Martin v. Löwis | 9ea6788 | 2003-01-25 15:29:56 +0000 | [diff] [blame] | 1 | """Implementation of the DOM Level 3 'LS-Load' feature.""" |
| 2 | |
| 3 | import copy |
| 4 | import xml.dom |
| 5 | |
| 6 | from xml.dom.minicompat import * |
| 7 | |
| 8 | from xml.dom.NodeFilter import NodeFilter |
| 9 | |
| 10 | |
| 11 | __all__ = ["DOMBuilder", "DOMEntityResolver", "DOMInputSource"] |
| 12 | |
| 13 | |
| 14 | class Options: |
| 15 | """Features object that has variables set for each DOMBuilder feature. |
| 16 | |
| 17 | The DOMBuilder class uses an instance of this class to pass settings to |
| 18 | the ExpatBuilder class. |
| 19 | """ |
| 20 | |
| 21 | # Note that the DOMBuilder class in LoadSave constrains which of these |
| 22 | # values can be set using the DOM Level 3 LoadSave feature. |
| 23 | |
| 24 | namespaces = 1 |
| 25 | namespace_declarations = True |
| 26 | validation = False |
| 27 | external_parameter_entities = True |
| 28 | external_general_entities = True |
| 29 | external_dtd_subset = True |
| 30 | validate_if_schema = False |
| 31 | validate = False |
| 32 | datatype_normalization = False |
| 33 | create_entity_ref_nodes = True |
| 34 | entities = True |
| 35 | whitespace_in_element_content = True |
| 36 | cdata_sections = True |
| 37 | comments = True |
| 38 | charset_overrides_xml_encoding = True |
| 39 | infoset = False |
| 40 | supported_mediatypes_only = False |
| 41 | |
| 42 | errorHandler = None |
| 43 | filter = None |
| 44 | |
| 45 | |
| 46 | class DOMBuilder: |
| 47 | entityResolver = None |
| 48 | errorHandler = None |
| 49 | filter = None |
| 50 | |
| 51 | ACTION_REPLACE = 1 |
| 52 | ACTION_APPEND_AS_CHILDREN = 2 |
| 53 | ACTION_INSERT_AFTER = 3 |
| 54 | ACTION_INSERT_BEFORE = 4 |
| 55 | |
| 56 | _legal_actions = (ACTION_REPLACE, ACTION_APPEND_AS_CHILDREN, |
| 57 | ACTION_INSERT_AFTER, ACTION_INSERT_BEFORE) |
| 58 | |
| 59 | def __init__(self): |
| 60 | self._options = Options() |
| 61 | |
| 62 | def _get_entityResolver(self): |
| 63 | return self.entityResolver |
| 64 | def _set_entityResolver(self, entityResolver): |
| 65 | self.entityResolver = entityResolver |
| 66 | |
| 67 | def _get_errorHandler(self): |
| 68 | return self.errorHandler |
| 69 | def _set_errorHandler(self, errorHandler): |
| 70 | self.errorHandler = errorHandler |
| 71 | |
| 72 | def _get_filter(self): |
| 73 | return self.filter |
| 74 | def _set_filter(self, filter): |
| 75 | self.filter = filter |
| 76 | |
| 77 | def setFeature(self, name, state): |
| 78 | if self.supportsFeature(name): |
| 79 | state = state and 1 or 0 |
| 80 | try: |
| 81 | settings = self._settings[(_name_xform(name), state)] |
| 82 | except KeyError: |
| 83 | raise xml.dom.NotSupportedErr( |
Walter Dörwald | 70a6b49 | 2004-02-12 17:35:32 +0000 | [diff] [blame^] | 84 | "unsupported feature: %r" % (name,)) |
Martin v. Löwis | 9ea6788 | 2003-01-25 15:29:56 +0000 | [diff] [blame] | 85 | else: |
| 86 | for name, value in settings: |
| 87 | setattr(self._options, name, value) |
| 88 | else: |
| 89 | raise xml.dom.NotFoundErr("unknown feature: " + repr(name)) |
| 90 | |
| 91 | def supportsFeature(self, name): |
| 92 | return hasattr(self._options, _name_xform(name)) |
| 93 | |
| 94 | def canSetFeature(self, name, state): |
| 95 | key = (_name_xform(name), state and 1 or 0) |
| 96 | return self._settings.has_key(key) |
| 97 | |
| 98 | # This dictionary maps from (feature,value) to a list of |
| 99 | # (option,value) pairs that should be set on the Options object. |
| 100 | # If a (feature,value) setting is not in this dictionary, it is |
| 101 | # not supported by the DOMBuilder. |
| 102 | # |
| 103 | _settings = { |
| 104 | ("namespace_declarations", 0): [ |
| 105 | ("namespace_declarations", 0)], |
| 106 | ("namespace_declarations", 1): [ |
| 107 | ("namespace_declarations", 1)], |
| 108 | ("validation", 0): [ |
| 109 | ("validation", 0)], |
| 110 | ("external_general_entities", 0): [ |
| 111 | ("external_general_entities", 0)], |
| 112 | ("external_general_entities", 1): [ |
| 113 | ("external_general_entities", 1)], |
| 114 | ("external_parameter_entities", 0): [ |
| 115 | ("external_parameter_entities", 0)], |
| 116 | ("external_parameter_entities", 1): [ |
| 117 | ("external_parameter_entities", 1)], |
| 118 | ("validate_if_schema", 0): [ |
| 119 | ("validate_if_schema", 0)], |
| 120 | ("create_entity_ref_nodes", 0): [ |
| 121 | ("create_entity_ref_nodes", 0)], |
| 122 | ("create_entity_ref_nodes", 1): [ |
| 123 | ("create_entity_ref_nodes", 1)], |
| 124 | ("entities", 0): [ |
| 125 | ("create_entity_ref_nodes", 0), |
| 126 | ("entities", 0)], |
| 127 | ("entities", 1): [ |
| 128 | ("entities", 1)], |
| 129 | ("whitespace_in_element_content", 0): [ |
| 130 | ("whitespace_in_element_content", 0)], |
| 131 | ("whitespace_in_element_content", 1): [ |
| 132 | ("whitespace_in_element_content", 1)], |
| 133 | ("cdata_sections", 0): [ |
| 134 | ("cdata_sections", 0)], |
| 135 | ("cdata_sections", 1): [ |
| 136 | ("cdata_sections", 1)], |
| 137 | ("comments", 0): [ |
| 138 | ("comments", 0)], |
| 139 | ("comments", 1): [ |
| 140 | ("comments", 1)], |
| 141 | ("charset_overrides_xml_encoding", 0): [ |
| 142 | ("charset_overrides_xml_encoding", 0)], |
| 143 | ("charset_overrides_xml_encoding", 1): [ |
| 144 | ("charset_overrides_xml_encoding", 1)], |
| 145 | ("infoset", 0): [], |
| 146 | ("infoset", 1): [ |
| 147 | ("namespace_declarations", 0), |
| 148 | ("validate_if_schema", 0), |
| 149 | ("create_entity_ref_nodes", 0), |
| 150 | ("entities", 0), |
| 151 | ("cdata_sections", 0), |
| 152 | ("datatype_normalization", 1), |
| 153 | ("whitespace_in_element_content", 1), |
| 154 | ("comments", 1), |
| 155 | ("charset_overrides_xml_encoding", 1)], |
| 156 | ("supported_mediatypes_only", 0): [ |
| 157 | ("supported_mediatypes_only", 0)], |
| 158 | ("namespaces", 0): [ |
| 159 | ("namespaces", 0)], |
| 160 | ("namespaces", 1): [ |
| 161 | ("namespaces", 1)], |
| 162 | } |
| 163 | |
| 164 | def getFeature(self, name): |
| 165 | xname = _name_xform(name) |
| 166 | try: |
| 167 | return getattr(self._options, xname) |
| 168 | except AttributeError: |
| 169 | if name == "infoset": |
| 170 | options = self._options |
| 171 | return (options.datatype_normalization |
| 172 | and options.whitespace_in_element_content |
| 173 | and options.comments |
| 174 | and options.charset_overrides_xml_encoding |
| 175 | and not (options.namespace_declarations |
| 176 | or options.validate_if_schema |
| 177 | or options.create_entity_ref_nodes |
| 178 | or options.entities |
| 179 | or options.cdata_sections)) |
| 180 | raise xml.dom.NotFoundErr("feature %s not known" % repr(name)) |
| 181 | |
| 182 | def parseURI(self, uri): |
| 183 | if self.entityResolver: |
| 184 | input = self.entityResolver.resolveEntity(None, uri) |
| 185 | else: |
| 186 | input = DOMEntityResolver().resolveEntity(None, uri) |
| 187 | return self.parse(input) |
| 188 | |
| 189 | def parse(self, input): |
| 190 | options = copy.copy(self._options) |
| 191 | options.filter = self.filter |
| 192 | options.errorHandler = self.errorHandler |
| 193 | fp = input.byteStream |
| 194 | if fp is None and options.systemId: |
| 195 | import urllib2 |
| 196 | fp = urllib2.urlopen(input.systemId) |
| 197 | return self._parse_bytestream(fp, options) |
| 198 | |
| 199 | def parseWithContext(self, input, cnode, action): |
| 200 | if action not in self._legal_actions: |
| 201 | raise ValueError("not a legal action") |
| 202 | raise NotImplementedError("Haven't written this yet...") |
| 203 | |
| 204 | def _parse_bytestream(self, stream, options): |
| 205 | import xml.dom.expatbuilder |
| 206 | builder = xml.dom.expatbuilder.makeBuilder(options) |
| 207 | return builder.parseFile(stream) |
| 208 | |
| 209 | |
| 210 | def _name_xform(name): |
| 211 | return name.lower().replace('-', '_') |
| 212 | |
| 213 | |
| 214 | class DOMEntityResolver(NewStyle): |
| 215 | __slots__ = '_opener', |
| 216 | |
| 217 | def resolveEntity(self, publicId, systemId): |
| 218 | assert systemId is not None |
| 219 | source = DOMInputSource() |
| 220 | source.publicId = publicId |
| 221 | source.systemId = systemId |
| 222 | source.byteStream = self._get_opener().open(systemId) |
| 223 | |
| 224 | # determine the encoding if the transport provided it |
| 225 | source.encoding = self._guess_media_encoding(source) |
| 226 | |
| 227 | # determine the base URI is we can |
| 228 | import posixpath, urlparse |
| 229 | parts = urlparse.urlparse(systemId) |
| 230 | scheme, netloc, path, params, query, fragment = parts |
| 231 | # XXX should we check the scheme here as well? |
| 232 | if path and not path.endswith("/"): |
| 233 | path = posixpath.dirname(path) + "/" |
| 234 | parts = scheme, netloc, path, params, query, fragment |
| 235 | source.baseURI = urlparse.urlunparse(parts) |
| 236 | |
| 237 | return source |
| 238 | |
| 239 | def _get_opener(self): |
| 240 | try: |
| 241 | return self._opener |
| 242 | except AttributeError: |
| 243 | self._opener = self._create_opener() |
| 244 | return self._opener |
| 245 | |
| 246 | def _create_opener(self): |
| 247 | import urllib2 |
| 248 | return urllib2.build_opener() |
| 249 | |
| 250 | def _guess_media_encoding(self, source): |
| 251 | info = source.byteStream.info() |
| 252 | if info.has_key("Content-Type"): |
| 253 | for param in info.getplist(): |
| 254 | if param.startswith("charset="): |
| 255 | return param.split("=", 1)[1].lower() |
| 256 | |
| 257 | |
| 258 | class DOMInputSource(NewStyle): |
| 259 | __slots__ = ('byteStream', 'characterStream', 'stringData', |
| 260 | 'encoding', 'publicId', 'systemId', 'baseURI') |
| 261 | |
| 262 | def __init__(self): |
| 263 | self.byteStream = None |
| 264 | self.characterStream = None |
| 265 | self.stringData = None |
| 266 | self.encoding = None |
| 267 | self.publicId = None |
| 268 | self.systemId = None |
| 269 | self.baseURI = None |
| 270 | |
| 271 | def _get_byteStream(self): |
| 272 | return self.byteStream |
| 273 | def _set_byteStream(self, byteStream): |
| 274 | self.byteStream = byteStream |
| 275 | |
| 276 | def _get_characterStream(self): |
| 277 | return self.characterStream |
| 278 | def _set_characterStream(self, characterStream): |
| 279 | self.characterStream = characterStream |
| 280 | |
| 281 | def _get_stringData(self): |
| 282 | return self.stringData |
| 283 | def _set_stringData(self, data): |
| 284 | self.stringData = data |
| 285 | |
| 286 | def _get_encoding(self): |
| 287 | return self.encoding |
| 288 | def _set_encoding(self, encoding): |
| 289 | self.encoding = encoding |
| 290 | |
| 291 | def _get_publicId(self): |
| 292 | return self.publicId |
| 293 | def _set_publicId(self, publicId): |
| 294 | self.publicId = publicId |
| 295 | |
| 296 | def _get_systemId(self): |
| 297 | return self.systemId |
| 298 | def _set_systemId(self, systemId): |
| 299 | self.systemId = systemId |
| 300 | |
| 301 | def _get_baseURI(self): |
| 302 | return self.baseURI |
| 303 | def _set_baseURI(self, uri): |
| 304 | self.baseURI = uri |
| 305 | |
| 306 | |
| 307 | class DOMBuilderFilter: |
| 308 | """Element filter which can be used to tailor construction of |
| 309 | a DOM instance. |
| 310 | """ |
| 311 | |
| 312 | # There's really no need for this class; concrete implementations |
| 313 | # should just implement the endElement() and startElement() |
| 314 | # methods as appropriate. Using this makes it easy to only |
| 315 | # implement one of them. |
| 316 | |
| 317 | FILTER_ACCEPT = 1 |
| 318 | FILTER_REJECT = 2 |
| 319 | FILTER_SKIP = 3 |
| 320 | FILTER_INTERRUPT = 4 |
| 321 | |
| 322 | whatToShow = NodeFilter.SHOW_ALL |
| 323 | |
| 324 | def _get_whatToShow(self): |
| 325 | return self.whatToShow |
| 326 | |
| 327 | def acceptNode(self, element): |
| 328 | return self.FILTER_ACCEPT |
| 329 | |
| 330 | def startContainer(self, element): |
| 331 | return self.FILTER_ACCEPT |
| 332 | |
| 333 | del NodeFilter |
| 334 | |
| 335 | |
| 336 | class DocumentLS: |
| 337 | """Mixin to create documents that conform to the load/save spec.""" |
| 338 | |
| 339 | async = False |
| 340 | |
| 341 | def _get_async(self): |
| 342 | return False |
| 343 | def _set_async(self, async): |
| 344 | if async: |
| 345 | raise xml.dom.NotSupportedErr( |
| 346 | "asynchronous document loading is not supported") |
| 347 | |
| 348 | def abort(self): |
| 349 | # What does it mean to "clear" a document? Does the |
| 350 | # documentElement disappear? |
| 351 | raise NotImplementedError( |
| 352 | "haven't figured out what this means yet") |
| 353 | |
| 354 | def load(self, uri): |
| 355 | raise NotImplementedError("haven't written this yet") |
| 356 | |
| 357 | def loadXML(self, source): |
| 358 | raise NotImplementedError("haven't written this yet") |
| 359 | |
| 360 | def saveXML(self, snode): |
| 361 | if snode is None: |
| 362 | snode = self |
| 363 | elif snode.ownerDocument is not self: |
| 364 | raise xml.dom.WrongDocumentErr() |
| 365 | return snode.toxml() |
| 366 | |
| 367 | |
| 368 | class DOMImplementationLS: |
| 369 | MODE_SYNCHRONOUS = 1 |
| 370 | MODE_ASYNCHRONOUS = 2 |
| 371 | |
| 372 | def createDOMBuilder(self, mode, schemaType): |
| 373 | if schemaType is not None: |
| 374 | raise xml.dom.NotSupportedErr( |
| 375 | "schemaType not yet supported") |
| 376 | if mode == self.MODE_SYNCHRONOUS: |
| 377 | return DOMBuilder() |
| 378 | if mode == self.MODE_ASYNCHRONOUS: |
| 379 | raise xml.dom.NotSupportedErr( |
| 380 | "asynchronous builders are not supported") |
| 381 | raise ValueError("unknown value for mode") |
| 382 | |
| 383 | def createDOMWriter(self): |
| 384 | raise NotImplementedError( |
| 385 | "the writer interface hasn't been written yet!") |
| 386 | |
| 387 | def createDOMInputSource(self): |
| 388 | return DOMInputSource() |