parser_test.go 12 KB
Newer Older
1
package parser
2
3

import (
Michael Yang's avatar
Michael Yang committed
4
	"bytes"
5
	"encoding/binary"
Michael Yang's avatar
Michael Yang committed
6
7
	"fmt"
	"io"
8
9
	"strings"
	"testing"
10
	"unicode/utf16"
11
12

	"github.com/stretchr/testify/assert"
Michael Yang's avatar
lint  
Michael Yang committed
13
	"github.com/stretchr/testify/require"
14
15
	"golang.org/x/text/encoding"
	"golang.org/x/text/encoding/unicode"
16
17
)

Michael Yang's avatar
Michael Yang committed
18
func TestParseFileFile(t *testing.T) {
19
20
21
22
23
24
25
26
27
28
29
	input := `
FROM model1
ADAPTER adapter1
LICENSE MIT
PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
`

	reader := strings.NewReader(input)

Michael Yang's avatar
Michael Yang committed
30
	modelfile, err := ParseFile(reader)
Michael Yang's avatar
lint  
Michael Yang committed
31
	require.NoError(t, err)
32
33
34
35
36
37
38
39
40
41

	expectedCommands := []Command{
		{Name: "model", Args: "model1"},
		{Name: "adapter", Args: "adapter1"},
		{Name: "license", Args: "MIT"},
		{Name: "param1", Args: "value1"},
		{Name: "param2", Args: "value2"},
		{Name: "template", Args: "template1"},
	}

Michael Yang's avatar
Michael Yang committed
42
	assert.Equal(t, expectedCommands, modelfile.Commands)
43
44
}

Michael Yang's avatar
Michael Yang committed
45
func TestParseFileFrom(t *testing.T) {
Michael Yang's avatar
tests  
Michael Yang committed
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
	var cases = []struct {
		input    string
		expected []Command
		err      error
	}{
		{
			"FROM foo",
			[]Command{{Name: "model", Args: "foo"}},
			nil,
		},
		{
			"FROM /path/to/model",
			[]Command{{Name: "model", Args: "/path/to/model"}},
			nil,
		},
		{
			"FROM /path/to/model/fp16.bin",
			[]Command{{Name: "model", Args: "/path/to/model/fp16.bin"}},
			nil,
		},
		{
			"FROM llama3:latest",
			[]Command{{Name: "model", Args: "llama3:latest"}},
			nil,
		},
		{
			"FROM llama3:7b-instruct-q4_K_M",
			[]Command{{Name: "model", Args: "llama3:7b-instruct-q4_K_M"}},
			nil,
		},
		{
			"", nil, errMissingFrom,
		},
		{
			"PARAMETER param1 value1",
			nil,
			errMissingFrom,
		},
		{
			"PARAMETER param1 value1\nFROM foo",
			[]Command{{Name: "param1", Args: "value1"}, {Name: "model", Args: "foo"}},
			nil,
		},
	}
90

Michael Yang's avatar
tests  
Michael Yang committed
91
92
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
93
			modelfile, err := ParseFile(strings.NewReader(c.input))
Michael Yang's avatar
lint  
Michael Yang committed
94
			require.ErrorIs(t, err, c.err)
Michael Yang's avatar
Michael Yang committed
95
96
97
			if modelfile != nil {
				assert.Equal(t, c.expected, modelfile.Commands)
			}
Michael Yang's avatar
tests  
Michael Yang committed
98
99
		})
	}
100
101
}

Michael Yang's avatar
Michael Yang committed
102
func TestParseFileParametersMissingValue(t *testing.T) {
103
104
105
106
107
108
109
	input := `
FROM foo
PARAMETER param1
`

	reader := strings.NewReader(input)

Michael Yang's avatar
Michael Yang committed
110
	_, err := ParseFile(reader)
Michael Yang's avatar
lint  
Michael Yang committed
111
	require.ErrorIs(t, err, io.ErrUnexpectedEOF)
112
}
113

Michael Yang's avatar
Michael Yang committed
114
func TestParseFileBadCommand(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
115
116
117
118
	input := `
FROM foo
BADCOMMAND param1 value1
`
Michael Yang's avatar
Michael Yang committed
119
	_, err := ParseFile(strings.NewReader(input))
Michael Yang's avatar
lint  
Michael Yang committed
120
	require.ErrorIs(t, err, errInvalidCommand)
Michael Yang's avatar
Michael Yang committed
121
122
}

Michael Yang's avatar
Michael Yang committed
123
func TestParseFileMessages(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
124
125
126
127
128
129
130
131
	var cases = []struct {
		input    string
		expected []Command
		err      error
	}{
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
132
MESSAGE system You are a file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
133
134
135
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
136
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
137
138
139
140
141
			},
			nil,
		},
		{
			`
142
FROM foo
Michael Yang's avatar
Michael Yang committed
143
MESSAGE system You are a file parser. Always parse things.`,
Michael Yang's avatar
Michael Yang committed
144
145
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
146
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
147
148
149
150
151
152
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
153
MESSAGE system You are a file parser. Always parse things.
154
155
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
Michael Yang's avatar
Michael Yang committed
156
157
158
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
159
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
160
161
162
163
164
165
166
167
168
				{Name: "message", Args: "user: Hey there!"},
				{Name: "message", Args: "assistant: Hello, I want to parse all the things!"},
			},
			nil,
		},
		{
			`
FROM foo
MESSAGE system """
Michael Yang's avatar
Michael Yang committed
169
You are a multiline file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
170
171
172
173
"""
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
174
				{Name: "message", Args: "system: \nYou are a multiline file parser. Always parse things.\n"},
Michael Yang's avatar
Michael Yang committed
175
176
177
178
179
180
181
182
183
			},
			nil,
		},
		{
			`
FROM foo
MESSAGE badguy I'm a bad guy!
`,
			nil,
Michael Yang's avatar
Michael Yang committed
184
			errInvalidMessageRole,
Michael Yang's avatar
Michael Yang committed
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
		},
		{
			`
FROM foo
MESSAGE system
`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
MESSAGE system`,
			nil,
			io.ErrUnexpectedEOF,
		},
	}
202

Michael Yang's avatar
Michael Yang committed
203
204
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
205
			modelfile, err := ParseFile(strings.NewReader(c.input))
Michael Yang's avatar
lint  
Michael Yang committed
206
			require.ErrorIs(t, err, c.err)
Michael Yang's avatar
Michael Yang committed
207
208
209
			if modelfile != nil {
				assert.Equal(t, c.expected, modelfile.Commands)
			}
Michael Yang's avatar
Michael Yang committed
210
211
212
		})
	}
}
213

Michael Yang's avatar
Michael Yang committed
214
func TestParseFileQuoted(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
215
216
217
218
219
220
221
222
	var cases = []struct {
		multiline string
		expected  []Command
		err       error
	}{
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
223
SYSTEM """
Michael Yang's avatar
Michael Yang committed
224
This is a
Michael Yang's avatar
Michael Yang committed
225
multiline system.
Michael Yang's avatar
Michael Yang committed
226
227
228
229
"""
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
230
				{Name: "system", Args: "\nThis is a\nmultiline system.\n"},
Michael Yang's avatar
Michael Yang committed
231
232
233
234
235
236
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
237
SYSTEM """
Michael Yang's avatar
Michael Yang committed
238
This is a
Michael Yang's avatar
Michael Yang committed
239
multiline system."""
Michael Yang's avatar
Michael Yang committed
240
241
242
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
243
				{Name: "system", Args: "\nThis is a\nmultiline system."},
Michael Yang's avatar
Michael Yang committed
244
245
246
247
248
249
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
250
251
SYSTEM """This is a
multiline system."""
Michael Yang's avatar
Michael Yang committed
252
253
254
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
255
				{Name: "system", Args: "This is a\nmultiline system."},
Michael Yang's avatar
Michael Yang committed
256
257
258
259
260
261
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
262
SYSTEM """This is a multiline system."""
Michael Yang's avatar
Michael Yang committed
263
264
265
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
266
				{Name: "system", Args: "This is a multiline system."},
Michael Yang's avatar
Michael Yang committed
267
268
269
270
271
272
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
273
SYSTEM """This is a multiline system.""
Michael Yang's avatar
Michael Yang committed
274
275
276
277
278
279
280
			`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
281
SYSTEM "
Michael Yang's avatar
Michael Yang committed
282
283
284
285
286
287
288
			`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
289
290
SYSTEM """
This is a multiline system with "quotes".
Michael Yang's avatar
Michael Yang committed
291
292
293
294
"""
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
295
				{Name: "system", Args: "\nThis is a multiline system with \"quotes\".\n"},
Michael Yang's avatar
Michael Yang committed
296
297
298
299
300
301
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
302
SYSTEM """"""
Michael Yang's avatar
Michael Yang committed
303
304
305
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
306
				{Name: "system", Args: ""},
Michael Yang's avatar
Michael Yang committed
307
308
309
310
311
312
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
313
SYSTEM ""
Michael Yang's avatar
Michael Yang committed
314
315
316
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
317
				{Name: "system", Args: ""},
Michael Yang's avatar
Michael Yang committed
318
319
320
321
322
323
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
324
SYSTEM "'"
Michael Yang's avatar
Michael Yang committed
325
326
327
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
328
				{Name: "system", Args: "'"},
Michael Yang's avatar
Michael Yang committed
329
330
331
			},
			nil,
		},
Michael Yang's avatar
tests  
Michael Yang committed
332
333
334
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
335
SYSTEM """''"'""'""'"'''''""'""'"""
Michael Yang's avatar
tests  
Michael Yang committed
336
337
338
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
339
340
341
342
343
344
345
346
347
348
349
350
351
				{Name: "system", Args: `''"'""'""'"'''''""'""'`},
			},
			nil,
		},
		{
			`
FROM foo
TEMPLATE """
{{ .Prompt }}
"""`,
			[]Command{
				{Name: "model", Args: "foo"},
				{Name: "template", Args: "\n{{ .Prompt }}\n"},
Michael Yang's avatar
tests  
Michael Yang committed
352
353
354
			},
			nil,
		},
355
356
	}

Michael Yang's avatar
Michael Yang committed
357
358
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
359
			modelfile, err := ParseFile(strings.NewReader(c.multiline))
Michael Yang's avatar
lint  
Michael Yang committed
360
			require.ErrorIs(t, err, c.err)
Michael Yang's avatar
Michael Yang committed
361
362
363
			if modelfile != nil {
				assert.Equal(t, c.expected, modelfile.Commands)
			}
Michael Yang's avatar
Michael Yang committed
364
365
		})
	}
366
367
}

Michael Yang's avatar
Michael Yang committed
368
func TestParseFileParameters(t *testing.T) {
Michael Yang's avatar
tests  
Michael Yang committed
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
	var cases = map[string]struct {
		name, value string
	}{
		"numa true":                    {"numa", "true"},
		"num_ctx 1":                    {"num_ctx", "1"},
		"num_batch 1":                  {"num_batch", "1"},
		"num_gqa 1":                    {"num_gqa", "1"},
		"num_gpu 1":                    {"num_gpu", "1"},
		"main_gpu 1":                   {"main_gpu", "1"},
		"low_vram true":                {"low_vram", "true"},
		"f16_kv true":                  {"f16_kv", "true"},
		"logits_all true":              {"logits_all", "true"},
		"vocab_only true":              {"vocab_only", "true"},
		"use_mmap true":                {"use_mmap", "true"},
		"use_mlock true":               {"use_mlock", "true"},
		"num_thread 1":                 {"num_thread", "1"},
		"num_keep 1":                   {"num_keep", "1"},
		"seed 1":                       {"seed", "1"},
		"num_predict 1":                {"num_predict", "1"},
		"top_k 1":                      {"top_k", "1"},
		"top_p 1.0":                    {"top_p", "1.0"},
		"tfs_z 1.0":                    {"tfs_z", "1.0"},
		"typical_p 1.0":                {"typical_p", "1.0"},
		"repeat_last_n 1":              {"repeat_last_n", "1"},
		"temperature 1.0":              {"temperature", "1.0"},
		"repeat_penalty 1.0":           {"repeat_penalty", "1.0"},
		"presence_penalty 1.0":         {"presence_penalty", "1.0"},
		"frequency_penalty 1.0":        {"frequency_penalty", "1.0"},
		"mirostat 1":                   {"mirostat", "1"},
		"mirostat_tau 1.0":             {"mirostat_tau", "1.0"},
		"mirostat_eta 1.0":             {"mirostat_eta", "1.0"},
		"penalize_newline true":        {"penalize_newline", "true"},
		"stop ### User:":               {"stop", "### User:"},
		"stop ### User: ":              {"stop", "### User: "},
		"stop \"### User:\"":           {"stop", "### User:"},
		"stop \"### User: \"":          {"stop", "### User: "},
		"stop \"\"\"### User:\"\"\"":   {"stop", "### User:"},
		"stop \"\"\"### User:\n\"\"\"": {"stop", "### User:\n"},
		"stop <|endoftext|>":           {"stop", "<|endoftext|>"},
		"stop <|eot_id|>":              {"stop", "<|eot_id|>"},
		"stop </s>":                    {"stop", "</s>"},
Michael Yang's avatar
Michael Yang committed
410
	}
411

Michael Yang's avatar
tests  
Michael Yang committed
412
413
	for k, v := range cases {
		t.Run(k, func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
414
415
			var b bytes.Buffer
			fmt.Fprintln(&b, "FROM foo")
Michael Yang's avatar
tests  
Michael Yang committed
416
			fmt.Fprintln(&b, "PARAMETER", k)
Michael Yang's avatar
Michael Yang committed
417
			modelfile, err := ParseFile(&b)
Michael Yang's avatar
lint  
Michael Yang committed
418
			require.NoError(t, err)
Michael Yang's avatar
tests  
Michael Yang committed
419
420
421
422

			assert.Equal(t, []Command{
				{Name: "model", Args: "foo"},
				{Name: v.name, Args: v.value},
Michael Yang's avatar
Michael Yang committed
423
			}, modelfile.Commands)
Michael Yang's avatar
Michael Yang committed
424
425
426
427
		})
	}
}

Michael Yang's avatar
Michael Yang committed
428
func TestParseFileComments(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
429
430
431
432
433
434
435
	var cases = []struct {
		input    string
		expected []Command
	}{
		{
			`
# comment
436
FROM foo
Michael Yang's avatar
Michael Yang committed
437
438
439
440
441
442
	`,
			[]Command{
				{Name: "model", Args: "foo"},
			},
		},
	}
443

Michael Yang's avatar
Michael Yang committed
444
445
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
446
			modelfile, err := ParseFile(strings.NewReader(c.input))
Michael Yang's avatar
lint  
Michael Yang committed
447
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
448
			assert.Equal(t, c.expected, modelfile.Commands)
Michael Yang's avatar
Michael Yang committed
449
450
		})
	}
451
}
Michael Yang's avatar
Michael Yang committed
452

Michael Yang's avatar
Michael Yang committed
453
func TestParseFileFormatParseFile(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
454
455
456
457
458
459
460
461
	var cases = []string{
		`
FROM foo
ADAPTER adapter1
LICENSE MIT
PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
Michael Yang's avatar
Michael Yang committed
462
MESSAGE system You are a file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
`,
		`
FROM foo
ADAPTER adapter1
LICENSE MIT
PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
MESSAGE system """
You are a store greeter. Always responsed with "Hello!".
"""
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
`,
		`
FROM foo
ADAPTER adapter1
LICENSE """
Very long and boring legal text.
Blah blah blah.
"Oh look, a quote!"
"""

PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
MESSAGE system """
You are a store greeter. Always responsed with "Hello!".
"""
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
496
497
498
499
`,
		`
FROM foo
SYSTEM ""
Michael Yang's avatar
Michael Yang committed
500
501
502
503
504
`,
	}

	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
505
			modelfile, err := ParseFile(strings.NewReader(c))
Michael Yang's avatar
lint  
Michael Yang committed
506
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
507

Michael Yang's avatar
Michael Yang committed
508
			modelfile2, err := ParseFile(strings.NewReader(modelfile.String()))
Michael Yang's avatar
lint  
Michael Yang committed
509
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
510

Michael Yang's avatar
Michael Yang committed
511
			assert.Equal(t, modelfile, modelfile2)
Michael Yang's avatar
Michael Yang committed
512
513
514
		})
	}
}
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529

func TestParseFileUTF16ParseFile(t *testing.T) {
	data := `FROM bob
PARAMETER param1 1
PARAMETER param2 4096
SYSTEM You are a utf16 file.
`

	expected := []Command{
		{Name: "model", Args: "bob"},
		{Name: "param1", Args: "1"},
		{Name: "param2", Args: "4096"},
		{Name: "system", Args: "You are a utf16 file."},
	}

530
531
532
533
	t.Run("le", func(t *testing.T) {
		var b bytes.Buffer
		require.NoError(t, binary.Write(&b, binary.LittleEndian, []byte{0xff, 0xfe}))
		require.NoError(t, binary.Write(&b, binary.LittleEndian, utf16.Encode([]rune(data))))
534

535
536
		actual, err := ParseFile(&b)
		require.NoError(t, err)
537

538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
		assert.Equal(t, expected, actual.Commands)
	})

	t.Run("be", func(t *testing.T) {
		var b bytes.Buffer
		require.NoError(t, binary.Write(&b, binary.BigEndian, []byte{0xfe, 0xff}))
		require.NoError(t, binary.Write(&b, binary.BigEndian, utf16.Encode([]rune(data))))

		actual, err := ParseFile(&b)
		require.NoError(t, err)
		assert.Equal(t, expected, actual.Commands)
	})
}

func TestParseMultiByte(t *testing.T) {
	input := `FROM test
	SYSTEM 你好👋`

	expect := []Command{
		{Name: "model", Args: "test"},
		{Name: "system", Args: "你好👋"},
	}

	encodings := []encoding.Encoding{
		unicode.UTF8,
		unicode.UTF16(unicode.LittleEndian, unicode.UseBOM),
		unicode.UTF16(unicode.BigEndian, unicode.UseBOM),
	}

	for _, encoding := range encodings {
		t.Run(fmt.Sprintf("%s", encoding), func(t *testing.T) {
			s, err := encoding.NewEncoder().String(input)
			require.NoError(t, err)

			actual, err := ParseFile(strings.NewReader(s))
			require.NoError(t, err)

			assert.Equal(t, expect, actual.Commands)
		})
	}
578
}