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
|
import pprint
from stone.ir import (
Boolean,
Bytes,
Float32,
Float64,
Int32,
Int64,
List,
Map,
String,
Timestamp,
UInt32,
UInt64,
Void,
is_struct_type,
is_boolean_type,
is_list_type,
is_map_type,
is_numeric_type,
is_string_type,
is_tag_ref,
is_user_defined_type,
unwrap_nullable,
)
from .helpers import split_words
# This file defines *stylistic* choices for Swift
# (ie, that class names are UpperCamelCase and that variables are lowerCamelCase)
_type_table = {
Boolean: 'Bool',
Bytes: 'Data',
Float32: 'Float',
Float64: 'Double',
Int32: 'Int32',
Int64: 'Int64',
List: 'Array',
Map: 'Dictionary',
String: 'String',
Timestamp: 'Date',
UInt32: 'UInt32',
UInt64: 'UInt64',
Void: 'Void',
}
_objc_type_table = {
Boolean: 'NSNumber',
Bytes: 'Data',
Float32: 'NSNumber',
Float64: 'NSNumber',
Int32: 'NSNumber',
Int64: 'NSNumber',
List: 'Array',
Map: 'Dictionary',
String: 'String',
Timestamp: 'Date',
UInt32: 'NSNumber',
UInt64: 'NSNumber',
Void: '',
}
_reserved_words = {
'description',
'bool',
'nsdata'
'float',
'double',
'int32',
'int64',
'list',
'string',
'timestamp',
'uint32',
'uint64',
'void',
'associatedtype',
'class',
'deinit',
'enum',
'extension',
'func',
'import',
'init',
'inout',
'internal',
'let',
'operator',
'private',
'protocol',
'public',
'static',
'struct',
'subscript',
'typealias',
'var',
'default',
'hash',
'client',
}
def fmt_obj(o):
assert not isinstance(o, dict), "Only use for base type literals"
if o is True:
return 'true'
if o is False:
return 'false'
if o is None:
return 'nil'
if o == '':
return '""'
elif isinstance(o, str):
return '"{}"'.format(o)
return pprint.pformat(o, width=1)
def _format_camelcase(name, lower_first=True):
words = [word.capitalize() for word in split_words(name)]
if lower_first:
words[0] = words[0].lower()
ret = ''.join(words)
if ret.lower() in _reserved_words:
ret += '_'
return ret
def fmt_class(name):
return _format_camelcase(name, lower_first=False)
def fmt_func(name, version):
if version > 1:
name = '{}_v{}'.format(name, version)
name = _format_camelcase(name)
return name
def fmt_type(data_type):
data_type, nullable = unwrap_nullable(data_type)
if is_user_defined_type(data_type):
result = '{}.{}'.format(fmt_class(data_type.namespace.name),
fmt_class(data_type.name))
else:
result = _type_table.get(data_type.__class__, fmt_class(data_type.name))
if is_list_type(data_type):
result = result + '<{}>'.format(fmt_type(data_type.data_type))
if is_map_type(data_type):
result = result + '<{}, {}>'.format(fmt_type(data_type.key_data_type),
fmt_type(data_type.value_data_type))
return result if not nullable else result + '?'
def fmt_objc_type(data_type, allow_nullable=True):
data_type, nullable = unwrap_nullable(data_type)
if is_user_defined_type(data_type):
result = 'DBX{}{}'.format(fmt_class(data_type.namespace.name),
fmt_class(data_type.name))
else:
result = _objc_type_table.get(data_type.__class__, fmt_class(data_type.name))
if is_list_type(data_type):
result = result + '<{}>'.format(fmt_objc_type(data_type.data_type, False))
elif is_map_type(data_type):
result = result + '<String, {}>'.format(fmt_objc_type(data_type.value_data_type))
return result if not nullable or not allow_nullable else result + '?'
def fmt_var(name):
return _format_camelcase(name)
def fmt_default_value(field):
if is_tag_ref(field.default):
return '{}.{}Serializer().serialize(.{})'.format(
fmt_class(field.default.union_data_type.namespace.name),
fmt_class(field.default.union_data_type.name),
fmt_var(field.default.tag_name))
elif is_list_type(field.data_type):
return '.array({})'.format(field.default)
elif is_numeric_type(field.data_type):
return '.number({})'.format(field.default)
elif is_string_type(field.data_type):
return '.str({})'.format(fmt_obj(field.default))
elif is_boolean_type(field.data_type):
if field.default:
bool_str = '1'
else:
bool_str = '0'
return '.number({})'.format(bool_str)
else:
raise TypeError('Can\'t handle default value type %r' %
type(field.data_type))
def fmt_route_name(route):
if route.version == 1:
return route.name
else:
return '{}_v{}'.format(route.name, route.version)
def fmt_route_name_namespace(route, namespace_name):
return '{}/{}'.format(namespace_name, fmt_route_name(route))
def fmt_func_namespace(name, version, namespace_name):
return '{}_{}'.format(namespace_name, fmt_func(name, version))
def check_route_name_conflict(namespace):
"""
Check name conflicts among generated route definitions. Raise a runtime exception when a
conflict is encountered.
"""
route_by_name = {}
for route in namespace.routes:
route_name = fmt_func(route.name, route.version)
if route_name in route_by_name:
other_route = route_by_name[route_name]
raise RuntimeError(
'There is a name conflict between {!r} and {!r}'.format(other_route, route))
route_by_name[route_name] = route
def mapped_list_info(data_type):
list_data_type, list_nullable = unwrap_nullable(data_type.data_type)
list_depth = 0
while is_list_type(list_data_type):
list_depth += 1
list_data_type, list_nullable = unwrap_nullable(list_data_type.data_type)
prefix = ''
suffix = ''
if list_depth > 0:
i = 0
while i < list_depth:
i += 1
prefix = '{}{{ $0.map '.format(prefix)
suffix = '{} }}'.format(suffix)
return (list_depth, prefix, suffix, list_data_type, list_nullable)
def field_is_user_defined(field):
data_type, nullable = unwrap_nullable(field.data_type)
return is_user_defined_type(data_type) and not nullable
def field_is_user_defined_optional(field):
data_type, nullable = unwrap_nullable(field.data_type)
return is_user_defined_type(data_type) and nullable
def field_is_user_defined_map(field):
data_type, _ = unwrap_nullable(field.data_type)
return is_map_type(data_type) and is_user_defined_type(data_type.value_data_type)
def field_is_user_defined_list(field):
data_type, _ = unwrap_nullable(field.data_type)
if is_list_type(data_type):
list_data_type, _ = unwrap_nullable(data_type.data_type)
return is_user_defined_type(list_data_type)
else:
return False
# List[typing.Tuple[let_name: str, swift_type: str, objc_type: str]]
def objc_datatype_value_type_tuples(data_type):
ret = []
# if list type get the data type of the item
if is_list_type(data_type):
data_type = data_type.data_type
# if map type get the data type of the value
if is_map_type(data_type):
data_type = data_type.value_data_type
# if data_type is a struct type and has subtypes, process them into labels and types
if is_struct_type(data_type) and data_type.has_enumerated_subtypes():
all_subtypes = data_type.get_all_subtypes_with_tags()
for subtype in all_subtypes:
# subtype[0] is the tag name and subtype[1] is the subtype struct itself
struct = subtype[1]
case_let_name = fmt_var(struct.name)
swift_type = fmt_type(struct)
objc_type = fmt_objc_type(struct)
ret.append((case_let_name, swift_type, objc_type))
return ret
def field_datatype_has_subtypes(field) -> bool:
return datatype_has_subtypes(field.data_type)
def datatype_has_subtypes(data_type) -> bool:
return len(objc_datatype_value_type_tuples(data_type)) > 0
|