weaviate

Форк
0
188 строк · 4.4 Кб
1
//                           _       _
2
// __      _____  __ ___   ___  __ _| |_ ___
3
// \ \ /\ / / _ \/ _` \ \ / / |/ _` | __/ _ \
4
//  \ V  V /  __/ (_| |\ V /| | (_| | ||  __/
5
//   \_/\_/ \___|\__,_| \_/ |_|\__,_|\__\___|
6
//
7
//  Copyright © 2016 - 2024 Weaviate B.V. All rights reserved.
8
//
9
//  CONTACT: hello@weaviate.io
10
//
11

12
package modtransformers
13

14
import (
15
	"context"
16
	"testing"
17

18
	"github.com/sirupsen/logrus"
19
	ltest "github.com/sirupsen/logrus/hooks/test"
20
	"github.com/stretchr/testify/assert"
21
	"github.com/stretchr/testify/require"
22
	"github.com/weaviate/weaviate/entities/models"
23
	"github.com/weaviate/weaviate/entities/schema"
24
)
25

26
func TestConfigDefaults(t *testing.T) {
27
	t.Run("for properties", func(t *testing.T) {
28
		def := New().ClassConfigDefaults()
29

30
		assert.Equal(t, true, def["vectorizeClassName"])
31
		assert.Equal(t, "masked_mean", def["poolingStrategy"])
32
	})
33

34
	t.Run("for the class", func(t *testing.T) {
35
		dt := schema.DataTypeText
36
		def := New().PropertyConfigDefaults(&dt)
37
		assert.Equal(t, false, def["vectorizePropertyName"])
38
		assert.Equal(t, false, def["skip"])
39
	})
40
}
41

42
func TestConfigValidator(t *testing.T) {
43
	t.Run("all usable props no-indexed", func(t *testing.T) {
44
		t.Run("all schema vectorization turned off", func(t *testing.T) {
45
			class := &models.Class{
46
				Vectorizer: "text2vec-contextionary",
47
				Class:      "ValidName",
48
				Properties: []*models.Property{
49
					{
50
						DataType: []string{"text"},
51
						Name:     "description",
52
					},
53
					{
54
						DataType:     schema.DataTypeText.PropString(),
55
						Tokenization: models.PropertyTokenizationWhitespace,
56
						Name:         "name",
57
					},
58
					{
59
						DataType: []string{"int"},
60
						Name:     "amount",
61
					},
62
				},
63
			}
64

65
			logger, _ := ltest.NewNullLogger()
66
			v := NewConfigValidator(logger)
67
			err := v.Do(context.Background(), class, nil, &fakeIndexChecker{
68
				vectorizePropertyName: false,
69
				vectorizeClassName:    false,
70
				propertyIndexed:       false,
71
			})
72
			assert.NotNil(t, err)
73
		})
74
	})
75
}
76

77
func TestConfigValidator_RiskOfDuplicateVectors(t *testing.T) {
78
	type test struct {
79
		name          string
80
		in            *models.Class
81
		expectWarning bool
82
		indexChecker  *fakeIndexChecker
83
	}
84

85
	tests := []test{
86
		{
87
			name: "usable properties",
88
			in: &models.Class{
89
				Class: "ValidName",
90
				Properties: []*models.Property{
91
					{
92
						DataType: []string{string(schema.DataTypeText)},
93
						Name:     "textProp",
94
					},
95
				},
96
			},
97
			expectWarning: false,
98
			indexChecker: &fakeIndexChecker{
99
				vectorizePropertyName: false,
100
				vectorizeClassName:    true,
101
				propertyIndexed:       true,
102
			},
103
		},
104
		{
105
			name: "no properties",
106
			in: &models.Class{
107
				Class: "ValidName",
108
			},
109
			expectWarning: true,
110
			indexChecker: &fakeIndexChecker{
111
				vectorizePropertyName: false,
112
				vectorizeClassName:    true,
113
				propertyIndexed:       false,
114
			},
115
		},
116
		{
117
			name: "usable properties, but they are no-indexed",
118
			in: &models.Class{
119
				Class: "ValidName",
120
				Properties: []*models.Property{
121
					{
122
						DataType: []string{string(schema.DataTypeText)},
123
						Name:     "textProp",
124
					},
125
				},
126
			},
127
			expectWarning: true,
128
			indexChecker: &fakeIndexChecker{
129
				vectorizePropertyName: false,
130
				vectorizeClassName:    true,
131
				propertyIndexed:       false,
132
			},
133
		},
134
		{
135
			name: "only unusable properties",
136
			in: &models.Class{
137
				Class: "ValidName",
138
				Properties: []*models.Property{
139
					{
140
						DataType: []string{string(schema.DataTypeInt)},
141
						Name:     "intProp",
142
					},
143
				},
144
			},
145
			expectWarning: true,
146
			indexChecker: &fakeIndexChecker{
147
				vectorizePropertyName: false,
148
				vectorizeClassName:    true,
149
				propertyIndexed:       false,
150
			},
151
		},
152
	}
153

154
	for _, test := range tests {
155
		t.Run(test.name, func(t *testing.T) {
156
			logger, hook := ltest.NewNullLogger()
157
			v := NewConfigValidator(logger)
158
			err := v.Do(context.Background(), test.in, nil, test.indexChecker)
159
			require.Nil(t, err)
160

161
			entry := hook.LastEntry()
162
			if test.expectWarning {
163
				require.NotNil(t, entry)
164
				assert.Equal(t, logrus.WarnLevel, entry.Level)
165
			} else {
166
				assert.Nil(t, entry)
167
			}
168
		})
169
	}
170
}
171

172
type fakeIndexChecker struct {
173
	vectorizeClassName    bool
174
	vectorizePropertyName bool
175
	propertyIndexed       bool
176
}
177

178
func (f *fakeIndexChecker) VectorizeClassName() bool {
179
	return f.vectorizeClassName
180
}
181

182
func (f *fakeIndexChecker) VectorizePropertyName(propName string) bool {
183
	return f.vectorizePropertyName
184
}
185

186
func (f *fakeIndexChecker) PropertyIndexed(propName string) bool {
187
	return f.propertyIndexed
188
}
189

Использование cookies

Мы используем файлы cookie в соответствии с Политикой конфиденциальности и Политикой использования cookies.

Нажимая кнопку «Принимаю», Вы даете АО «СберТех» согласие на обработку Ваших персональных данных в целях совершенствования нашего веб-сайта и Сервиса GitVerse, а также повышения удобства их использования.

Запретить использование cookies Вы можете самостоятельно в настройках Вашего браузера.