-
Notifications
You must be signed in to change notification settings - Fork 45
/
generate_test_timestamps.py
302 lines (259 loc) · 16.4 KB
/
generate_test_timestamps.py
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
import datetime
import pytz
import sys
from collections import namedtuple
def __merge_dicts(*dict_args):
# Only needed for Python <3.5 support. In Python 3.5+, you can use the {**a, **b} syntax.
"""
From: https://stackoverflow.com/a/26853961
Given any number of dicts, shallow copy and merge into a new dict,
precedence goes to key value pairs in latter dicts.
"""
result = {}
for dictionary in dict_args:
result.update(dictionary)
return result
NumberField = namedtuple('NumberField', ['min_width', 'max_width', 'min_value', 'max_value'])
NUMBER_FIELDS = {
"year": NumberField(4, 4, 1, 9999),
"month": NumberField(2, 2, 1, 12),
"day": NumberField(2, 2, 1, 31),
"ordinal_day": NumberField(3, 3, 1, 365), # Intentionally missing leap year case
"iso_week": NumberField(2, 2, 1, 53),
"iso_day": NumberField(1, 1, 1, 7),
"hour": NumberField(2, 2, 0, 24), # 24 = special midnight value
"minute": NumberField(2, 2, 0, 59),
"second": NumberField(2, 2, 0, 60), # 60 = Leap second
"microsecond": NumberField(1, None, 0, None), # Can have unbounded characters
"tzhour": NumberField(2, 2, 0, 23),
"tzminute": NumberField(2, 2, 0, 59),
}
PADDED_NUMBER_FIELD_FORMATS = {
field_name: "{{{field_name}:0>{max_width}}}".format(
field_name=field_name,
max_width=field.max_width if field.max_width is not None else 1,
)
for field_name, field in NUMBER_FIELDS.items()
}
def __generate_valid_formats(year=2014, month=2, day=3, iso_week=6, iso_day=1, ordinal_day=34, hour=1, minute=23, second=45, microsecond=123456, tzhour=4, tzminute=30):
# Given a set of values, generates the 400+ different combinations of those values within a valid ISO 8601 string.
# Returns a Python format string, the fields in the format string, and the corresponding parameters you could pass to the datetime constructor
# These can be used by generate_valid_timestamp_and_datetime and generate_invalid_timestamp_and_datetime to produce test cases
valid_basic_calendar_date_formats = [
("{year}{month}{day}", set(["year", "month", "day"]), {"year": year, "month": month, "day": day})
]
valid_extended_calendar_date_formats = [
("{year}-{month}", set(["year", "month"]), {"year": year, "month": month, "day": 1}),
("{year}-{month}-{day}", set(["year", "month", "day"]), {"year": year, "month": month, "day": day}),
]
valid_basic_week_date_formats = [
("{year}W{iso_week}", set(["year", "iso_week"]), {"year": year, "iso_week": iso_week, "iso_day": 1}),
("{year}W{iso_week}{iso_day}", set(["year", "iso_week", "iso_day"]), {"year": year, "iso_week": iso_week, "iso_day": iso_day})
]
valid_extended_week_date_formats = [
("{year}-W{iso_week}", set(["year", "iso_week"]), {"year": year, "iso_week": iso_week, "iso_day": 1}),
("{year}-W{iso_week}-{iso_day}", set(["year", "iso_week", "iso_day"]), {"year": year, "iso_week": iso_week, "iso_day": iso_day})
]
valid_basic_ordinal_date_formats = [
("{year}{ordinal_day}", set(["year", "ordinal_day"]), {"year": year, "ordinal_day": ordinal_day}),
]
valid_extended_ordinal_date_formats = [
("{year}-{ordinal_day}", set(["year", "ordinal_day"]), {"year": year, "ordinal_day": ordinal_day}),
]
valid_date_and_time_separators = [None, "T", "t", " "]
valid_basic_time_formats = [
("{hour}", set(["hour"]), {"hour": hour}),
("{hour}{minute}", set(["hour", "minute"]), {"hour": hour, "minute": minute}),
("{hour}{minute}{second}", set(["hour", "minute", "second"]), {"hour": hour, "minute": minute, "second": second})
]
valid_extended_time_formats = [
("{hour}", set(["hour"]), {"hour": hour}),
("{hour}:{minute}", set(["hour", "minute"]), {"hour": hour, "minute": minute}),
("{hour}:{minute}:{second}", set(["hour", "minute", "second"]), {"hour": hour, "minute": minute, "second": second}),
]
valid_subseconds = [
("", set(), {}),
(".{microsecond}", set(["microsecond"]), {"microsecond": microsecond}), # TODO: Generate the trimmed 0's version?
(",{microsecond}", set(["microsecond"]), {"microsecond": microsecond}),
]
valid_tz_info_formats = [
("", set(), {}),
("Z", set(), {"tzinfo": pytz.UTC}),
("z", set(), {"tzinfo": pytz.UTC}),
("-{tzhour}", set(["tzhour"]), {"tzinfo": pytz.FixedOffset(-1 * tzhour * 60)}),
("+{tzhour}", set(["tzhour"]), {"tzinfo": pytz.FixedOffset(1 * tzhour * 60)}),
("-{tzhour}{tzminute}", set(["tzhour", "tzminute"]), {"tzinfo": pytz.FixedOffset(-1 * ((tzhour * 60) + tzminute))}),
("+{tzhour}{tzminute}", set(["tzhour", "tzminute"]), {"tzinfo": pytz.FixedOffset(1 * ((tzhour * 60) + tzminute))}),
("-{tzhour}:{tzminute}", set(["tzhour", "tzminute"]), {"tzinfo": pytz.FixedOffset(-1 * ((tzhour * 60) + tzminute))}),
("+{tzhour}:{tzminute}", set(["tzhour", "tzminute"]), {"tzinfo": pytz.FixedOffset(1 * ((tzhour * 60) + tzminute))})
]
format_date_time_combinations = [
(valid_basic_calendar_date_formats, valid_basic_time_formats),
(valid_extended_calendar_date_formats, valid_extended_time_formats),
(valid_basic_ordinal_date_formats, valid_basic_time_formats),
(valid_extended_ordinal_date_formats, valid_extended_time_formats),
]
if (sys.version_info.major, sys.version_info.minor) >= (3, 8):
# We rely on datetime.datetime.fromisocalendar
# to generate the expected values, but that was added in Python 3.8
format_date_time_combinations += [
(valid_basic_week_date_formats, valid_basic_time_formats),
(valid_extended_week_date_formats, valid_extended_time_formats)
]
for valid_calendar_date_formats, valid_time_formats in format_date_time_combinations:
for calendar_format, calendar_fields, calendar_params in valid_calendar_date_formats:
if "iso_week" in calendar_fields:
dt = datetime.datetime.fromisocalendar(calendar_params["year"], calendar_params["iso_week"], calendar_params["iso_day"])
calendar_params = __merge_dicts(calendar_params, { "month": dt.month, "day": dt.day })
del(calendar_params["iso_week"])
del(calendar_params["iso_day"])
if "ordinal_day" in calendar_fields:
dt = datetime.datetime(calendar_params["year"], 1, 1) + (datetime.timedelta(days=(calendar_params["ordinal_day"] - 1)))
calendar_params = __merge_dicts(calendar_params, { "month": dt.month, "day": dt.day })
del(calendar_params["ordinal_day"])
for date_and_time_separator in valid_date_and_time_separators:
if date_and_time_separator is None:
full_format = calendar_format
datetime_params = calendar_params
yield (full_format, calendar_fields, datetime_params)
else:
for time_format, time_fields, time_params in valid_time_formats:
for subsecond_format, subsecond_fields, subsecond_params in valid_subseconds:
for tz_info_format, tz_info_fields, tz_info_params in valid_tz_info_formats:
if "second" in time_fields:
# Add subsecond
full_format = calendar_format + date_and_time_separator + time_format + subsecond_format + tz_info_format
fields = set().union(calendar_fields, time_fields, subsecond_fields, tz_info_fields)
datetime_params = __merge_dicts(calendar_params, time_params, subsecond_params, tz_info_params)
elif subsecond_format == "": # Arbitrary choice of subsecond format. We don't want duplicates, so we only yield for one of them.
full_format = calendar_format + date_and_time_separator + time_format + tz_info_format
fields = set().union(calendar_fields, time_fields, tz_info_fields)
datetime_params = __merge_dicts(calendar_params, time_params, tz_info_params)
else:
# Ignore other subsecond formats
continue
yield (full_format, fields, datetime_params)
def __pad_params(**kwargs):
# Pads parameters to the required field widths.
return {key: PADDED_NUMBER_FIELD_FORMATS[key].format(**{key: value}) if key in PADDED_NUMBER_FIELD_FORMATS else value for key, value in kwargs.items()}
def generate_valid_timestamp_and_datetime(year=2014, month=2, day=3, iso_week=6, iso_day=1, ordinal_day=34, hour=1, minute=23, second=45, microsecond=123456, tzhour=4, tzminute=30):
# Given a set of values, generates the 400+ different combinations of those values within a valid ISO 8601 string, and the corresponding datetime
# This can be used to generate test cases of valid ISO 8601 timestamps.
# Note that this will produce many test cases that exercise the exact same code pathways (i.e., offer no additional coverage).
# Given a knowledge of the code, this is excessive, but these serve as a good set of black box tests (i.e., You could apply these to any ISO 8601 parse).
kwargs = {
"year": year,
"month": month,
"day": day,
"iso_week": iso_week,
"iso_day": iso_day,
"ordinal_day": ordinal_day,
"hour": hour,
"minute": minute,
"second": second,
"microsecond": microsecond,
"tzhour": tzhour,
"tzminute": tzminute,
}
for timestamp_format, _fields, datetime_params in __generate_valid_formats(**kwargs):
# Pad each field to the appropriate width
padded_kwargs = __pad_params(**kwargs)
timestamp = timestamp_format.format(**padded_kwargs)
yield (timestamp, datetime.datetime(**datetime_params))
def generate_invalid_timestamp(year=2014, month=2, day=3, iso_week=6, iso_day=1, ordinal_day=34, hour=1, minute=23, second=45, microsecond=123456, tzhour=4, tzminute=30):
# At the very least, each field can be invalid in the following ways:
# - Have too few characters
# - Have too many characters
# - Contain invalid characters
# - Have a value that is too small
# - Have a value that is too large
#
# This function takes each valid format (from `__generate_valid_formats()`), and mangles each field within the format to be invalid in each of the above ways.
# It also tests the case of trailing characters after each format.
# Note that this will produce many test cases that exercise the exact same code pathways (i.e., offer no additional coverage).
# Given a knowledge of the code, this is excessive, but these serve as a good set of black box tests (i.e., You could apply these to any ISO 8601 parse).
# This does not produce every invalid timestamp format though. For simplicity of the code, it does not cover the cases of:
# - The fields having 0 characters (Many fields (like day, minute, second etc.) are optional. So unless the field follows a separator, it is valid to have 0 characters)
# - Invalid day numbers for a given month (ex. "2014-02-31")
# - Invalid separators (ex. "2014=04=01")
# - Ordinal dates in leap years
# - Missing/Mismatched separators (ex. "2014-0101T0000:00")
# - Hour = 24, but not Special midnight case (ex. "24:00:01")
# - Timestamps that bear no resemblance to ISO 8601
# These cases will need to be test separately
kwargs = {
"year": year,
"month": month,
"day": day,
"iso_week": iso_week,
"iso_day": iso_day,
"ordinal_day": ordinal_day,
"hour": hour,
"minute": minute,
"second": second,
"microsecond": microsecond,
"tzhour": tzhour,
"tzminute": tzminute,
}
for timestamp_format, fields, _datetime_params in __generate_valid_formats(**kwargs):
for field_name in fields:
mangled_kwargs = __pad_params(**kwargs)
field = NUMBER_FIELDS.get(field_name, None)
if field is not None:
# Too few characters
for length in range(1, field.min_width):
if timestamp_format.startswith("{year}W{iso_week}{iso_day}") and field_name == "iso_week":
# If you reduce the iso_week field to 1 character, then the iso_day will make it into
# a valid "{year}W{iso_week}" timestamp
continue
if timestamp_format.startswith("{year}{month}{day}") and (field_name == "month" or field_name == "day"):
# If you reduce the month or day field to 1 character, then it will make it into
# a valid "{year}{ordinal_day}" timestamp
continue
if timestamp_format.startswith("{year}{month}{day}") and field_name == "year" and length == 3:
# If you reduce the year field to 3 characters, then it will make it into
# a valid "{year}{ordinal_day}" timestamp
continue
if timestamp_format.startswith("{year}-{ordinal_day}") and field_name == "ordinal_day" and length == 2:
# If you reduce the ordinal_day field to 2 characters, then it will make it into
# a valid "{year}-{month}" timestamp
continue
str_value = str(__pad_params(**{field_name: kwargs[field_name]})[field_name])[0:length]
mangled_kwargs[field_name] = "{{:0>{length}}}".format(length=length).format(str_value)
timestamp = timestamp_format.format(**mangled_kwargs)
yield (timestamp, "{0} has too few characters".format(field_name))
# Too many characters
if field.max_width is not None:
if timestamp_format.startswith("{year}-{month}") and field_name == "month":
# If you extend the month field to 3 characters, then it will make it into
# a valid "{year}{ordinal_day}" timestamp
continue
mangled_kwargs[field_name] = "{{:0>{length}}}".format(length=field.max_width + 1).format(kwargs[field_name])
timestamp = timestamp_format.format(**mangled_kwargs)
yield (timestamp, "{0} has too many characters".format(field_name))
# Too small of value
if (field.min_value - 1) >= 0:
mangled_kwargs[field_name] = __pad_params(**{field_name: field.min_value - 1})[field_name]
timestamp = timestamp_format.format(**mangled_kwargs)
yield (timestamp, "{0} has too small value".format(field_name))
# Too large of value
if field.max_value is not None:
mangled_kwargs[field_name] = __pad_params(**{field_name: field.max_value + 1})[field_name]
timestamp = timestamp_format.format(**mangled_kwargs)
yield (timestamp, "{0} has too large value".format(field_name))
# Invalid characters
max_invalid_characters = field.max_width if field.max_width is not None else 1
# ex. 2014 -> a, aa, aaa
for length in range(1, max_invalid_characters):
mangled_kwargs[field_name] = "a" * length
timestamp = timestamp_format.format(**mangled_kwargs)
yield (timestamp, "{0} has invalid characters".format(field_name))
# ex. 2014 -> aaaa, 2aaa, 20aa, 201a
for length in range(0, max_invalid_characters):
str_value = str(__pad_params(**{field_name: kwargs[field_name]})[field_name])[0:length]
mangled_kwargs[field_name] = "{{:a<{length}}}".format(length=max_invalid_characters).format(str_value)
timestamp = timestamp_format.format(**mangled_kwargs)
yield (timestamp, "{0} has invalid characters".format(field_name))
# Trailing characters
timestamp = timestamp_format.format(**__pad_params(**kwargs)) + "EXTRA"
yield (timestamp, "{0} has extra characters".format(field_name))