-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathCHANGES.txt
94 lines (67 loc) · 2.31 KB
/
CHANGES.txt
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
1.3 (2022-02-14)
------------------
- remove normalization; it does not belong here
- switch to ISO release date format
1.2.4 (11.2. 2022)
------------------
- small re fix
- fix README
1.2.3 (10.2. 2022)
------------------
- have bestmatch return match quality also
1.2.1 (6.2. 2017)
-----------------
- constrain acceptable multiple match reduction by alternative
length (difference from term length)
1.2 (2.2. 2017)
---------------
- add special handling for cases where there's a continuous-
substring match but it's not enough to produce a sufficiently
high-quality match (see code for specific algorithm)
- now has 99,7% correct match rate with real-life
test data consisting of about 1000 unique samples
1.1 (1.2. 2017)
---------------
- switch to assuming that data passed in is already normalized
to requirements, whatever they are (keep turning everything
into lowercase though); change test data and test data
generation accordingly (things run a bit faster now)
- improved detection; common substring matches are now
distinguished by length, and the normalized match whose
length is closest to term in character length, is chosen as
the match (closes #4)
- cleanup and refactoring (new util module)
- improved tests
1.0 (31.3. 2016)
----------------
- assert variations passed to `closestvariation` is nonempty
- add support for handling multiple matches, with tests
0.9 (14.9. 2015)
----------------
- add support for Damerau-Levenshtein algorithm
- add support for detecting multiple equally good matches
- add support for optionally disqualifying terms from being
used as best match results
- add support for generating additional variations from dashed
and dotted names to improve matching
- make sure using unicode throughout works
- return more info in matching result
- add draft Sift4 matching algorithm support
- tests improvements & refactoring
- add test data for future testing of ambiguous matches
0.8 (9.9. 2015)
---------------
- use cleanco 1.3 unicode mode
0.7 (2.9. 2015)
---------------
- support Jaccard, Levensthein & Sorensen as
matching algorithms
0.6 (2.9. 2015)
---------------
- normalize also the name, not just variations
0.5 (1.9. 2015)
---------------
- store match info in NoMatchError
0.4 (27.8. 2015)
-----------------
- raise a custom NoMatchError upon failure