parser.py 17.7 KB
Newer Older
1
#!/usr/bin/env python3
2
3
#SCT log parser

4

5
import sys
Vincent Stehlé's avatar
Vincent Stehlé committed
6
import argparse
Vincent Stehlé's avatar
Vincent Stehlé committed
7
import csv
Vincent Stehlé's avatar
Vincent Stehlé committed
8
import logging
Vincent Stehlé's avatar
Vincent Stehlé committed
9
import json
Vincent Stehlé's avatar
Vincent Stehlé committed
10

Vincent Stehlé's avatar
Vincent Stehlé committed
11
12
13
14
15
try:
    import yaml
except ImportError:
    print('No yaml...')

16

17
#based loosley on https://stackoverflow.com/a/4391978
Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
18
# returns a filtered dict of dicts that meet some Key-value pair.
19
# I.E. key="result" value="FAILURE"
20
def key_value_find(list_1, key, value):
21
    found = list()
22
    for test in list_1:
23
        if test[key] == value:
24
            found.append(test)
25
    return found
26

27
28

#Were we intrept test logs into test dicts
29
def test_parser(string, current):
30
    test_list = {
Vincent Stehlé's avatar
Vincent Stehlé committed
31
      "name": string[2], #FIXME:Sometimes, SCT has name and Description,
32
      "result": string[1],
33
      **current,
34
      "guid": string[0], #FIXME:GUID's overlap
Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
35
      #"comment": string[-1], #FIXME:need to hash this out, sometime there is no comments
36
      "log": ' '.join(string[3:])
37
    }
38
    return test_list
Vincent Stehlé's avatar
Vincent Stehlé committed
39

40
#Parse the ekl file, and create a map of the tests
41
42
def ekl_parser (file):
    #create our "database" dict
43
    temp_list = list()
44
45
46
47
48
49
50
51
    # All tests are grouped by the "HEAD" line, which precedes them.
    current = {}

    # Count number of tests since beginning of the set
    n = 0

    # Number of skipped tests sets
    s = 0
52

53
    for i, line in enumerate(file):
Vincent Stehlé's avatar
Vincent Stehlé committed
54
55
56
57
58
59
60
        # Strip the line from trailing whitespaces
        line = line.rstrip()

        # Skip empty line
        if line == '':
            continue

61
62
        # strip the line of | & || used for sepration
        split_line = line.split('|')
63

64
65
66
67
        # TERM marks the end of a test set
        # In case of empty test set we generate an artificial skipped test
        # entry. Then reset current as a precaution, as well as our test
        # counter.
68
        if split_line[0] == '' and split_line[1] == "TERM":
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
            if not n:
                logging.debug(f"Skipped test set `{current['sub set']}'")

                temp_list.append({
                    **current,
                    'name': '',
                    'guid': '',
                    'log': '',
                    'result': 'SKIPPED',
                })

                s += 1

            current = {}
            n = 0
84
85
            continue

86
87
88
        # The "HEAD" tag is the only indcation we are on a new test set
        if split_line[0] == '' and split_line[1] == "HEAD":
            # split the header into test group and test set.
89
            try:
90
91
92
                group, Set = split_line[12].split('\\')
            except Exception:
                group, Set = '', split_line[12]
93
94
95
            current = {
                'group': group,
                'test set': Set,
96
97
98
99
100
101
102
103
                'sub set': split_line[10],
                'set guid': split_line[8],
                'iteration': split_line[4],
                'start date': split_line[6],
                'start time': split_line[7],
                'revision': split_line[9],
                'descr': split_line[11],
                'device path': '|'.join(split_line[13:]),
104
            }
105

Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
106
        #FIXME:? EKL file has an inconsistent line structure,
107
108
        # sometime we see a line that consits ' dump of GOP->I\n'
        #easiest way to skip is check for blank space in the first char
109
        elif split_line[0] != '' and split_line[0][0] != " ":
110
111
112
113
            try:
                #deliminiate on ':' for tests
                split_test = [new_string for old_string in split_line for new_string in old_string.split(':')]
                #put the test into a dict, and then place that dict in another dict with GUID as key
114
                tmp_dict = test_parser(split_test, current)
115
                temp_list.append(tmp_dict)
116
                n += 1
117
118
119
            except:
                print("Line:",split_line)
                sys.exit("your log may be corrupted")
120
121
122
        else:
            logging.error(f"Unparsed line {i} `{line}'")

123
124
125
    if s:
        logging.debug(f'{s} skipped test set(s)')

126
    return temp_list
127

Jeff Booher-Kaeding's avatar
Jeff Booher-Kaeding committed
128
#Parse Seq file, used to tell which tests should run.
129
def seq_parser(file):
130
    temp_dict = list()
131
    lines=file.readlines()
Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
132
    magic=7 #a test in a seq file is 7 lines, if not mod7, something wrong..
133
134
    if len(lines)%magic != 0:
        sys.exit("seqfile cut short, should be mod7")
135
136
137
138
139
140
141
142
    #the utf-16 char makes this looping a bit harder, so we use x+(i) where i is next 0-6th
    for x in range(0,len(lines),magic): #loop ever "7 lines"
        #(x+0)[Test Case]
        #(x+1)Revision=0x10000
        #(x+2)Guid=XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX
        #(x+3)Name=InstallAcpiTableFunction
        #(x+4)Order=0xFFFFFFFF
        #(x+5)Iterations=0xFFFFFFFF
Vincent Stehlé's avatar
Vincent Stehlé committed
143
        #(x+6)(utf-16 char)
144
        #currently only add tests that are supposed to run, should add all?
Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
145
        #0xFFFFFFFF in "Iterations" means the test is NOT supposed to run
146
147
148
149
150
151
152
153
        if not "0xFFFFFFFF" in lines[x+5]:
            seq_dict = {
                "name": lines[x+3][5:-1],#from after "Name=" to end (5char long)
                "guid": lines[x+2][5:-1],#from after"Guid=" to the end, (5char long)
                "Iteration": lines[x+5][11:-1],#from after "Iterations=" (11char long)
                "rev": lines[x+1][9:-1],#from after "Revision=" (9char long)
                "Order": lines[x+4][6:-1]#from after "Order=" (6char long)
            }
154
            temp_dict.append(seq_dict) #put in a dict based on guid
155

156
    return temp_dict
157

158
159
160
161
#group items by key, and print by key
#we slowly iterate through the list, group and print groups
def key_tree_2_md(input_list,file,key):
    #make a copy so we don't destroy the first list.
Vincent Stehlé's avatar
Vincent Stehlé committed
162
    temp_list = input_list.copy()
163
164
165
166
167
168
169
170
171
172
173
174
    while temp_list:
        test_dict = temp_list.pop()
        found, not_found = [test_dict],[]
        #go through whole list looking for key match
        while temp_list:
            next_dict = temp_list.pop()
            if next_dict[key] == test_dict[key]: #if match add to found
                found.append(next_dict)
            else: #else not found
                not_found.append(next_dict)
        temp_list = not_found #start over with found items removed
        file.write("### " + test_dict[key])
175
        dict_2_md(found,file)
Vincent Stehlé's avatar
Vincent Stehlé committed
176

177
178


179
180
181
#generic writer, takes a list of dicts and turns the dicts into an MD table.
def dict_2_md(input_list,file):
    if len(input_list) > 0:
Jeff Booher-Kaeding's avatar
Jeff Booher-Kaeding committed
182
        file.write("\n\n")
183
184
185
186
        #create header for MD table using dict keys
        temp_string1, temp_string2 = "|", "|"
        for x in (input_list[0].keys()):
            temp_string1 += (x + "|")
Jeff Booher-Kaeding's avatar
Jeff Booher-Kaeding committed
187
            temp_string2 += ("---|")
188
189
190
191
192
193
194
        file.write(temp_string1+"\n"+temp_string2+"\n")
        #print each item from the dict into the table
        for x in input_list:
            test_string = "|"
            for y in x.keys():
                test_string += (x[y] + "|")
            file.write(test_string+'\n')
Vincent Stehlé's avatar
Vincent Stehlé committed
195
    #seprate table from other items in MD
196
    file.write("\n\n")
197

Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
198

Vincent Stehlé's avatar
Vincent Stehlé committed
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
# Sanitize our YAML configuration
# We modify conf in-place
# TODO: use a proper validator instead
def sanitize_yaml(conf):
    rules = set()

    for i in range(len(conf)):
        r = conf[i]

        # Generate a rule name if needed
        if 'rule' not in r:
            r['rule'] = f'r{i}'
            logging.debug(f"Auto-naming rule {i} `{r['rule']}'")
            conf[i] = r

        if r['rule'] in rules:
            logging.warning(f"Duplicate rule {i} `{r['rule']}'")

        rules.add(r['rule'])

        if 'criteria' not in r or not type(r['criteria']) is dict or \
           'update' not in r or not type(r['update']) is dict:
            logging.error(f"Bad rule {i} `{r}'")
            raise Exception()


# Evaluate if a test dict matches a criteria
# The criteria is a dict of Key-value pairs.
# I.E. crit = {"result": "FAILURE", "xxx": "yyy", ...}
# All key-value pairs must be present and match for a test dict to match.
# A test value and a criteria value match if the criteria value string is
# present anywhere in the test value string.
# For example, the test value "abcde" matches the criteria value "cd".
# This allows for more "relaxed" criteria than strict comparison.
def matches_crit(test, crit):
    for key, value in crit.items():
        if key not in test or test[key].find(value) < 0:
            return False

    return True


# Apply all configuration rules to the tests
# We modify cross_check in-place
def apply_rules(cross_check, conf):
    # Prepare statistics counters
    stats = {}

    for r in conf:
        stats[r['rule']] = 0

    # Apply rules on each test data
    s = len(cross_check)

    for i in range(s):
        test = cross_check[i]

        for r in conf:
            if not matches_crit(test, r['criteria']):
                continue

            rule = r['rule']

            logging.debug(
                f"Applying rule `{rule}'"
                f" to test {i} `{test['name']}'")

            test.update({
                **r['update'],
                'Updated by': rule,
            })

            stats[rule] += 1
            break

    # Statistics
    n = 0

    for rule, cnt in stats.items():
        logging.debug(f"{cnt} matche(s) for rule `{rule}'")
        n += cnt

    if n:
        r = len(conf)
        logging.info(
            f'Updated {n} test(s) out of {s} after applying {r} rule(s)')


# Use YAML configuration file and perform all the transformations described in
# there.
# See the README.md for details on the file format.
# We modify cross_check in-place
def use_config(cross_check, filename):
    assert('yaml' in sys.modules)

    # Load configuration file
    logging.debug(f'Read {filename}')

    with open(filename, 'r') as yamlfile:
        conf = yaml.load(yamlfile, Loader=yaml.FullLoader)

    logging.debug('{} rule(s)'.format(len(conf)))
    sanitize_yaml(conf)
    apply_rules(cross_check, conf)

Vincent Stehlé's avatar
Vincent Stehlé committed
304
305
306
307
308
309
310
311
312
313
# Sort tests data in-place
# sort_keys is a comma-separated list
# The first key has precedence, then the second, etc.
# To use python list in-place sorting, we use the keys in reverse order.
def sort_data(cross_check, sort_keys):
    logging.debug(f"Sorting on `{sort_keys}'")
    for k in reversed(sort_keys.split(',')):
        cross_check.sort(key=lambda x: x[k])


Vincent Stehlé's avatar
Vincent Stehlé committed
314
315
316
317
318
319
320
321
322
# Generate csv
def gen_csv(cross_check, filename):
    # Find keys
    keys = set()

    for x in cross_check:
        keys = keys.union(x.keys())

    # Write csv
Vincent Stehlé's avatar
Vincent Stehlé committed
323
324
    logging.debug(f'Generate {filename}')

Vincent Stehlé's avatar
Vincent Stehlé committed
325
326
327
328
329
330
331
    with open(filename, 'w', newline='') as csvfile:
        writer = csv.DictWriter(
            csvfile, fieldnames=sorted(keys), delimiter=';')
        writer.writeheader()
        writer.writerows(cross_check)


Vincent Stehlé's avatar
Vincent Stehlé committed
332
333
334
335
336
337
338
339
# Generate json
def gen_json(cross_check, filename):
    logging.debug(f'Generate {filename}')

    with open(filename, 'w') as jsonfile:
        json.dump(cross_check, jsonfile, sort_keys=True, indent=2)


340
341
342
343
344
345
346
347
348
# Generate yaml
def gen_yaml(cross_check, filename):
    assert('yaml' in sys.modules)
    logging.debug(f'Generate {filename}')

    with open(filename, 'w') as yamlfile:
        yaml.dump(cross_check, yamlfile)


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
# Combine or two databases db1 and db2 coming from ekl and seq files
# respectively into a single cross_check database
# Tests in db1, which were not meant to be run according to db2 have their
# results forced to SPURIOUS.
# Tests sets in db2, which were not run according to db1 have an artificial
# test entry created with result DROPPED.
def combine_dbs(db1, db2):
    cross_check = db1

    # Do a pass to verify that all tests in db1 were meant to be run.
    # Otherwise, force the result to SPURIOUS.
    s = set()

    for x in db2:
        s.add(x['guid'])

    n = 0

    for i in range(len(cross_check)):
        if cross_check[i]['set guid'] not in s:
            logging.debug(f"Spurious test {i} `{cross_check[i]['name']}'")
            cross_check[i]['result'] = 'SPURIOUS'
            n += 1

    if n:
        logging.debug(f'{n} spurious test(s)')

    # Do a pass to autodetect all tests fields in case we need to merge dropped
    # tests sets entries
    f = {}

    for x in cross_check:
        for k in x.keys():
            f[k] = ''

    logging.debug(f'Test fields: {f.keys()}')

    # Do a pass to find the test sets that did not run for whatever reason.
    s = set()

    for x in cross_check:
        s.add(x['set guid'])

    n = 0

    for i in range(len(db2)):
        x = db2[i]

        if not x['guid'] in s:
            logging.debug(f"Dropped test set {i} `{x['name']}'")

            # Create an artificial test entry to reflect the dropped test set
            cross_check.append({
                **f,
                'sub set': x['name'],
                'set guid': x['guid'],
                'revision': x['rev'],
                'group': 'Unknown',
                'result': 'DROPPED',
            })

            n += 1

    if n:
        logging.debug(f'{n} dropped test set(s)')

    return cross_check


418
def main():
Vincent Stehlé's avatar
Vincent Stehlé committed
419
420
421
422
    parser = argparse.ArgumentParser(
        description='Process SCT results.'
                    ' This program takes the SCT summary and sequence files,'
                    ' and generates a nice report in mardown format.',
Vincent Stehlé's avatar
Vincent Stehlé committed
423
424
425
426
427
        epilog='When sorting is requested, tests data are sorted'
               ' according to the first sort key, then the second, etc.'
               ' Sorting happens after update by the configuration rules.'
               ' Useful example: --sort'
               ' "group,descr,set guid,test set,sub set,guid,name,log"',
Vincent Stehlé's avatar
Vincent Stehlé committed
428
        formatter_class=argparse.ArgumentDefaultsHelpFormatter)
Vincent Stehlé's avatar
Vincent Stehlé committed
429
    parser.add_argument('--csv', help='Output .csv filename')
Vincent Stehlé's avatar
Vincent Stehlé committed
430
    parser.add_argument('--json', help='Output .json filename')
Vincent Stehlé's avatar
Vincent Stehlé committed
431
432
    parser.add_argument(
        '--md', help='Output .md filename', default='result.md')
Vincent Stehlé's avatar
Vincent Stehlé committed
433
434
    parser.add_argument(
        '--debug', action='store_true', help='Turn on debug messages')
Vincent Stehlé's avatar
Vincent Stehlé committed
435
436
    parser.add_argument(
        '--sort', help='Comma-separated list of keys to sort output on')
Vincent Stehlé's avatar
Vincent Stehlé committed
437
438
439
440
441
442
443
444
    parser.add_argument(
        'log_file', nargs='?', default='sample.ekl',
        help='Input .ekl filename')
    parser.add_argument(
        'seq_file', nargs='?', default='sample.seq',
        help='Input .seq filename')
    parser.add_argument('find_key', nargs='?', help='Search key')
    parser.add_argument('find_value', nargs='?', help='Search value')
Vincent Stehlé's avatar
Vincent Stehlé committed
445
446
447
448
449
450

    # A few command line switches depend on yaml. We enable those only if we
    # could actually import yaml.
    if 'yaml' in sys.modules:
        parser.add_argument(
            '--config', help='Input .yaml configuration filename')
451
        parser.add_argument('--yaml', help='Output .yaml filename')
Vincent Stehlé's avatar
Vincent Stehlé committed
452

Vincent Stehlé's avatar
Vincent Stehlé committed
453
454
    args = parser.parse_args()

Vincent Stehlé's avatar
Vincent Stehlé committed
455
456
457
458
    logging.basicConfig(
        format='%(levelname)s %(funcName)s: %(message)s',
        level=logging.DEBUG if args.debug else logging.INFO)

Vincent Stehlé's avatar
Vincent Stehlé committed
459
    #Command line argument 1, ekl file to open
460
    db1 = list() #"database 1" all tests.
Vincent Stehlé's avatar
Vincent Stehlé committed
461
462
    logging.debug(f'Read {args.log_file}')

Vincent Stehlé's avatar
Vincent Stehlé committed
463
    with open(args.log_file,"r",encoding="utf-16") as f: #files are encoded in utf-16
Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
464
465
        db1 = ekl_parser(f.readlines())

Vincent Stehlé's avatar
Vincent Stehlé committed
466
467
    logging.debug('{} test(s)'.format(len(db1)))

Vincent Stehlé's avatar
Vincent Stehlé committed
468
    #Command line argument 2, seq file to open
469
    db2 = dict() #"database 2" all test sets that should run
Vincent Stehlé's avatar
Vincent Stehlé committed
470
471
    logging.debug(f'Read {args.seq_file}')

Vincent Stehlé's avatar
Vincent Stehlé committed
472
    with open(args.seq_file,"r",encoding="utf-16") as f: #files are encoded in utf-16
473
        db2 = seq_parser(f)
Vincent Stehlé's avatar
Vincent Stehlé committed
474

Vincent Stehlé's avatar
Vincent Stehlé committed
475
476
    logging.debug('{} test set(s)'.format(len(db2)))

477
478
    # Produce a single cross_check database from our two db1 and db2 databases.
    cross_check = combine_dbs(db1, db2)
479

Vincent Stehlé's avatar
Vincent Stehlé committed
480
481
482
483
    # Take configuration file into account. This can perform transformations on
    # the tests results.
    if 'config' in args and args.config is not None:
        use_config(cross_check, args.config)
Vincent Stehlé's avatar
Vincent Stehlé committed
484

Vincent Stehlé's avatar
Vincent Stehlé committed
485
486
487
488
    # Sort tests data in-place, if requested
    if args.sort is not None:
        sort_data(cross_check, args.sort)

489
490
491
    # search for failures, warnings, passes & others
    # We detect all present keys in additions to the expected ones. This is
    # handy with config rules overriding the result field with arbitrary values.
492
    res_keys = set(['DROPPED', 'FAILURE', 'WARNING', 'PASS'])
Vincent Stehlé's avatar
Vincent Stehlé committed
493

494
495
496
497
498
499
500
501
502
503
    for x in cross_check:
        res_keys.add(x['result'])

    # Now we fill some bins with tests according to their result
    bins = {}

    for k in res_keys:
        bins[k] = key_value_find(cross_check, "result", k)

    # Print a one-line summary
504
    s = map(
505
506
507
508
        lambda k: '{} {}(s)'.format(len(bins[k]), k.lower()),
        sorted(res_keys))

    logging.info(', '.join(s))
509
510

    # generate MD summary
Vincent Stehlé's avatar
Vincent Stehlé committed
511
512
    logging.debug(f'Generate {args.md}')

Vincent Stehlé's avatar
Vincent Stehlé committed
513
    with open(args.md, 'w') as resultfile:
514
515
516
        resultfile.write("# SCT Summary \n\n")
        resultfile.write("|  |  |\n")
        resultfile.write("|--|--|\n")
517
518
519
520
521
522

        # Loop on all the result values we found for the summary
        for k in sorted(res_keys):
            resultfile.write(
                "|{}:|{}|\n".format(k.title(), len(bins[k])))

523
524
        resultfile.write("\n\n")

525
526
        # Loop on all the result values we found (except PASS) for the sections
        # listing the tests by group
527
        n = 1
528
529
        res_keys_np = set(res_keys)
        res_keys_np.remove('PASS')
Vincent Stehlé's avatar
Vincent Stehlé committed
530

531
532
533
534
        for k in sorted(res_keys_np):
            resultfile.write("## {}. {} by group\n\n".format(n, k.title()))
            key_tree_2_md(bins[k], resultfile, "group")
            n += 1
535

Vincent Stehlé's avatar
Vincent Stehlé committed
536
537
538
    # Generate csv if requested
    if args.csv is not None:
        gen_csv(cross_check, args.csv)
Vincent Stehlé's avatar
Vincent Stehlé committed
539

Vincent Stehlé's avatar
Vincent Stehlé committed
540
541
542
543
    # Generate json if requested
    if args.json is not None:
        gen_json(cross_check, args.json)

544
545
546
547
    # Generate yaml if requested
    if 'yaml' in args and args.yaml is not None:
        gen_yaml(cross_check, args.yaml)

Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
548
549
    #command line argument 3&4, key are to support a key & value search.
    #these will be displayed in CLI
Vincent Stehlé's avatar
Vincent Stehlé committed
550
551
    if args.find_key is not None and args.find_value is not None:
        found = key_value_find(db1, args.find_key, args.find_value)
Jeff Booher-Kaeding's avatar
V1.0?    
Jeff Booher-Kaeding committed
552
553
554
        #print the dict
        print("found:",len(found),"items with search constraints")
        for x in found:
Vincent Stehlé's avatar
Vincent Stehlé committed
555
556
557
558
            print(
                x["guid"], ":", x["name"], "with", args.find_key, ":",
                x[args.find_key])

559

560
main()