def submission_fulltext_download(obj, eng):
submission_pdf = obj.extra_data.get('submission_pdf')
if submission_pdf and is_pdf_link(submission_pdf):
filename = secure_filename('fulltext.pdf')
pdf = download_file_to_workflow(
workflow=obj,
name=filename,
url=submission_pdf,
)
if pdf:
obj.data['documents'] = [
document for document in obj.data.get('documents', ())
if document.get('key') != filename
]
lb = LiteratureBuilder(source=obj.data['acquisition_source']['source'], record=obj.data)
lb.add_document(
filename,
fulltext=True,
original_url=submission_pdf,
url='/api/files/{bucket}/{key}'.format(bucket=obj.files[filename].bucket_id, key=filename)
)
obj.data = lb.record
obj.log.info('PDF provided by user from %s', submission_pdf)
return obj.files[filename].file.uri
else:
obj.log.info('Cannot fetch PDF provided by user from %s', submission_pdf)
def populate_arxiv_document(obj, eng):
arxiv_id = get_arxiv_id(obj.data)
url = current_app.config['ARXIV_PDF_URL'].format(arxiv_id=arxiv_id)
if not is_pdf_link(url):
if NO_PDF_ON_ARXIV in requests.get(url).content:
obj.log.info('No PDF is available for %s', arxiv_id)
return
raise DownloadError("{url} is not serving a PDF file.".format(url=url))
filename = secure_filename('{0}.pdf'.format(arxiv_id))
obj.data['documents'] = [
document for document in obj.data.get('documents', ())
if document.get('key') != filename
]
lb = LiteratureBuilder(source='arxiv', record=obj.data)
lb.add_document(
filename,
fulltext=True,
hidden=True,
material='preprint',
original_url=url,
url=url,
)
obj.data = lb.record
def populate_arxiv_document(obj, eng):
arxiv_id = LiteratureReader(obj.data).arxiv_id
for conf_name in ('ARXIV_PDF_URL', 'ARXIV_PDF_URL_ALTERNATIVE'):
url = current_app.config[conf_name].format(arxiv_id=arxiv_id)
is_valid_pdf_link = is_pdf_link(url)
if is_valid_pdf_link:
break
try:
if NO_PDF_ON_ARXIV in requests.get(url).content:
obj.log.info('No PDF is available for %s', arxiv_id)
return
except requests.exceptions.RequestException:
raise DownloadError("Error accessing url {url}".format(url=url))
if not is_valid_pdf_link:
raise DownloadError("{url} is not serving a PDF file.".format(url=url))
filename = secure_filename('{0}.pdf'.format(arxiv_id))
obj.data['documents'] = [
document for document in obj.data.get('documents', ())
if document.get('key') != filename
]
lb = LiteratureBuilder(source='arxiv', record=obj.data)
lb.add_document(
filename,
fulltext=True,
hidden=True,
material='preprint',
original_url=url,
url=url,
)
obj.data = lb.record
def populate_arxiv_document(obj, eng):
arxiv_id = get_arxiv_id(obj.data)
for conf_name in ('ARXIV_PDF_URL', 'ARXIV_PDF_URL_ALTERNATIVE'):
url = current_app.config[conf_name].format(arxiv_id=arxiv_id)
is_valid_pdf_link = is_pdf_link(url)
if is_valid_pdf_link:
break
if NO_PDF_ON_ARXIV in requests.get(url).content:
obj.log.info('No PDF is available for %s', arxiv_id)
return
if not is_valid_pdf_link:
raise DownloadError("{url} is not serving a PDF file.".format(url=url))
filename = secure_filename('{0}.pdf'.format(arxiv_id))
obj.data['documents'] = [
document for document in obj.data.get('documents', ())
if document.get('key') != filename
]
lb = LiteratureBuilder(source='arxiv', record=obj.data)
lb.add_document(
filename,
fulltext=True,
hidden=True,
material='preprint',
original_url=url,
url=url,
)
obj.data = lb.record
def add_document_or_figure(
self,
metadata,
stream=None,
is_document=True,
file_name=None,
key=None,
):
"""Add a document or figure to the record.
Args:
metadata(dict): metadata of the document or figure, see the schemas
for more details, will be validated.
stream(file like object): if passed, will extract the file contents
from it.
is_document(bool): if the given information is for a document,
set to ```False``` for a figure.
file_name(str): Name of the file, used as a basis of the key for
the files store.
key(str): if passed, will use this as the key for the files store
and ignore ``file_name``, use it to overwrite existing keys.
Returns:
dict: metadata of the added document or figure.
Raises:
TypeError: if not ``file_name`` nor ``key`` are passed (one of
them is required).
"""
if not key and not file_name:
raise TypeError(
'No file_name and no key passed, at least one of them is '
'needed.'
)
if not key:
key = self._get_unique_files_key(base_file_name=file_name)
if stream is not None:
self.files[key] = stream
builder = LiteratureBuilder(record=self.to_dict())
metadata['key'] = key
metadata['url'] = '/api/files/{bucket}/{key}'.format(
bucket=self.files[key].bucket_id,
key=key,
)
if is_document:
builder.add_document(**metadata)
else:
builder.add_figure(**metadata)
super(InspireRecord, self).update(builder.record)
return metadata
def add_document_or_figure(
self,
metadata,
stream=None,
is_document=True,
file_name=None,
key=None,
):
"""Add a document or figure to the record.
Args:
metadata(dict): metadata of the document or figure, see the schemas
for more details, will be validated.
stream(file like object): if passed, will extract the file contents
from it.
is_document(bool): if the given information is for a document,
set to ```False``` for a figure.
file_name(str): Name of the file, used as a basis of the key for
the files store.
key(str): if passed, will use this as the key for the files store
and ignore ``file_name``, use it to overwrite existing keys.
Returns:
dict: metadata of the added document or figure.
Raises:
TypeError: if not ``file_name`` nor ``key`` are passed (one of
them is required).
"""
if not key and not file_name:
raise TypeError(
'No file_name and no key passed, at least one of them is '
'needed.'
)
if not key:
key = self._get_unique_files_key(base_file_name=file_name)
if stream is not None:
self.files[key] = stream
builder = LiteratureBuilder(record=self.dumps())
metadata['key'] = key
metadata['url'] = '/api/files/{bucket}/{key}'.format(
bucket=self.files[key].bucket_id,
key=key,
)
if is_document:
builder.add_document(**metadata)
else:
builder.add_figure(**metadata)
super(InspireRecord, self).update(builder.record)
return metadata
def add_files(self, documents=None, figures=None):
"""Public method for adding documents and figures
Args:
documents (list[dict]): List of documents which should be added to this
record
figures (list[dict]): List of figures which should be added to this record
Documents and figures are lists of dicts.
Most obscure dict which whould be provided for each file is:
{
'url': 'http:// or /api/file/bucket_id/file_key'
'is_document': True or False(default)
}
Returns:
list: list of added keys
"""
if not documents and not figures:
raise TypeError("No files passed, at least one is needed")
if not current_app.config.get("FEATURE_FLAG_ENABLE_FILES", False):
if figures:
self.setdefault("figures", []).extend(figures)
if documents:
self.setdefault("documents", []).extend(documents)
return []
files = []
builder = LiteratureBuilder(record=self)
if documents:
doc_keys = [
doc_metadata["key"]
for doc_metadata in self.get("documents", [])
]
for doc in documents:
metadata = self._add_file(document=True, **doc)
if metadata["key"] not in doc_keys:
builder.add_document(**metadata)
files.append(metadata)
if figures:
fig_keys = [
fig_metadata["key"]
for fig_metadata in self.get("figures", [])
]
for fig in figures:
metadata = self._add_file(**fig)
if metadata["key"] not in fig_keys:
builder.add_figure(**metadata)
files.append(metadata)
# FIXME: this is wrong every time it goes to ``update``` function
# which means update refs, pidstore etc..
super().update(builder.record.dumps())
return files
def populate_submission_document(obj, eng):
submission_pdf = obj.extra_data.get('submission_pdf')
if submission_pdf and is_pdf_link(submission_pdf):
filename = secure_filename('fulltext.pdf')
obj.data['documents'] = [
document for document in obj.data.get('documents', ())
if document.get('key') != filename
]
lb = LiteratureBuilder(source=obj.data['acquisition_source']['source'],
record=obj.data)
lb.add_document(
filename,
fulltext=True,
url=submission_pdf,
original_url=submission_pdf,
)
obj.data = lb.record
def populate_submission_document(obj, eng):
submission_pdf = obj.extra_data.get('submission_pdf')
if submission_pdf and is_pdf_link(submission_pdf):
filename = secure_filename('fulltext.pdf')
obj.data['documents'] = [
document for document in obj.data.get('documents', ())
if document.get('key') != filename
]
lb = LiteratureBuilder(
source=obj.data['acquisition_source']['source'], record=obj.data)
lb.add_document(
filename,
fulltext=True,
url=submission_pdf,
original_url=submission_pdf,
)
obj.data = lb.record
def arxiv_fulltext_download(obj, eng):
"""Perform the fulltext download step for arXiv records.
:param obj: Workflow Object to process
:param eng: Workflow Engine processing the object
"""
arxiv_id = get_arxiv_id(obj.data)
filename = secure_filename('{0}.pdf'.format(arxiv_id))
url = current_app.config['ARXIV_PDF_URL'].format(arxiv_id=arxiv_id)
if not is_pdf_link(url):
if NO_PDF_ON_ARXIV in requests.get(url).content:
obj.log.info('No PDF is available for %s', arxiv_id)
return
raise DownloadError("{url} is not serving a PDF file.".format(url=url))
pdf = download_file_to_workflow(
workflow=obj,
name=filename,
url=url,
)
if pdf:
obj.data['documents'] = [
document for document in obj.data.get('documents', ())
if document.get('key') != filename
]
lb = LiteratureBuilder(source='arxiv', record=obj.data)
lb.add_document(filename,
fulltext=True,
hidden=True,
material='preprint',
original_url=url,
url='/api/files/{bucket}/{key}'.format(
bucket=obj.files[filename].bucket_id,
key=filename))
obj.data = lb.record
obj.log.info('PDF retrieved from arXiv for %s', arxiv_id)
else:
obj.log.error('Cannot retrieve PDF from arXiv for %s', arxiv_id)
