1
|
|
|
""" |
2
|
|
|
Support classes to help with querying and processing web data. |
3
|
|
|
""" |
4
|
|
|
from __future__ import annotations |
5
|
|
|
|
6
|
|
|
import logging |
7
|
|
|
import re |
8
|
|
|
from dataclasses import dataclass |
9
|
|
|
from typing import ( |
10
|
|
|
Any, |
11
|
|
|
Callable, |
12
|
|
|
Sequence, |
13
|
|
|
Set, |
14
|
|
|
List, |
15
|
|
|
Union, |
16
|
|
|
Optional, |
17
|
|
|
Type, |
18
|
|
|
TypeVar, |
19
|
|
|
Iterable, |
20
|
|
|
FrozenSet, |
21
|
|
|
) |
22
|
|
|
|
23
|
|
|
import numpy as np |
|
|
|
|
24
|
|
|
from pocketutils.tools.base_tools import BaseTools |
|
|
|
|
25
|
|
|
from pocketutils.tools.string_tools import StringTools |
|
|
|
|
26
|
|
|
from pocketutils.core.dot_dict import NestedDotDict |
|
|
|
|
27
|
|
|
|
28
|
|
|
logger = logging.getLogger("mandos") |
29
|
|
|
|
30
|
|
|
empty_frozenset = frozenset([]) |
31
|
|
|
|
32
|
|
|
|
33
|
|
|
@dataclass(frozen=True) |
|
|
|
|
34
|
|
|
class FilterFn: |
35
|
|
|
keep_if: Callable[[Any], bool] |
36
|
|
|
|
37
|
|
|
def __call__(self, *args, **kwargs): |
38
|
|
|
return self.keep_if(*args, **kwargs) |
39
|
|
|
|
40
|
|
|
|
41
|
|
|
class Fns: |
|
|
|
|
42
|
|
|
@classmethod |
43
|
|
|
def has_key(cls, key: str) -> FilterFn: |
|
|
|
|
44
|
|
|
return FilterFn(lambda content: key in content) |
45
|
|
|
|
46
|
|
|
@classmethod |
47
|
|
|
def key_does_not_equal(cls, key: str, disallowed_value: Any) -> FilterFn: |
|
|
|
|
48
|
|
|
return FilterFn(lambda content: content.get(key) != disallowed_value) |
49
|
|
|
|
50
|
|
|
@classmethod |
51
|
|
|
def key_equals(cls, key: str, allowed_value: Any) -> FilterFn: |
|
|
|
|
52
|
|
|
return FilterFn(lambda content: content.get(key) == allowed_value) |
53
|
|
|
|
54
|
|
|
@classmethod |
55
|
|
|
def key_is_not_in(cls, key: str, disallowed_values: Set[Any]) -> FilterFn: |
|
|
|
|
56
|
|
|
return FilterFn(lambda content: content.get(key) not in disallowed_values) |
57
|
|
|
|
58
|
|
|
@classmethod |
59
|
|
|
def key_is_in(cls, key: str, allowed_values: Set[Any]) -> FilterFn: |
|
|
|
|
60
|
|
|
return FilterFn(lambda content: content.get(key) in allowed_values) |
61
|
|
|
|
62
|
|
|
@classmethod |
63
|
|
|
def construct(cls, tp: Type[T]) -> Callable[[Iterable[Any]], T]: |
|
|
|
|
64
|
|
|
""" |
65
|
|
|
Function that constructs an instance whose attributes are from the passed list. |
66
|
|
|
""" |
67
|
|
|
|
68
|
|
|
def construct(things: Iterable[str]) -> Optional[str]: |
69
|
|
|
x = [s.strip() for s in things] |
|
|
|
|
70
|
|
|
return tp(*x) |
71
|
|
|
|
72
|
|
|
return construct |
73
|
|
|
|
74
|
|
|
@classmethod |
75
|
|
|
def join_nonnulls(cls) -> Callable[[Iterable[str]], Optional[str]]: |
|
|
|
|
76
|
|
|
def opt_join(things: Iterable[str]) -> Optional[str]: |
77
|
|
|
x = [s.strip() for s in things] |
|
|
|
|
78
|
|
|
return None if len(x) == 0 else ";".join(x) |
79
|
|
|
|
80
|
|
|
return opt_join |
81
|
|
|
|
82
|
|
|
@classmethod |
83
|
|
|
def split_and_flatten_nonnulls( |
|
|
|
|
84
|
|
|
cls, sep: str, skip_nulls: bool = False |
|
|
|
|
85
|
|
|
) -> Callable[[Iterable[Union[str, int, float, None]]], Set[str]]: |
86
|
|
|
def split_flat(things: Iterable[str]) -> Set[str]: |
87
|
|
|
results = set() |
88
|
|
|
for thing in things: |
89
|
|
|
if thing is not None and thing != float("NaN") or not skip_nulls: |
90
|
|
|
# let it fail if skip_nulls is False |
91
|
|
|
for bit in str(thing).split(sep): |
92
|
|
|
results.add(bit.strip()) |
93
|
|
|
return results |
94
|
|
|
|
95
|
|
|
return split_flat |
96
|
|
|
|
97
|
|
|
@classmethod |
98
|
|
|
def request_only(cls) -> Callable[[Iterable[str]], Optional[str]]: |
|
|
|
|
99
|
|
|
def only_nonreq(things: Iterable[str]) -> Optional[str]: |
100
|
|
|
things = [s.strip() for s in things] |
101
|
|
|
if len(things) > 1: |
|
|
|
|
102
|
|
|
raise ValueError(f"{len(things)} items in {things}") |
103
|
|
|
elif len(things) == 0: |
104
|
|
|
return None |
105
|
|
|
else: |
106
|
|
|
return things[0] |
107
|
|
|
|
108
|
|
|
return only_nonreq |
109
|
|
|
|
110
|
|
|
@classmethod |
111
|
|
|
def roman_to_arabic( |
|
|
|
|
112
|
|
|
cls, min_val: Optional[int] = None, max_val: Optional[int] = None |
|
|
|
|
113
|
|
|
) -> Callable[[str], int]: |
114
|
|
|
def roman_to_arabic(s: str) -> int: |
|
|
|
|
115
|
|
|
return StringTools.roman_to_arabic(s.strip(), min_val=min_val, max_val=max_val) |
116
|
|
|
|
117
|
|
|
return roman_to_arabic |
118
|
|
|
|
119
|
|
|
@classmethod |
120
|
|
|
def require_only(cls) -> Callable[[Iterable[str]], Optional[str]]: |
|
|
|
|
121
|
|
|
return BaseTools.only |
122
|
|
|
|
123
|
|
|
@classmethod |
124
|
|
|
def extract_group_1( |
|
|
|
|
125
|
|
|
cls, pattern: Union[str, re.Pattern] |
|
|
|
|
126
|
|
|
) -> Callable[[Optional[Any]], Optional[str]]: |
127
|
|
|
pattern = pattern if isinstance(pattern, re.Pattern) else re.compile(pattern) |
128
|
|
|
|
129
|
|
|
def _match(thing: Optional[str]) -> Optional[str]: |
130
|
|
|
if thing is None: |
131
|
|
|
return None |
132
|
|
|
match = pattern.fullmatch(str(thing)) |
133
|
|
|
if match is None: |
134
|
|
|
return None |
135
|
|
|
return match.group(1) |
136
|
|
|
|
137
|
|
|
return _match |
138
|
|
|
|
139
|
|
|
@classmethod |
140
|
|
|
def lowercase_unless_acronym(cls) -> Callable[[str], str]: |
|
|
|
|
141
|
|
|
def lowercase_unless_acronym(s: str) -> str: |
|
|
|
|
142
|
|
|
s = s.strip() |
143
|
|
|
return s if s.isupper() else s.lower() |
144
|
|
|
|
145
|
|
|
return lowercase_unless_acronym |
146
|
|
|
|
147
|
|
|
@classmethod |
148
|
|
|
def split_bars_to_int(cls, sep: str = "||") -> Callable[[Optional[str]], FrozenSet[int]]: |
|
|
|
|
149
|
|
|
def split_bars_to_int(value: str) -> FrozenSet[int]: |
150
|
|
|
return frozenset([int(x) for x in cls.split(sep)(value)]) |
151
|
|
|
|
152
|
|
|
return split_bars_to_int |
153
|
|
|
|
154
|
|
|
@classmethod |
155
|
|
|
def split(cls, sep: str = "||") -> Callable[[Optional[str]], FrozenSet[str]]: |
|
|
|
|
156
|
|
|
def split_bars(value: str) -> FrozenSet[str]: |
157
|
|
|
if value is None: |
158
|
|
|
return empty_frozenset |
159
|
|
|
return frozenset([s.strip() for s in value.split(sep)]) |
160
|
|
|
|
161
|
|
|
return split_bars |
162
|
|
|
|
163
|
|
|
@classmethod |
164
|
|
|
def n_bar_items(cls, sep: str = "||") -> Callable[[Optional[str]], int]: |
|
|
|
|
165
|
|
|
def n_bar_items(value: str) -> int: |
166
|
|
|
return len(cls.split(sep)(value)) |
167
|
|
|
|
168
|
|
|
return n_bar_items |
169
|
|
|
|
170
|
|
|
@classmethod |
171
|
|
|
def not_null(cls) -> Callable[[Any], bool]: |
|
|
|
|
172
|
|
|
return lambda value: value is not None |
173
|
|
|
|
174
|
|
|
@classmethod |
175
|
|
|
def identity(cls) -> Callable[[T], T]: |
|
|
|
|
176
|
|
|
return lambda value: value |
177
|
|
|
|
178
|
|
|
@classmethod |
179
|
|
|
def req_is(cls, type_, nullable: bool = False, then_convert=None) -> Callable[[str], str]: |
|
|
|
|
180
|
|
|
def req_is(value): |
181
|
|
|
if not isinstance(value, type_): |
182
|
|
|
raise ValueError(f"{value} is a {type(value)}, not {type_}") |
183
|
|
|
return value if then_convert is None else then_convert(value) |
184
|
|
|
|
185
|
|
|
req_is.__name__ = f"req_is_{type_}" + ("_or_null" if nullable else "") |
186
|
|
|
return req_is |
187
|
|
|
|
188
|
|
|
|
189
|
|
|
def _get_conversion_fn(fn: Union[None, str, Callable[[Any], Any]]) -> Callable[[Any], Any]: |
|
|
|
|
190
|
|
|
if fn is None: |
191
|
|
|
return Fns.identity |
192
|
|
|
if isinstance(fn, str): |
|
|
|
|
193
|
|
|
return Fns.request_only() |
194
|
|
|
else: |
195
|
|
|
return fn |
196
|
|
|
|
197
|
|
|
|
198
|
|
|
T = TypeVar("T", covariant=True) |
|
|
|
|
199
|
|
|
V = TypeVar("V", covariant=True) |
|
|
|
|
200
|
|
|
|
201
|
|
|
|
202
|
|
|
@dataclass(frozen=True, eq=True) |
|
|
|
|
203
|
|
|
class AbstractJsonNavigator: |
204
|
|
|
"""""" |
205
|
|
|
|
206
|
|
|
|
207
|
|
|
@dataclass(frozen=True, eq=True) |
|
|
|
|
208
|
|
|
class JsonNavigatorSingleOptional(AbstractJsonNavigator): |
209
|
|
|
contents: Optional[T] |
210
|
|
|
|
211
|
|
|
@property |
212
|
|
|
def get(self) -> T: |
|
|
|
|
213
|
|
|
return self.contents |
214
|
|
|
|
215
|
|
|
def __floordiv__( |
216
|
|
|
self, conversion: Union[Type[T], Callable[[T], V]] |
|
|
|
|
217
|
|
|
) -> JsonNavigatorSingleOptional: |
218
|
|
|
return JsonNavigatorSingleOptional(conversion(self.contents)) |
219
|
|
|
|
220
|
|
|
|
221
|
|
|
@dataclass(frozen=True, eq=True) |
|
|
|
|
222
|
|
|
class JsonNavigatorListOfOptionals(AbstractJsonNavigator): |
223
|
|
|
contents: List[Optional[T]] |
224
|
|
|
|
225
|
|
|
@property |
226
|
|
|
def to_list(self) -> List[T]: |
|
|
|
|
227
|
|
|
return [k for k in self.contents if k is not None] |
228
|
|
|
|
229
|
|
|
@property |
230
|
|
|
def to_set(self) -> FrozenSet[T]: |
|
|
|
|
231
|
|
|
return frozenset([k for k in self.contents if k is not None]) |
232
|
|
|
|
233
|
|
|
def __truediv__( |
234
|
|
|
self, key: Union[None, str, FilterFn, Callable[[Any], Any]] |
|
|
|
|
235
|
|
|
) -> JsonNavigatorListOfOptionals: |
236
|
|
|
if isinstance(key, FilterFn): |
|
|
|
|
237
|
|
|
return self._filter(key) |
238
|
|
|
else: |
239
|
|
|
return self._go_inside(key) |
240
|
|
|
|
241
|
|
|
def __rshift__(self, key: Callable[[List[Optional[Any]]], Any]) -> JsonNavigatorSingleOptional: |
242
|
|
|
# we can't skip 2 |
243
|
|
|
return self // key |
244
|
|
|
|
245
|
|
|
def __floordiv__( |
246
|
|
|
self, key: Callable[[List[Optional[Any]]], Any] |
|
|
|
|
247
|
|
|
) -> JsonNavigatorSingleOptional: |
248
|
|
|
return JsonNavigatorSingleOptional(key(self.contents)) |
249
|
|
|
|
250
|
|
|
def _go_inside( |
251
|
|
|
self, key: Union[None, str, Callable[[Any], Any]] |
|
|
|
|
252
|
|
|
) -> JsonNavigatorListOfOptionals: |
253
|
|
|
fn = _get_conversion_fn(key) |
|
|
|
|
254
|
|
|
return JsonNavigatorListOfOptionals( |
255
|
|
|
[None if content is None else fn(content) for content in self.contents] |
256
|
|
|
) |
257
|
|
|
|
258
|
|
|
def _filter( |
259
|
|
|
self, keep_if: Union[Callable[[Optional[T]], bool]] |
|
|
|
|
260
|
|
|
) -> JsonNavigatorListOfOptionals: |
261
|
|
|
return JsonNavigatorListOfOptionals([z for z in self.contents if keep_if(z)]) |
262
|
|
|
|
263
|
|
|
|
264
|
|
|
@dataclass(frozen=True, eq=True) |
|
|
|
|
265
|
|
|
class JsonNavigatorListOfLists(AbstractJsonNavigator): |
266
|
|
|
contents: List[List[Any]] |
267
|
|
|
|
268
|
|
|
def __truediv__( |
269
|
|
|
self, |
|
|
|
|
270
|
|
|
keys: Union[ |
|
|
|
|
271
|
|
|
Sequence[Union[None, str, Callable[[Any], Any]]], FilterFn, Callable[[List[T]], Any] |
272
|
|
|
], |
273
|
|
|
) -> JsonNavigatorListOfLists: |
274
|
|
|
if isinstance(keys, FilterFn): |
|
|
|
|
275
|
|
|
return self._filter(keys) |
276
|
|
|
else: |
277
|
|
|
return self._go_inside(keys) |
278
|
|
|
|
279
|
|
|
def __rshift__(self, conversion: Callable[[List[List[T]]], Any]) -> JsonNavigatorSingleOptional: |
280
|
|
|
return JsonNavigatorSingleOptional(conversion(self.contents)) |
281
|
|
|
|
282
|
|
|
def __floordiv__(self, conversion: Callable[[List[T]], Any]) -> JsonNavigatorListOfOptionals: |
283
|
|
|
return JsonNavigatorListOfOptionals([conversion(z) for z in self.contents]) |
284
|
|
|
|
285
|
|
|
def _go_inside( |
286
|
|
|
self, keys: Sequence[Union[None, str, Callable[[Any], Any]]] |
|
|
|
|
287
|
|
|
) -> JsonNavigatorListOfLists: |
288
|
|
|
fns = [_get_conversion_fn(fn) for fn in keys] |
289
|
|
|
return JsonNavigatorListOfLists( |
290
|
|
|
[ |
291
|
|
|
[fn(value) for value, fn in BaseTools.zip_list(contents, fns)] |
292
|
|
|
for contents in self.contents |
293
|
|
|
] |
294
|
|
|
) |
295
|
|
|
|
296
|
|
|
def _filter(self, keep_if: FilterFn) -> JsonNavigatorListOfLists: |
297
|
|
|
return JsonNavigatorListOfLists([z for z in self.contents if keep_if(z)]) |
298
|
|
|
|
299
|
|
|
|
300
|
|
|
@dataclass(frozen=True, eq=True) |
|
|
|
|
301
|
|
|
class JsonNavigator(AbstractJsonNavigator): |
302
|
|
|
contents: List[NestedDotDict] |
303
|
|
|
|
304
|
|
|
@property |
305
|
|
|
def get(self) -> List[NestedDotDict]: |
|
|
|
|
306
|
|
|
return self.contents |
307
|
|
|
|
308
|
|
|
@classmethod |
309
|
|
|
def create( |
|
|
|
|
310
|
|
|
cls, dct: Union[dict, NestedDotDict, Sequence[dict], Sequence[NestedDotDict]] |
|
|
|
|
311
|
|
|
) -> JsonNavigator: |
312
|
|
|
if hasattr(dct, "items"): |
313
|
|
|
dct = [dct] |
314
|
|
|
return JsonNavigator([NestedDotDict(dict(**d, _landmark="")) for d in dct]) |
315
|
|
|
|
316
|
|
|
def __truediv__( |
317
|
|
|
self, key: Union[int, str, FilterFn, Callable[[NestedDotDict], NestedDotDict]] |
|
|
|
|
318
|
|
|
) -> JsonNavigator: |
319
|
|
|
if isinstance(key, FilterFn): |
|
|
|
|
320
|
|
|
return self._filter(key) |
321
|
|
|
else: |
322
|
|
|
return self._go_inside(key) |
323
|
|
|
|
324
|
|
|
def _go_inside(self, key: Union[int, str]) -> JsonNavigator: |
325
|
|
|
new = [] |
326
|
|
|
for z in self.contents: |
|
|
|
|
327
|
|
|
if key in z: |
328
|
|
|
# nav = z.get_as("_nav", list, []) |
329
|
|
|
# nav.append(z[key]) |
330
|
|
|
if isinstance(z.get(key), list): |
331
|
|
|
new.extend([NestedDotDict(dict(**m)) for m in z[key]]) |
332
|
|
|
elif isinstance(z.get(key), NestedDotDict): |
333
|
|
|
new.append(NestedDotDict(dict(**z[key]))) |
334
|
|
|
elif isinstance(z.get(key), dict): |
335
|
|
|
new.append(NestedDotDict(dict(**z[key]))) |
336
|
|
|
else: |
337
|
|
|
raise ValueError(f"{key} value is {type(z[key])}: {z[key]}") |
338
|
|
|
return JsonNavigator(new) |
339
|
|
|
|
340
|
|
|
def _filter(self, keep_where: FilterFn) -> JsonNavigator: |
341
|
|
|
if callable(keep_where): |
|
|
|
|
342
|
|
|
return JsonNavigator([z for z in self.contents if keep_where(z)]) |
343
|
|
|
else: |
344
|
|
|
key, values = keep_where |
345
|
|
|
if not isinstance(values, (Set, FrozenSet, List)): |
|
|
|
|
346
|
|
|
values = {values} |
347
|
|
|
return JsonNavigator([z for z in self.contents if z.get(key) in values]) |
348
|
|
|
|
349
|
|
|
def __mod__(self, key: Union[int, str]) -> JsonNavigator: |
350
|
|
|
new = {} |
351
|
|
|
for z in self.contents: |
|
|
|
|
352
|
|
|
if z[key] in new: |
353
|
|
|
raise ValueError(f"{key} found twice") |
354
|
|
|
new[z[key]] = z |
355
|
|
|
return JsonNavigator([NestedDotDict(new)]) |
356
|
|
|
|
357
|
|
|
def __floordiv__(self, keys: Sequence[str]) -> JsonNavigatorListOfLists: |
358
|
|
|
return JsonNavigatorListOfLists([[z.get(key) for key in keys] for z in self.contents]) |
359
|
|
|
|
360
|
|
|
def __rshift__(self, key: str) -> JsonNavigatorListOfOptionals: |
361
|
|
|
return JsonNavigatorListOfOptionals([z.get(key) for z in self.contents]) |
362
|
|
|
|
363
|
|
|
|
364
|
|
|
__all__ = ["JsonNavigator", "JsonNavigatorListOfLists", "Fns", "FilterFn"] |
365
|
|
|
|