|
|
kwargs[parts[0]] = parts[1] |
|
100
|
|
|
return posargs, kwargs |
|
101
|
|
|
|
|
102
|
|
|
|
|
103
|
|
|
def boolean(val): |
|
104
|
|
|
"""Convert the given value to a boolean True/False value, if it isn't already. |
|
105
|
|
|
True values are '1', 'yes', 'true', and 'on' (case insensitive), everything |
|
106
|
|
|
else is False.""" |
|
107
|
|
|
|
|
108
|
|
|
return str(val).lower() in ("1", "yes", "true", "on") |
|
109
|
|
|
|
|
110
|
|
|
|
|
111
|
|
|
def identity(x): |
|
112
|
|
|
"""Identity function: return the given argument unchanged""" |
|
113
|
|
|
return x |
|
114
|
|
|
|
|
115
|
|
|
|
|
116
|
|
|
def metric_code(metric): |
|
117
|
|
|
"""Convert a human-readable metric name into an alphanumeric string""" |
|
118
|
|
|
return metric.translate(metric.maketrans(" ", "_", "()")) |
|
119
|
|
|
|
NatLibFi /
Annif
