|
1 | 1 | import unittest |
| 2 | +from typing import Any |
2 | 3 |
|
3 | | -from ..tools import commands # noqa |
| 4 | +from ..tools import commands, parse, validate |
| 5 | +from ..tools.utils import JsonType |
| 6 | + |
| 7 | + |
| 8 | +def evaluation_function( |
| 9 | + response: Any, answer: Any, params: JsonType |
| 10 | +) -> JsonType: |
| 11 | + if params.get("raise", False): |
| 12 | + raise ValueError("raised") |
| 13 | + |
| 14 | + force_true = params.get("force", False) |
| 15 | + |
| 16 | + return {"is_correct": (response == answer) or force_true} |
4 | 17 |
|
5 | 18 |
|
6 | 19 | class TestCommandsModule(unittest.TestCase): |
| 20 | + def setUp(self) -> None: |
| 21 | + commands.evaluation_function = evaluation_function |
| 22 | + return super().setUp() |
| 23 | + |
| 24 | + def tearDown(self) -> None: |
| 25 | + commands.evaluation_function = None |
| 26 | + return super().tearDown() |
| 27 | + |
7 | 28 | def test_valid_eval_command(self): |
8 | | - ... |
| 29 | + event = { |
| 30 | + "body": {"response": "hello", "answer": "world", "params": {}} |
| 31 | + } |
| 32 | + response = commands.evaluate(event) |
| 33 | + |
| 34 | + self.assertIn("result", response) |
| 35 | + self.assertIn("is_correct", response["result"]) # type: ignore |
9 | 36 |
|
10 | 37 | def test_invalid_eval_args_raises_parse_error(self): |
11 | | - ... |
| 38 | + event = {"headers": "any", "other": "params"} |
| 39 | + |
| 40 | + with self.assertRaises(parse.ParseError) as e: |
| 41 | + commands.evaluate(event) |
| 42 | + |
| 43 | + self.assertEqual( |
| 44 | + e.exception.message, "No data supplied in request body." |
| 45 | + ) |
12 | 46 |
|
13 | 47 | def test_invalid_eval_schema_raises_validation_error(self): |
14 | | - ... |
| 48 | + event = {"body": {"response": "hello", "params": {}}} |
| 49 | + |
| 50 | + with self.assertRaises(validate.ValidationError) as e: |
| 51 | + commands.evaluate(event) |
| 52 | + |
| 53 | + self.assertEqual( |
| 54 | + e.exception.message, |
| 55 | + "Failed to validate body against the evaluation schema.", |
| 56 | + ) |
15 | 57 |
|
16 | 58 | def test_single_feedback_case(self): |
17 | | - ... |
| 59 | + event = { |
| 60 | + "body": { |
| 61 | + "response": "hello", |
| 62 | + "answer": "hello", |
| 63 | + "params": { |
| 64 | + "cases": [ |
| 65 | + { |
| 66 | + "answer": "other", |
| 67 | + "feedback": "should be 'other'.", |
| 68 | + "mark": 0, |
| 69 | + } |
| 70 | + ] |
| 71 | + }, |
| 72 | + } |
| 73 | + } |
| 74 | + |
| 75 | + response = commands.evaluate(event) |
| 76 | + result = response["result"] # type: ignore |
| 77 | + |
| 78 | + self.assertTrue(result["is_correct"]) |
| 79 | + self.assertNotIn("match", result) |
| 80 | + self.assertNotIn("feedback", result) |
18 | 81 |
|
19 | 82 | def test_single_feedback_case_match(self): |
20 | | - ... |
| 83 | + event = { |
| 84 | + "body": { |
| 85 | + "response": "hello", |
| 86 | + "answer": "world", |
| 87 | + "params": { |
| 88 | + "cases": [ |
| 89 | + { |
| 90 | + "answer": "hello", |
| 91 | + "feedback": "should be 'hello'.", |
| 92 | + } |
| 93 | + ] |
| 94 | + }, |
| 95 | + } |
| 96 | + } |
| 97 | + |
| 98 | + response = commands.evaluate(event) |
| 99 | + result = response["result"] # type: ignore |
| 100 | + |
| 101 | + self.assertFalse(result["is_correct"]) |
| 102 | + self.assertEqual(result["match"], 0) |
| 103 | + self.assertEqual(result["feedback"], "should be 'hello'.") |
21 | 104 |
|
22 | 105 | def test_case_warning_data_structure(self): |
23 | | - ... |
| 106 | + event = { |
| 107 | + "body": { |
| 108 | + "response": "hello", |
| 109 | + "answer": "world", |
| 110 | + "params": { |
| 111 | + "cases": [ |
| 112 | + { |
| 113 | + "feedback": "should be 'hello'.", |
| 114 | + } |
| 115 | + ] |
| 116 | + }, |
| 117 | + } |
| 118 | + } |
| 119 | + |
| 120 | + response = commands.evaluate(event) |
| 121 | + result = response["result"] # type: ignore |
| 122 | + |
| 123 | + self.assertIn("warnings", result) |
| 124 | + warning = result["warnings"].pop() # type: ignore |
| 125 | + |
| 126 | + self.assertDictEqual( |
| 127 | + warning, |
| 128 | + {"case": 0, "message": "Missing answer/feedback field"}, |
| 129 | + ) |
24 | 130 |
|
25 | 131 | def test_multiple_feedback_cases_single_match(self): |
26 | | - ... |
| 132 | + event = { |
| 133 | + "body": { |
| 134 | + "response": "yes", |
| 135 | + "answer": "world", |
| 136 | + "params": { |
| 137 | + "cases": [ |
| 138 | + { |
| 139 | + "answer": "hello", |
| 140 | + "feedback": "should be 'hello'.", |
| 141 | + }, |
| 142 | + { |
| 143 | + "answer": "yes", |
| 144 | + "feedback": "should be 'yes'.", |
| 145 | + }, |
| 146 | + { |
| 147 | + "answer": "no", |
| 148 | + "feedback": "should be 'no'.", |
| 149 | + }, |
| 150 | + ] |
| 151 | + }, |
| 152 | + } |
| 153 | + } |
| 154 | + |
| 155 | + response = commands.evaluate(event) |
| 156 | + result = response["result"] # type: ignore |
| 157 | + |
| 158 | + self.assertFalse(result["is_correct"]) |
| 159 | + self.assertEqual(result["match"], 1) |
| 160 | + self.assertEqual(result["feedback"], "should be 'yes'.") |
27 | 161 |
|
28 | 162 | def test_multiple_feedback_cases_multiple_matches(self): |
29 | | - ... |
| 163 | + event = { |
| 164 | + "body": { |
| 165 | + "response": "yes", |
| 166 | + "answer": "world", |
| 167 | + "params": { |
| 168 | + "cases": [ |
| 169 | + { |
| 170 | + "answer": "hello", |
| 171 | + "feedback": "should be 'hello'.", |
| 172 | + "params": {"force": True}, |
| 173 | + }, |
| 174 | + { |
| 175 | + "answer": "yes", |
| 176 | + "feedback": "should be 'yes'.", |
| 177 | + }, |
| 178 | + { |
| 179 | + "answer": "no", |
| 180 | + "feedback": "should be 'no'.", |
| 181 | + }, |
| 182 | + ] |
| 183 | + }, |
| 184 | + } |
| 185 | + } |
| 186 | + |
| 187 | + response = commands.evaluate(event) |
| 188 | + result = response["result"] # type: ignore |
| 189 | + |
| 190 | + self.assertFalse(result["is_correct"]) |
| 191 | + self.assertEqual(result["match"], 0) |
| 192 | + self.assertEqual(result["feedback"], "should be 'hello'.") |
| 193 | + |
| 194 | + def test_case_params_overwrite_eval_params(self): |
| 195 | + event = { |
| 196 | + "body": { |
| 197 | + "response": "hello", |
| 198 | + "answer": "world", |
| 199 | + "params": { |
| 200 | + "force": True, |
| 201 | + "cases": [ |
| 202 | + { |
| 203 | + "answer": "yes", |
| 204 | + "feedback": "should be 'yes'.", |
| 205 | + "params": {"force": False}, |
| 206 | + } |
| 207 | + ], |
| 208 | + }, |
| 209 | + } |
| 210 | + } |
| 211 | + |
| 212 | + response = commands.evaluate(event) |
| 213 | + result = response["result"] # type: ignore |
| 214 | + |
| 215 | + self.assertTrue(result["is_correct"]) |
| 216 | + self.assertNotIn("match", result) |
| 217 | + self.assertNotIn("feedback", result) |
30 | 218 |
|
31 | 219 | def test_invalid_case_entry_doesnt_raise_exception(self): |
32 | | - ... |
| 220 | + event = { |
| 221 | + "body": { |
| 222 | + "response": "hello", |
| 223 | + "answer": "world", |
| 224 | + "params": { |
| 225 | + "cases": [ |
| 226 | + { |
| 227 | + "answer": "hello", |
| 228 | + "feedback": "should be 'hello'.", |
| 229 | + "params": {"raise": True}, |
| 230 | + } |
| 231 | + ] |
| 232 | + }, |
| 233 | + } |
| 234 | + } |
| 235 | + |
| 236 | + response = commands.evaluate(event) |
| 237 | + result = response["result"] # type: ignore |
| 238 | + |
| 239 | + self.assertIn("warnings", result) |
| 240 | + warning = result["warnings"].pop() # type: ignore |
| 241 | + |
| 242 | + self.assertDictEqual( |
| 243 | + warning, |
| 244 | + { |
| 245 | + "case": 0, |
| 246 | + "message": "An exception was raised while executing " |
| 247 | + "the evaluation function.", |
| 248 | + "detail": "raised", |
| 249 | + }, |
| 250 | + ) |
33 | 251 |
|
34 | 252 | def test_multiple_matched_cases_are_combined_and_warned(self): |
35 | | - ... |
| 253 | + event = { |
| 254 | + "body": { |
| 255 | + "response": "yes", |
| 256 | + "answer": "world", |
| 257 | + "params": { |
| 258 | + "cases": [ |
| 259 | + { |
| 260 | + "answer": "hello", |
| 261 | + "feedback": "should be 'hello'.", |
| 262 | + "params": {"force": True}, |
| 263 | + }, |
| 264 | + { |
| 265 | + "answer": "yes", |
| 266 | + "feedback": "should be 'yes'.", |
| 267 | + }, |
| 268 | + { |
| 269 | + "answer": "no", |
| 270 | + "feedback": "should be 'no'.", |
| 271 | + }, |
| 272 | + ] |
| 273 | + }, |
| 274 | + } |
| 275 | + } |
36 | 276 |
|
37 | | - def test_overriding_eval_feedback_in_case(self): |
38 | | - ... |
| 277 | + response = commands.evaluate(event) |
| 278 | + result = response["result"] # type: ignore |
39 | 279 |
|
40 | | - def test_overriding_is_correct_with_case_mark(self): |
41 | | - ... |
| 280 | + self.assertIn("warnings", result) |
| 281 | + warning = result["warnings"].pop() # type: ignore |
42 | 282 |
|
43 | | - def test_valid_preview_command(self): |
44 | | - ... |
| 283 | + self.assertDictEqual( |
| 284 | + warning, |
| 285 | + { |
| 286 | + "message": "Cases 0, 1 were matched. " |
| 287 | + "Only the first one's feedback was returned", |
| 288 | + }, |
| 289 | + ) |
| 290 | + |
| 291 | + def test_overriding_eval_feedback_to_correct_case(self): |
| 292 | + event = { |
| 293 | + "body": { |
| 294 | + "response": "hello", |
| 295 | + "answer": "world", |
| 296 | + "params": { |
| 297 | + "cases": [ |
| 298 | + { |
| 299 | + "answer": "hello", |
| 300 | + "feedback": "should be 'hello'.", |
| 301 | + "mark": 1, |
| 302 | + } |
| 303 | + ] |
| 304 | + }, |
| 305 | + } |
| 306 | + } |
| 307 | + |
| 308 | + response = commands.evaluate(event) |
| 309 | + result = response["result"] # type: ignore |
| 310 | + |
| 311 | + self.assertTrue(result["is_correct"]) |
| 312 | + self.assertEqual(result["match"], 0) |
| 313 | + self.assertEqual(result["feedback"], "should be 'hello'.") |
45 | 314 |
|
46 | | - def test_invalid_preview_args_raises_parse_error(self): |
47 | | - ... |
| 315 | + def test_overriding_eval_feedback_to_incorrect_case(self): |
| 316 | + event = { |
| 317 | + "body": { |
| 318 | + "response": "hello", |
| 319 | + "answer": "hello", |
| 320 | + "params": { |
| 321 | + "cases": [ |
| 322 | + { |
| 323 | + "answer": "hello", |
| 324 | + "feedback": "should be 'hello'.", |
| 325 | + "mark": 0, |
| 326 | + } |
| 327 | + ] |
| 328 | + }, |
| 329 | + } |
| 330 | + } |
| 331 | + |
| 332 | + response = commands.evaluate(event) |
| 333 | + result = response["result"] # type: ignore |
| 334 | + |
| 335 | + self.assertFalse(result["is_correct"]) |
| 336 | + self.assertEqual(result["match"], 0) |
| 337 | + self.assertEqual(result["feedback"], "should be 'hello'.") |
| 338 | + |
| 339 | + def test_valid_preview_command(self): |
| 340 | + event = {"body": {"response": "hello"}} |
48 | 341 |
|
49 | | - def test_invalid_preview_schema_raises_validation_error(self): |
50 | | - ... |
| 342 | + response = commands.preview(event) |
| 343 | + result = response["result"] # type: ignore |
51 | 344 |
|
52 | | - def test_handle_healthcheck_command(self): |
53 | | - ... |
| 345 | + self.assertEqual(result["preview"], "hello") |
54 | 346 |
|
55 | 347 |
|
56 | 348 | if __name__ == "__main__": |
|
0 commit comments