File: test_query_knowledgebase.py

package info (click to toggle)
python-azure 20250603%2Bgit-1
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 851,724 kB
  • sloc: python: 7,362,925; ansic: 804; javascript: 287; makefile: 195; sh: 145; xml: 109
file content (470 lines) | stat: -rw-r--r-- 18,440 bytes parent folder | download
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
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
# coding=utf-8
# ------------------------------------
# Copyright (c) Microsoft Corporation.
# Licensed under the MIT License.
# ------------------------------------
import pytest

from azure.ai.language.questionanswering import QuestionAnsweringClient
from azure.ai.language.questionanswering._operations._operations import \
    build_question_answering_get_answers_request as build_get_answers_request
from azure.ai.language.questionanswering.models import (
    AnswersOptions,
    KnowledgeBaseAnswerContext,
    ShortAnswerOptions,
    MetadataFilter,
    QueryFilters,
)
from azure.core.credentials import AzureKeyCredential

from testcase import QuestionAnsweringTestCase


class TestQnAKnowledgeBase(QuestionAnsweringTestCase):

    def test_query_knowledgebase_llc(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        json_content = {
            "question": "Ports and connectors",
            "top": 3,
            "context": {
                "previousUserQuery": "Meet Surface Pro 4",
                "previousQnAId": 4
            }
        }
        request = build_get_answers_request(
            json=json_content,
            project_name=qna_creds["qna_project"],
            deployment_name='test'
        )
        with client:
            response = client.send_request(request)
            assert response.status_code == 200

        output = response.json()
        assert output
        assert output.get('answers')
        for answer in output['answers']:
            assert answer.get('answer')
            assert answer.get('confidenceScore')
            assert answer.get('id')
            assert answer.get('source')
            assert answer.get('metadata') is not None
            assert not answer.get('answerSpan')

            assert answer.get('questions')
            for question in answer['questions']:
                assert question

            assert answer.get('dialog')
            assert answer['dialog'].get('isContextOnly') is not None
            assert answer['dialog'].get('prompts') is not None
            if answer['dialog'].get('prompts'):
                for prompt in answer['dialog']['prompts']:
                    assert prompt.get('displayOrder') is not None
                    assert prompt.get('qnaId')
                    assert prompt.get('displayText')

    def test_query_knowledgebase_llc_with_answerspan(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        json_content = {
            "question": "Ports and connectors",
            "top": 3,
            "context": {
                "previousUserQuery": "Meet Surface Pro 4",
                "previousQnAId": 4
            },
            "answerSpanRequest": {
                "enable": True,
                "confidenceScoreThreshold": 0.1,
                "topAnswersWithSpan": 1
            }
        }
        request = build_get_answers_request(
            json=json_content,
            project_name=qna_creds["qna_project"],
            deployment_name='test'
        )
        with client:
            response = client.send_request(request)
            assert response.status_code == 200

        output = response.json()
        assert output
        assert output.get('answers')
        for answer in output['answers']:
            assert answer.get('answer')
            assert answer.get('confidenceScore')
            assert answer.get('id')
            assert answer.get('source')
            assert answer.get('metadata') is not None

            if answer.get('answerSpan'):
                assert answer['answerSpan'].get('text')
                assert answer['answerSpan'].get('confidenceScore')
                assert answer['answerSpan'].get('offset') is not None
                assert answer['answerSpan'].get('length')

            assert answer.get('questions')
            for question in answer['questions']:
                assert question

            assert answer.get('dialog')
            assert answer['dialog'].get('isContextOnly') is not None
            assert answer['dialog'].get('prompts') is not None
            if answer['dialog'].get('prompts'):
                for prompt in answer['dialog']['prompts']:
                    assert prompt.get('displayOrder') is not None
                    assert prompt.get('qnaId')
                    assert prompt.get('displayText')

    def test_query_knowledgebase(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        query_params = AnswersOptions(
            question="Ports and connectors",
            top=3,
            answer_context=KnowledgeBaseAnswerContext(
                previous_question="Meet Surface Pro 4",
                previous_qna_id=4
            )
        )

        with client:
            output = client.get_answers(
                query_params,
                project_name=qna_creds["qna_project"],
                deployment_name='test'
            )

        assert output.answers
        for answer in output.answers:
            assert answer.answer
            assert answer.confidence
            assert answer.qna_id
            assert answer.source
            assert answer.metadata is not None
            assert not answer.short_answer

            assert answer.questions
            for question in answer.questions:
                assert question

            assert answer.dialog
            assert answer.dialog.is_context_only is not None
            assert answer.dialog.prompts is not None
            if answer.dialog.prompts:
                for prompt in answer.dialog.prompts:
                    assert prompt.display_order is not None
                    assert prompt.qna_id
                    assert prompt.display_text

    def test_query_knowledgebase_aad(self, recorded_test, qna_creds):
        token = self.get_credential(QuestionAnsweringClient)
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], token)
        query_params = AnswersOptions(
            question="Ports and connectors",
            top=3,
            answer_context=KnowledgeBaseAnswerContext(
                previous_question="Meet Surface Pro 4",
                previous_qna_id=4
            )
        )

        with client:
            output = client.get_answers(
                query_params,
                project_name=qna_creds["qna_project"],
                deployment_name='test'
            )

        assert output.answers
        for answer in output.answers:
            assert answer.answer
            assert answer.confidence
            assert answer.qna_id
            assert answer.source
            assert answer.metadata is not None
            assert not answer.short_answer

            assert answer.questions
            for question in answer.questions:
                assert question

            assert answer.dialog
            assert answer.dialog.is_context_only is not None
            assert answer.dialog.prompts is not None
            if answer.dialog.prompts:
                for prompt in answer.dialog.prompts:
                    assert prompt.display_order is not None
                    assert prompt.qna_id
                    assert prompt.display_text

    def test_query_knowledgebase_with_answerspan(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        query_params = AnswersOptions(
            question="Ports and connectors",
            top=3,
            answer_context=KnowledgeBaseAnswerContext(
                previous_question="Meet Surface Pro 4",
                previous_qna_id=4
            ),
            short_answer_options=ShortAnswerOptions(
                confidence_threshold=0.1,
                top=2
            )
        )

        with client:
            output = client.get_answers(
                query_params,
                project_name=qna_creds["qna_project"],
                deployment_name='test'
            )

        assert output.answers
        for answer in output.answers:
            assert answer.answer
            assert answer.confidence
            assert answer.qna_id
            assert answer.source
            assert answer.metadata is not None

            if answer.short_answer:
                assert answer.short_answer.text
                assert answer.short_answer.confidence
                assert answer.short_answer.offset is not None
                assert answer.short_answer.length

            assert answer.questions
            for question in answer.questions:
                assert question

            assert answer.dialog
            assert answer.dialog.is_context_only is not None
            assert answer.dialog.prompts is not None
            if answer.dialog.prompts:
                for prompt in answer.dialog.prompts:
                    assert prompt.display_order is not None
                    assert prompt.qna_id
                    assert prompt.display_text

    @pytest.mark.live_test_only("Needs re-recording to work with new common sanitizers")
    def test_query_knowledgebase_with_dictparams(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        query_params = {
            "question": "How long should my Surface battery last?",
            "top": 3,
            "userId": "sd53lsY=",
            "confidenceScoreThreshold": 0.2,
            "answerSpanRequest": {
                "enable": True,
                "confidenceScoreThreshold": 0.2,
                "topAnswersWithSpan": 1
            },
            "includeUnstructuredSources": True
        }

        with client:
            output = client.get_answers(
                query_params,
                project_name=qna_creds["qna_project"],
                deployment_name='test'
            )

        assert len(output.answers) == 3
        confident_answers = [a for a in output.answers if a.confidence > 0.7]
        assert len(confident_answers) == 1
        assert confident_answers[0].source == "surface-book-user-guide-EN.pdf"

    @pytest.mark.live_test_only("Needs re-recording to work with new common sanitizers")
    def test_query_knowledgebase_overload(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        with client:
            output = client.get_answers(
                project_name=qna_creds["qna_project"],
                deployment_name='test',
                question="How long should my Surface battery last?",
                top=3,
                user_id="sd53lsY=",
                confidence_threshold=0.2,
                short_answer_options=ShortAnswerOptions(
                    confidence_threshold=0.2,
                    top=1
                ),
                include_unstructured_sources=True
            )

        assert len(output.answers) == 3
        confident_answers = [a for a in output.answers if a.confidence > 0.7]
        assert len(confident_answers) == 1
        assert confident_answers[0].source == "surface-book-user-guide-EN.pdf"

    @pytest.mark.live_test_only("Needs re-recording to work with new common sanitizers")
    def test_query_knowledgebase_with_followup(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        with client:
            query_params = AnswersOptions(
                question="How long should my Surface battery last?",
                top=3,
                user_id="sd53lsY=",
                confidence_threshold=0.2,
                short_answer_options=ShortAnswerOptions(
                    confidence_threshold=0.2,
                    top=1
                ),
                include_unstructured_sources=True
            )

            output = client.get_answers(
                query_params,
                project_name=qna_creds["qna_project"],
                deployment_name='test'
            )
            confident_answers = [a for a in output.answers if a.confidence > 0.7]
            assert len(confident_answers) == 1
            assert confident_answers[0].source == "surface-book-user-guide-EN.pdf"

            query_params = AnswersOptions(
                question="How long it takes to charge Surface?",
                top=3,
                user_id="sd53lsY=",
                confidence_threshold=0.2,
                answer_context=KnowledgeBaseAnswerContext(
                    previous_question="How long should my Surface battery last?",
                    previous_qna_id=confident_answers[0].qna_id
                ),
                short_answer_options=ShortAnswerOptions(
                    confidence_threshold=0.2,
                    top=1
                ),
                include_unstructured_sources=True
            )
            output = client.get_answers(
                query_params,
                project_name=qna_creds["qna_project"],
                deployment_name='test'
            )

            assert output.answers
            confident_answers = [a for a in output.answers if a.confidence > 0.48]
            assert len(confident_answers) == 1
            assert confident_answers[0].short_answer.text == " two to four hours"


    def test_query_knowledgebase_only_id(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        with client:
            query_params = AnswersOptions(
                qna_id=19
            )

            output = client.get_answers(
                query_params,
                project_name=qna_creds["qna_project"],
                deployment_name='test'
            )

            assert len(output.answers) == 1

    def test_query_knowledgebase_python_dict(self, recorded_test, qna_creds):
        client = QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"]))
        with client:
            query_params = {"qna_id": 19}

            output = client.get_answers(
                query_params,
                project_name=qna_creds["qna_project"],
                deployment_name='test'
            )

            assert len(output.answers) == 1

    def test_query_knowledgebase_overload_positional_and_kwarg(self):
        with QuestionAnsweringClient("http://fake.com", AzureKeyCredential("123")) as client:
            with pytest.raises(TypeError):
                client.get_answers("positional_one", "positional_two")
            with pytest.raises(TypeError):
                client.get_answers("positional_options_bag", options="options bag by name")
            with pytest.raises(TypeError):
                client.get_answers(
                    options={'qnaId': 15},
                    project_name="hello",
                    deployment_name='test'
                )
            with pytest.raises(TypeError):
                client.get_answers(
                    {'qnaId': 15},
                    question='Why?',
                    project_name="hello",
                    deployment_name='test'
                )

    def test_query_knowledgebase_question_or_qna_id(self):
        with QuestionAnsweringClient("http://fake.com", AzureKeyCredential("123")) as client:

            options = AnswersOptions()
            with pytest.raises(TypeError):
                client.get_answers(
                    options,
                    project_name="hello",
                    deployment_name='test'
                )

            with pytest.raises(TypeError):
                client.get_answers(
                    project_name="hello",
                    deployment_name='test'
                )

    def test_query_knowledgebase_filter(self, recorded_test, qna_creds):
        """Thanks to @heaths for this test!"""
        filters = QueryFilters(
            metadata_filter=MetadataFilter(
                metadata=[
                    ("explicitlytaggedheading", "check the battery level"),
                    ("explicitlytaggedheading", "make your battery last")
                ],
                logical_operation="OR"
            ),
        )
        with QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"])) as client:
            response = client.get_answers(
                project_name=qna_creds["qna_project"],
                deployment_name='test',
                question="Battery life",
                filters=filters,
                top=3,
            )
            assert len(response.answers) == 2
            assert any(
                [a for a in response.answers if a.metadata.get('explicitlytaggedheading') == "check the battery level"]
            )
            assert any(
                [a for a in response.answers if a.metadata.get('explicitlytaggedheading') == "make your battery last"]
            )

    def test_query_knowledgebase_filter_dict_params(self, recorded_test, qna_creds):
        filters = {
            "metadataFilter": {
                "metadata": [
                    ("explicitlytaggedheading", "check the battery level"),
                    ("explicitlytaggedheading", "make your battery last")
                ],
                "logicalOperation": "or"
            },
        }
        with QuestionAnsweringClient(qna_creds["qna_endpoint"], AzureKeyCredential(qna_creds["qna_key"])) as client:
            response = client.get_answers(
                project_name=qna_creds["qna_project"],
                deployment_name='test',
                question="Battery life",
                filters=filters,
                top=3,
            )
            assert len(response.answers) == 2
            assert any(
                [a for a in response.answers if a.metadata.get('explicitlytaggedheading') == "check the battery level"]
            )
            assert any(
                [a for a in response.answers if a.metadata.get('explicitlytaggedheading') == "make your battery last"]
            )