hat.json

18def equals(a: Data,
19           b: Data
20           ) -> bool:
21    """Equality comparison of json serializable data.
22
23    Tests for equality of data according to JSON format. Notably, ``bool``
24    values are not considered equal to numeric values in any case. This is
25    different from default equality comparison, which considers `False`
26    equal to `0` and `0.0`; and `True` equal to `1` and `1.0`.
27
28    Example::
29
30        assert equals(0, 0.0) is True
31        assert equals({'a': 1, 'b': 2}, {'b': 2, 'a': 1}) is True
32        assert equals(1, True) is False
33
34    """
35    if a is None:
36        return b is None
37
38    if isinstance(a, bool):
39        return isinstance(b, bool) and a == b
40
41    if isinstance(a, (int, float)):
42        return (isinstance(b, (int, float)) and
43                not isinstance(b, bool) and
44                a == b)
45
46    if isinstance(a, str):
47        return isinstance(b, str) and a == b
48
49    if isinstance(a, list):
50        return (isinstance(b, list) and
51                len(a) == len(b) and
52                all(equals(i, j) for i, j in zip(a, b)))
53
54    if isinstance(a, dict):
55        return (isinstance(b, dict) and
56                len(a) == len(b) and
57                all(equals(a[key], b[key]) for key in a.keys()))
58
59    raise TypeError('invalid json type')

62def clone(data: Data) -> Data:
63    """Deep clone data
64
65    This function recursively creates new instances of array and object data
66    based on input data. Resulting json data is equal to provided data.
67
68    Example::
69
70        x = {'a': [1, 2, 3]}
71        y = clone(x)
72        assert x is not y
73        assert x['a'] is not y['a']
74        assert equals(x, y)
75
76    """
77    if isinstance(data, list):
78        return [clone(i) for i in data]
79
80    if isinstance(data, dict):
81        return {k: clone(v) for k, v in data.items()}
82
83    return data

 86def flatten(data: Data) -> Iterable[Data]:
 87    """Flatten JSON data
 88
 89    If `data` is array, this generator recursively yields result of `flatten`
 90    call with each element of input list. For other `Data` types, input data is
 91    yielded.
 92
 93    Example::
 94
 95        data = [1, [], [2], {'a': [3]}]
 96        result = [1, 2, {'a': [3]}]
 97        assert list(flatten(data)) == result
 98
 99    """
100    if isinstance(data, list):
101        for i in data:
102            yield from flatten(i)
103
104    else:
105        yield data

17def get(data: Data,
18        path: Path,
19        default: Data | None = None
20        ) -> Data:
21    """Get data element referenced by path
22
23    Example::
24
25        data = {'a': [1, 2, [3, 4]]}
26        path = ['a', 2, 0]
27        assert get(data, path) == 3
28
29        data = [1, 2, 3]
30        assert get(data, 0) == 1
31        assert get(data, 5) is None
32        assert get(data, 5, default=123) == 123
33
34    """
35    for i in flatten(path):
36        if isinstance(i, str):
37            if not isinstance(data, dict) or i not in data:
38                return default
39            data = data[i]
40
41        elif isinstance(i, int) and not isinstance(i, bool):
42            if not isinstance(data, list):
43                return default
44            try:
45                data = data[i]
46            except IndexError:
47                return default
48
49        else:
50            raise ValueError('invalid path')
51
52    return data

 55def set_(data: Data,
 56         path: Path,
 57         value: Data
 58         ) -> Data:
 59    """Create new data by setting data path element value
 60
 61    Example::
 62
 63        data = [1, {'a': 2, 'b': 3}, 4]
 64        path = [1, 'b']
 65        result = set_(data, path, 5)
 66        assert result == [1, {'a': 2, 'b': 5}, 4]
 67        assert result is not data
 68
 69        data = [1, 2, 3]
 70        result = set_(data, 4, 4)
 71        assert result == [1, 2, 3, None, 4]
 72
 73    """
 74    parents = collections.deque()
 75
 76    for i in flatten(path):
 77        parent = data
 78
 79        if isinstance(i, str):
 80            data = data.get(i) if isinstance(data, dict) else None
 81
 82        elif isinstance(i, int) and not isinstance(i, bool):
 83            try:
 84                data = data[i] if isinstance(data, list) else None
 85            except IndexError:
 86                data = None
 87
 88        else:
 89            raise ValueError('invalid path')
 90
 91        parents.append((parent, i))
 92
 93    while parents:
 94        parent, i = parents.pop()
 95
 96        if isinstance(i, str):
 97            parent = dict(parent) if isinstance(parent, dict) else {}
 98            parent[i] = value
 99
100        elif isinstance(i, int) and not isinstance(i, bool):
101            if not isinstance(parent, list):
102                parent = []
103
104            if i >= len(parent):
105                parent = [*parent,
106                          *itertools.repeat(None, i - len(parent) + 1)]
107
108            elif i < 0 and (-i) > len(parent):
109                parent = [*itertools.repeat(None, (-i) - len(parent)),
110                          *parent]
111
112            else:
113                parent = list(parent)
114
115            parent[i] = value
116
117        else:
118            raise ValueError('invalid path')
119
120        value = parent
121
122    return value

125def remove(data: Data,
126           path: Path
127           ) -> Data:
128    """Create new data by removing part of data referenced by path
129
130    Example::
131
132        data = [1, {'a': 2, 'b': 3}, 4]
133        path = [1, 'b']
134        result = remove(data, path)
135        assert result == [1, {'a': 2}, 4]
136        assert result is not data
137
138        data = [1, 2, 3]
139        result = remove(data, 4)
140        assert result == [1, 2, 3]
141
142    """
143    result = data
144    parents = collections.deque()
145
146    for i in flatten(path):
147        parent = data
148
149        if isinstance(i, str):
150            if not isinstance(data, dict) or i not in data:
151                return result
152            data = data[i]
153
154        elif isinstance(i, int) and not isinstance(i, bool):
155            if not isinstance(data, list):
156                return result
157            try:
158                data = data[i]
159            except IndexError:
160                return result
161
162        else:
163            raise ValueError('invalid path')
164
165        parents.append((parent, i))
166
167    result = None
168
169    while parents:
170        parent, i = parents.pop()
171
172        if isinstance(i, str):
173            parent = dict(parent)
174
175        elif isinstance(i, int) and not isinstance(i, bool):
176            parent = list(parent)
177
178        else:
179            raise ValueError('invalid path')
180
181        if result is None:
182            del parent[i]
183
184        else:
185            parent[i] = result
186
187        result = parent
188
189    return result

192class Storage:
193    """JSON data storage
194
195    Helper class representing observable JSON data state manipulated with
196    path based get/set/remove functions.
197
198    """
199
200    def __init__(self, data: Data = None):
201        self._data = data
202        self._change_cbs = util.CallbackRegistry()
203
204    @property
205    def data(self) -> Data:
206        """Data"""
207        return self._data
208
209    def register_change_cb(self,
210                           cb: Callable[[Data], None]
211                           ) -> util.RegisterCallbackHandle:
212        """Register data change callback"""
213        return self._change_cbs.register(cb)
214
215    def get(self, path: Path, default: Data | None = None):
216        """Get data"""
217        return get(self._data, path, default)
218
219    def set(self, path: Path, value: Data):
220        """Set data"""
221        self._data = set_(self._data, path, value)
222        self._change_cbs.notify(self._data)
223
224    def remove(self, path: Path):
225        """Remove data"""
226        self._data = remove(self._data, path)
227        self._change_cbs.notify(self._data)

21class Format(enum.Enum):
22    """Encoding format"""
23    JSON = 'json'
24    YAML = 'yaml'
25    TOML = 'toml'

28def encode(data: Data,
29           format: Format = Format.JSON,
30           indent: int | None = None
31           ) -> str:
32    """Encode JSON data.
33
34    In case of TOML format, data must be JSON Object.
35
36    Args:
37        data: JSON data
38        format: encoding format
39        indent: indentation size
40
41    """
42    if format == Format.JSON:
43        return json.dumps(data, indent=indent, allow_nan=False)
44
45    if format == Format.YAML:
46        dumper = (yaml.CSafeDumper if hasattr(yaml, 'CSafeDumper')
47                  else yaml.SafeDumper)
48        return str(yaml.dump(data, indent=indent, Dumper=dumper))
49
50    if format == Format.TOML:
51        return tomli_w.dumps(data)
52
53    raise ValueError('unsupported format')

56def decode(data_str: str,
57           format: Format = Format.JSON
58           ) -> Data:
59    """Decode JSON data.
60
61    Args:
62        data_str: encoded JSON data
63        format: encoding format
64
65    """
66    if format == Format.JSON:
67        return json.loads(data_str)
68
69    if format == Format.YAML:
70        loader = (yaml.CSafeLoader if hasattr(yaml, 'CSafeLoader')
71                  else yaml.SafeLoader)
72        return yaml.load(io.StringIO(data_str), Loader=loader)
73
74    if format == Format.TOML:
75        return toml.loads(data_str)
76
77    raise ValueError('unsupported format')

80def get_file_format(path: pathlib.PurePath) -> Format:
81    """Detect file format based on path suffix"""
82    if path.suffix == '.json':
83        return Format.JSON
84
85    if path.suffix in ('.yaml', '.yml'):
86        return Format.YAML
87
88    if path.suffix == '.toml':
89        return Format.TOML
90
91    raise ValueError('can not determine format from path suffix')

 94def encode_file(data: Data,
 95                path: pathlib.PurePath,
 96                format: Format | None = None,
 97                indent: int | None = 4):
 98    """Encode JSON data to file.
 99
100    If `format` is ``None``, encoding format is derived from path suffix.
101
102    In case of TOML format, data must be JSON Object.
103
104    Args:
105        data: JSON data
106        path: file path
107        format: encoding format
108        indent: indentation size
109
110    """
111    if format is None:
112        format = get_file_format(path)
113
114    flags = 'w' if format != Format.TOML else 'wb'
115    encoding = 'utf-8' if format != Format.TOML else None
116
117    with open(path, flags, encoding=encoding) as f:
118        encode_stream(data, f, format, indent)

121def decode_file(path: pathlib.PurePath,
122                format: Format | None = None
123                ) -> Data:
124    """Decode JSON data from file.
125
126    If `format` is ``None``, encoding format is derived from path suffix.
127
128    Args:
129        path: file path
130        format: encoding format
131
132    """
133    if format is None:
134        format = get_file_format(path)
135
136    flags = 'r' if format != Format.TOML else 'rb'
137    encoding = 'utf-8' if format != Format.TOML else None
138
139    with open(path, flags, encoding=encoding) as f:
140        return decode_stream(f, format)

143def encode_stream(data: Data,
144                  stream: io.TextIOBase | io.RawIOBase,
145                  format: Format = Format.JSON,
146                  indent: int | None = 4):
147    """Encode JSON data to stream.
148
149    In case of TOML format, data must be JSON Object.
150
151    In case of TOML format, `stream` should be `io.RawIOBase`. For
152    other formats, `io.TextIOBase` is expected.
153
154    Args:
155        data: JSON data
156        stream: output stream
157        format: encoding format
158        indent: indentation size
159
160    """
161    if format == Format.JSON:
162        json.dump(data, stream, indent=indent, allow_nan=False)
163
164    elif format == Format.YAML:
165        dumper = (yaml.CSafeDumper if hasattr(yaml, 'CSafeDumper')
166                  else yaml.SafeDumper)
167        yaml.dump(data, stream, indent=indent, Dumper=dumper,
168                  explicit_start=True, explicit_end=True)
169
170    elif format == Format.TOML:
171        tomli_w.dump(data, stream)
172
173    else:
174        raise ValueError('unsupported format')

177def decode_stream(stream: io.TextIOBase | io.RawIOBase,
178                  format: Format = Format.JSON
179                  ) -> Data:
180    """Decode JSON data from stream.
181
182    In case of TOML format, `stream` should be `io.RawIOBase`. For
183    other formats, `io.TextIOBase` is expected.
184
185    Args:
186        stream: input stream
187        format: encoding format
188
189    """
190    if format == Format.JSON:
191        return json.load(stream)
192
193    if format == Format.YAML:
194        loader = (yaml.CSafeLoader if hasattr(yaml, 'CSafeLoader')
195                  else yaml.SafeLoader)
196        return yaml.load(stream, Loader=loader)
197
198    if format == Format.TOML:
199        return toml.load(stream)
200
201    raise ValueError('unsupported format')

204def read_conf(path: pathlib.Path | None,
205              default_path: pathlib.Path | None = None,
206              default_suffixes: list[str] = ['.yaml', '.yml', '.toml', '.json'],  # NOQA
207              stdio_path: pathlib.Path | None = pathlib.Path('-')
208              ) -> Data:
209    """Read configuration formated as JSON data"""
210    if stdio_path and path == stdio_path:
211        return decode_stream(sys.stdin)
212
213    if path:
214        return decode_file(path)
215
216    if not default_path:
217        raise Exception('invalid configuration path')
218
219    for suffix in default_suffixes:
220        path = default_path.with_suffix(suffix)
221        if path.exists():
222            break
223
224    return decode_file(path)

 9def diff(src: Data,
10         dst: Data
11         ) -> Data:
12    """Generate JSON Patch diff.
13
14    Example::
15
16        src = [1, {'a': 2}, 3]
17        dst = [1, {'a': 4}, 3]
18        result = diff(src, dst)
19        assert result == [{'op': 'replace', 'path': '/1/a', 'value': 4}]
20
21    """
22    return jsonpatch.JsonPatch.from_diff(src, dst).patch

25def patch(data: Data,
26          diff: Data
27          ) -> Data:
28    """Apply JSON Patch diff.
29
30    Example::
31
32        data = [1, {'a': 2}, 3]
33        d = [{'op': 'replace', 'path': '/1/a', 'value': 4}]
34        result = patch(data, d)
35        assert result == [1, {'a': 4}, 3]
36
37    """
38    return jsonpatch.apply_patch(data, diff)

34def create_schema_repository(*args: pathlib.PurePath | Schema
35                             ) -> SchemaRepository:
36    """Create JSON Schema repository.
37
38    Repository can be initialized with multiple arguments, which can be
39    instances of ``pathlib.PurePath`` or ``Schema``.
40
41    If an argument is of type ``pathlib.PurePath``, and path points to file
42    with a suffix '.json', '.yml' or '.yaml', json serializable data is decoded
43    from the file. Otherwise, it is assumed that path points to a directory,
44    which is recursively searched for json and yaml files. All decoded schemas
45    are added to the repository. If a schema with the same `id` was previously
46    added, an exception is raised.
47
48    If an argument is of type ``Schema``, it should be a json serializable data
49    representation of a JSON schema. If a schema with the same `id` was
50    previously added, an exception is raised.
51
52    """
53    repo = {}
54
55    for arg in args:
56        if isinstance(arg, pathlib.PurePath):
57            if arg.suffix in _schema_path_suffixes:
58                paths = [arg]
59
60            else:
61                paths = itertools.chain.from_iterable(
62                    arg.rglob(f'*{i}') for i in _schema_path_suffixes)
63
64            schemas = (decode_file(path) for path in paths)
65
66        elif isinstance(arg, dict):
67            schemas = [arg]
68
69        else:
70            raise TypeError('invalid argument type')
71
72        for schema in schemas:
73            if '$schema' in schema:
74                meta_schema_id = urllib.parse.urldefrag(schema['$schema']).url
75                if meta_schema_id not in _meta_schema_ids:
76                    raise Exception(
77                        f"unsupported meta schema id {meta_schema_id}")
78
79            else:
80                schema = {'$schema': _default_meta_schema_id,
81                          **schema}
82
83            schema_id = schema.get('$id')
84            if not schema_id:
85                schema_id = schema.get('id')
86            if not schema_id:
87                raise Exception('invalid schema id')
88
89            sanitized_schema_id = urllib.parse.urldefrag(schema_id).url
90            if sanitized_schema_id in repo:
91                raise Exception(f"duplicate schema id {sanitized_schema_id}")
92
93            repo[sanitized_schema_id] = schema
94
95    return repo

 98def merge_schema_repositories(*repos: SchemaRepository
 99                              ) -> SchemaRepository:
100    """Merge JSON Schema repositories.
101
102    Exception is raised is multiple repositories contain same schema id with
103    diferent schemas.
104
105    """
106    result = {}
107
108    for repo in repos:
109        for schema_id, schema in repo.items():
110            other_schema = result.get(schema_id)
111            if other_schema is not None and other_schema != schema:
112                raise Exception(f"conflict for schema id {schema_id}")
113
114            result[schema_id] = schema
115
116    return result