async def _fetch_document_from_folder_path_async(
self, document_path: str, adapter: 'StorageAdapterBase',
force_reload: bool) -> 'CdmDocumentDefinition':
"""Gets the document from folder path.
arguments:
path: The path.
adapter: The storage adapter where the document can be found."""
from .cdm_corpus_def import FOLIO_EXTENSION, MANIFEST_EXTENSION, MODEL_JSON_EXTENSION
doc_name = None
first = document_path.find('/')
if first < 0:
doc_name = document_path
else:
doc_name = document_path[0:first]
if doc_name in self._document_lookup and not force_reload:
return self._document_lookup[doc_name]
is_cdm_folder = doc_name.endswith(
FOLIO_EXTENSION) or doc_name.endswith(MANIFEST_EXTENSION)
is_model_json = doc_name.endswith(MODEL_JSON_EXTENSION)
# got that doc?
doc = None
# go get the doc
doc_path = '{}{}'.format(self.folder_path, doc_name)
json_data = None
fs_modified_time = None
try:
if adapter.can_read():
json_data = await adapter.read_async(doc_path)
fs_modified_time = await adapter.compute_last_modified_time_async(
adapter.create_adapter_path(doc_path))
except Exception as e:
self.corpus.ctx.logger.exception(
'Could not read %s from the \'%s\' namespace.', doc_path,
self.namespace)
return None
if is_cdm_folder:
from cdm.persistence.cdmfolder import ManifestPersistence
from cdm.persistence.cdmfolder.types import ManifestContent
manifest = ManifestContent()
manifest.decode(json_data)
doc = await ManifestPersistence.from_data(self.ctx, doc_name,
self.namespace,
self.folder_path,
manifest)
doc.folder = self
self.documents.append(doc)
self.corpus._add_document_objects(self, doc)
self._document_lookup[doc_name] = doc
elif is_model_json:
from cdm.persistence.modeljson import ManifestPersistence
from cdm.persistence.modeljson.types import Model
model = Model()
model.decode(json_data)
doc = await ManifestPersistence.from_data(self.ctx, model, self)
doc.folder = self
doc.folder_path = self.folder_path
else:
from cdm.persistence.cdmfolder import DocumentPersistence
from cdm.persistence.cdmfolder.types import DocumentContent
document = DocumentContent()
document.decode(json_data)
doc = self.documents.append(await DocumentPersistence.from_data(
self.ctx, doc_name, self.namespace, self.folder_path,
document))
doc._file_system_modified_time = fs_modified_time
doc._is_dirty = False
return doc
async def _load_document_from_path_async(self, folder: 'CdmFolderDefinition', doc_name: str,
doc_container: 'CdmDocumentDefinition',
res_opt: Optional[ResolveOptions] = None) \
-> 'CdmDocumentDefinition':
# go get the doc
doc_content = None # type: Optional[CdmDocumentDefinition]
json_data = None
fs_modified_time = None
doc_path = folder.folder_path + doc_name
adapter = self._ctx.corpus.storage.fetch_adapter(
folder.namespace) # type: StorageAdapter
try:
if adapter.can_read():
# log message used by navigator, do not change or remove
logger.debug(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path, 'request file: {}'.format(doc_path))
json_data = await adapter.read_async(doc_path)
# log message used by navigator, do not change or remove
logger.debug(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path, 'received file: {}'.format(doc_path))
else:
raise Exception('Storage Adapter is not enabled to read.')
except Exception as e:
# log message used by navigator, do not change or remove
logger.debug(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path, 'fail file: {}'.format(doc_path))
# when shallow validation is enabled, log messages about being unable to find referenced documents as warnings instead of errors.
if res_opt and res_opt.shallow_validation:
logger.warning(
self._ctx, self._TAG,
PersistenceLayer._load_document_from_path_async.__name__,
doc_path, CdmLogCode.WARN_PERSIST_FILE_READ_FAILURE,
doc_path, folder.Namespace, e)
else:
logger.error(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path,
CdmLogCode.ERR_PERSIST_FILE_READ_FAILURE,
doc_path, str(folder.namespace), e)
return None
try:
fs_modified_time = await adapter.compute_last_modified_time_async(
doc_path)
except Exception as e:
logger.warning(
self._ctx, self._TAG,
PersistenceLayer._load_document_from_path_async.__name__,
doc_path, CdmLogCode.WARN_PERSIST_FILE_MOD_COMPUTE_FAILED,
e.Message)
if not doc_name:
logger.error(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path, CdmLogCode.ERR_PERSIST_NULL_DOC_NAME)
return None
doc_name_lower = doc_name.lower()
# If loading an model.json file, check that it is named correctly.
if doc_name_lower.endswith(
self.MODEL_JSON_EXTENSION
) and not doc_name.lower() == self.MODEL_JSON_EXTENSION:
logger.error(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path,
CdmLogCode.ERR_PERSIST_DOC_NAME_LOAD_FAILURE,
doc_name, self.MODEL_JSON_EXTENSION)
return None
try:
if doc_name_lower.endswith(PersistenceLayer.MANIFEST_EXTENSION
) or doc_name_lower.endswith(
PersistenceLayer.FOLIO_EXTENSION):
from cdm.persistence.cdmfolder import ManifestPersistence
from cdm.persistence.cdmfolder.types import ManifestContent
manifest = ManifestContent()
manifest.decode(json_data)
doc_content = ManifestPersistence.from_object(
self._ctx, doc_name, folder.namespace, folder.folder_path,
manifest)
elif doc_name_lower.endswith(
PersistenceLayer.MODEL_JSON_EXTENSION):
if doc_name_lower != PersistenceLayer.MODEL_JSON_EXTENSION:
logger.error(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path,
CdmLogCode.ERR_PERSIST_DOC_NAME_LOAD_FAILURE,
doc_name, self.MODEL_JSON_EXTENSION)
return None
from cdm.persistence.modeljson import ManifestPersistence
from cdm.persistence.modeljson.types import Model
model = Model()
model.decode(json_data)
doc_content = await ManifestPersistence.from_object(
self._ctx, model, folder)
elif doc_name_lower.endswith(PersistenceLayer.CDM_EXTENSION):
from cdm.persistence.cdmfolder import DocumentPersistence
from cdm.persistence.cdmfolder.types import DocumentContent
document = DocumentContent()
document.decode(json_data)
doc_content = DocumentPersistence.from_object(
self._ctx, doc_name, folder.namespace, folder.folder_path,
document)
else:
# Could not find a registered persistence class to handle this document type.
logger.error(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path, CdmLogCode.ERR_PERSIST_CLASS_MISSING,
doc_name)
return None
except Exception as e:
logger.error(self._ctx, self._TAG,
self._load_document_from_path_async.__name__,
doc_path,
CdmLogCode.ERR_PERSIST_DOC_CONVERSION_FAILURE,
doc_path, e)
return None
# add document to the folder, this sets all the folder/path things, caches name to content association and may trigger indexing on content
if doc_content is not None:
if doc_container:
# there are situations where a previously loaded document must be re-loaded.
# the end of that chain of work is here where the old version of the document has been removed from
# the corpus and we have created a new document and loaded it from storage and after this call we will probably
# add it to the corpus and index it, etc.
# it would be really rude to just kill that old object and replace it with this replicant, especially because
# the caller has no idea this happened. so... sigh ... instead of returning the new object return the one that
# was just killed off but make it contain everything the new document loaded.
doc_content = doc_content.copy(
ResolveOptions(wrt_doc=doc_container,
directives=self._ctx.corpus.
default_resolution_directives),
doc_container)
folder.documents.append(doc_content, doc_name)
doc_content._file_system_modified_time = fs_modified_time
doc_content._is_dirty = False
return doc_content
async def load_document_from_path_async(
folder: 'CdmFolderDefinition', doc_name: str,
doc_container: 'CdmDocumentDefinition') -> 'CdmDocumentDefinition':
is_cdm_folder = doc_name.lower().endswith(
CdmCorpusDefinition._FOLIO_EXTENSION) or doc_name.lower().endswith(
CdmCorpusDefinition._MANIFEST_EXTENSION)
is_model_json = doc_name.lower().endswith(
CdmCorpusDefinition._MODEL_JSON_EXTENSION)
# go get the doc
doc_content = None # type: Optional[CdmDocumentDefinition]
json_data = None
fs_modified_time = None
ctx = folder.ctx
doc_path = folder.folder_path + doc_name
adapter = ctx.corpus.storage.fetch_adapter(
folder.namespace) # type: StorageAdapter
try:
if adapter.can_read():
json_data = await adapter.read_async(doc_path)
fs_modified_time = await adapter.compute_last_modified_time_async(
adapter.create_adapter_path(doc_path))
ctx.logger.info('read file: {}'.format(doc_path))
except Exception as e:
ctx.logger.error(
'could not read %s from the \'%s\' namespace.\n Reason: \n%s',
doc_path, folder.namespace, e)
return None
try:
if is_cdm_folder:
from cdm.persistence.cdmfolder import ManifestPersistence
from cdm.persistence.cdmfolder.types import ManifestContent
manifest = ManifestContent()
manifest.decode(json_data)
doc_content = await ManifestPersistence.from_data(
ctx, doc_name, folder.namespace, folder.folder_path, manifest)
elif is_model_json:
if doc_name.lower() != CdmCorpusDefinition._MODEL_JSON_EXTENSION:
ctx.logger.error(
'Failed to load \'{}\', as it\'s not an acceptable filename. It must be model.json'
.format(doc_name))
return None
from cdm.persistence.modeljson import ManifestPersistence
from cdm.persistence.modeljson.types import Model
model = Model()
model.decode(json_data)
doc_content = await ManifestPersistence.from_data(
ctx, model, folder)
else:
from cdm.persistence.cdmfolder import DocumentPersistence
from cdm.persistence.cdmfolder.types import DocumentContent
document = DocumentContent()
document.decode(json_data)
doc_content = await DocumentPersistence.from_data(
ctx, doc_name, folder.namespace, folder.folder_path, document)
except Exception as e:
ctx.logger.error('Could not convert \'{}\'. Reason \'{}\''.format(
doc_path, e))
return None
# add document to the folder, this sets all the folder/path things, caches name to content association and may trigger indexing on content
if doc_content is not None:
if doc_container:
# there are situations where a previously loaded document must be re-loaded.
# the end of that chain of work is here where the old version of the document has been removed from
# the corpus and we have created a new document and loaded it from storage and after this call we will probably
# add it to the corpus and index it, etc.
# it would be really rude to just kill that old object and replace it with this replicant, especially because
# the caller has no idea this happened. so... sigh ... instead of returning the new object return the one that
# was just killed off but make it contain everything the new document loaded.
doc_content = doc_content.copy(
ResolveOptions(wrt_doc=doc_container), doc_container)
folder.documents.append(doc_content, doc_name)
doc_content._file_system_modified_time = fs_modified_time
doc_content._is_dirty = False
return doc_content