-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathCHANGELOG
86 lines (75 loc) · 3.86 KB
/
CHANGELOG
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
v4.2.0
- Library:
- Refactoring: moved Serializations namespace to NW.Shared.Serialization.
v4.1.0
- Library:
- Refactoring: moved Files and Validation to NW.Shared.Files and NW.Shared.Validation.
v4.0.0 - BREAKING CHANGES
- Library:
- Feature: "Bag Pattern" adopted into the codebase (*Settings => SettingBag, *Components => ComponentBag, ...).
- Framework Freeze: migrated projects to .NET 8.0 and NUnit 4.x; updated packages accordingly, implemented new NUnit assertion model.
- Documentation:
- Feature: updated to v4.0.0.
v3.7.0
- Library:
- Performance: ClassifyMany() now performs 75% faster thanks to parallelism.
- Documentation:
- Feature: added Documentation-NW.NGramTextClassificationScripts.md.
- Scripts:
- Feature: added ngramtchelper.ps1.
v3.6.0
- Library:
- Feature: added support for the "CleanLabeledExamples" functionality.
- Feature: added TextSnippet property to TextClassifierResult.
- Feature: added support for the "disableindexserialization" flag for the SaveSession method.
- Performance: the tokenization process of the provided labeled examples now happens only once, despite of the number of text snippets.
- Client:
- Feature: added "--cleanlabeledexamples" option to the "classify" sub-command
- Feature: added "--disableindexserialization" option to the "classify" sub-command
- Documentation:
- Feature: updated to v3.6.0.
- File Formats:
- Feature: added TextSnippet field to each result object to Session Json.
- Feature: added a Session Json's variant with disabled index serialization.
v3.5.0 - BREAKING CHANGES
- Library:
- Feature: added support for the classification of a collection of snippets of texts.
- Feature: added support for Minimum Accuracy* fields in TextClassifierSettings.
- Feature: added support for "Files" namespace.
- Feature: added support for "Filenames" namespace.
- Feature: added the possibility to load and save labeled examples from/as JSON file.
- Feature: added the possibility to load and save text snippets from/as JSON file.
- Feature: added the possibility to load a NGramTokenizerRuleSet from a JSON file.
- Feature: added the possibility to save a TextClassifierSession as JSON file.
- Client:
- Feature: added "classify" sub-command and its options.
- Documentation:
- Feature: updated to v3.5.0.
v3.0.0 - BREAKING CHANGES
- Decoupled LabeledExample class from the tokenization process; TokenizedExample class added.
- Removed all the Invoke() statements to make the code is more compact.
- Split the monolithic ObjectMother class to namespace-based ObjectMother classes to improve maintainability.
- Split the monolithic MessageCollection class to namespace-based MessageCollection classes to improve maintainability.
- TextClassifier, added LogAsciiBanner() method.
- Add a preliminary client structure for NW.NGramTextClassificationClient.
v2.2.0
- Fixed root cause for a System.NullReferenceException when running TryPredictLabel() for "/".
- TextClassifierTests: added tests for TryPredictLabel().
- NW.NGramTextClassification: .NET Standard 2.0 => 2.1.
- NW.NGramTextClassification.UnitTests, NW.NGramTextClassificationClient: .NET Core 2.1 => 3.1.
v2.1.0
- Added support for XML documentation file.
- TextClassifier: added TryPredictLabel().
v2.0.0 - BREAKING CHANGES
- Added support for sub-namespaces.
- Added XML documentation for all the methods and constructors (summaries).
- Re-organized part of the unit tests (shared items in ObjectMother).
- NGramTokenizer: nearly a complete re-design.
- LabeledExampleFactory: nearly a complete re-design.
v1.1.0
- TextClassifierComponents: fixed Issue #1 (thanks @projectstann!) related to DefaultTextTruncatingFunction.
- NGramTokenizer: added "TheRuleCantBeAppliedTo" exception message.
- NGramTokenizer: fixed "System.ArgumentException: The 'length''s value is greater or equal than 'arr.Length''s value.".
- NGramTokenizer, DoFor<T>: added extra Validation.
v1.0.0
- Initial release.