def update_metric(user, org, name_id):
m = fetch_by_id_or_field(Metric, 'name', name_id, org_id=org.id)
if not m:
raise NotFoundError(
'Metric "{}" does not yet exist for Org "{}"'
.format(name_id, org.name))
# get the request data
req_data = request_data()
# filter out any non-columns
columns = get_table_columns(Metric)
for k in req_data.keys():
if k not in columns:
req_data.pop(k)
# don't ever overwrite these:
for k in ['id', 'recipe_id', 'name', 'org_id', 'created', 'updated']:
if k in req_data:
req_data.pop(k, None)
# update fields
for k, v in req_data.items():
setattr(m, k, v)
try:
db.session.add(m)
db.session.commit()
except Exception as e:
raise RequestError("Error updating Metric: {}".format(e.message))
return jsonify(m)
def update_author(user, org, author_id):
"""
Update an author.
"""
a = fetch_by_id_or_field(Author, 'name', author_id,
org_id=org.id, transform='upper')
if not a:
raise NotFoundError(
'Author with ID/Name "{}" does not exist."'
.format(author_id))
req_data = request_data()
cols = get_table_columns(Author)
for k in req_data.keys():
if k not in cols or k in ['id', 'org_id']:
req_data.pop(k, None)
for k, v in req_data.items():
if k == 'name':
if v:
v = v.upper()
setattr(a, k, v)
db.session.add(a)
db.session.commit()
return jsonify(a)
def create_author(user, org):
"""
Create an author.
"""
req_data = request_data()
cols = get_table_columns(Author)
if 'name' not in req_data:
raise RequestError("A 'name' is required to create an Author.")
for k in req_data.keys():
if k not in cols or k in ['id', 'org_id']:
req_data.pop(k, None)
# upper-case.
elif k == 'name':
req_data[k] = req_data[k].upper()
a = Author(org_id=org.id, **req_data)
try:
db.session.add(a)
db.session.commit()
except Exception as e:
raise RequestError(
'There was an error creating this Author: {}'.format(e.message))
return jsonify(a)
def update_author(user, org, author_id):
"""
Update an author.
"""
a = Author.query\
.filter_by(id=author_id, org_id=org.id)\
.first()
if not a:
raise NotFoundError(
'Author with ID "{}" does not exist."'.format(author_id))
req_data = request_data()
cols = get_table_columns(Author)
for k in req_data.keys():
if k not in cols or k in ['id', 'org_id']:
req_data.pop(k, None)
for k, v in req_data.items():
setattr(a, k, v)
db.session.add(a)
db.session.commit()
return jsonify(a)
def update_author(user, org, author_id):
"""
Update an author.
"""
a = fetch_by_id_or_field(Author,
'name',
author_id,
org_id=org.id,
transform='upper')
if not a:
raise NotFoundError(
'Author with ID/Name "{}" does not exist."'.format(author_id))
req_data = request_data()
cols = get_table_columns(Author)
for k in req_data.keys():
if k not in cols or k in ['id', 'org_id']:
req_data.pop(k, None)
for k, v in req_data.items():
if k == 'name':
if v:
v = v.upper()
setattr(a, k, v)
db.session.add(a)
db.session.commit()
return jsonify(a)
def create_author(user, org):
"""
Create an author.
"""
req_data = request_data()
cols = get_table_columns(Author)
if 'name' not in req_data:
raise RequestError(
"A 'name' is required to create an Author.")
for k in req_data.keys():
if k not in cols or k in ['id', 'org_id']:
req_data.pop(k, None)
# upper-case.
elif k == 'name':
req_data[k] = req_data[k].upper()
a = Author(org_id=org.id, **req_data)
try:
db.session.add(a)
db.session.commit()
except Exception as e:
raise RequestError(
'There was an error creating this Author: {}'
.format(e.message))
return jsonify(a)
def update_author(user, org, author_id):
"""
Update an author.
"""
a = Author.query\
.filter_by(id=author_id, org_id=org.id)\
.first()
if not a:
raise NotFoundError(
'Author with ID "{}" does not exist."'
.format(author_id))
req_data = request_data()
cols = get_table_columns(Author)
for k in req_data.keys():
if k not in cols or k in ['id', 'org_id']:
req_data.pop(k, None)
for k, v in req_data.items():
setattr(a, k, v)
db.session.add(a)
db.session.commit()
return jsonify(a)
def update_tag(user, org, tag_id):
"""
Update an individual tag.
"""
# fetch the tag object
tag = fetch_by_id_or_field(Tag, 'slug', tag_id, org_id=org.id)
if not tag:
raise NotFoundError(
'A Tag with ID {} does not exist'
.format(tag_id))
# fetch the request data.
req_data = request_data()
# check hex code
if 'color' in req_data:
validate_hex_code(req_data['color'])
# check tag type
if 'type' in req_data:
validate_tag_types(req_data['type'])
# if tag type is "impact" ensure a proper category and
# level are included
if req_data['type'] == 'impact':
validate_tag_categories(req_data['category'])
validate_tag_levels(req_data['level'])
# check if levels + categories are being assigned to
# subject tags
if tag.type == 'subject':
if req_data.get('category') or req_data.get('level'):
raise RequestError(
'Categories and Levels can only be set for Impact Tags')
# set org id
req_data['org_id'] = org.id
# filter out non-table columns
columns = get_table_columns(Tag)
for k in req_data.keys():
if k not in columns:
req_data.pop(k)
# update attributes
for k, v in req_data.items():
setattr(tag, k, v)
db.session.add(tag)
# check for dupes
try:
db.session.commit()
except Exception as err:
raise RequestError(err.message)
return jsonify(tag)
def _clean():
__prepare = partial(
_prepare, requires=requires, recipe=recipe, org_id=org_id, type='event')
for obj in p.imap_unordered(__prepare, data):
# split out tags_ids + content_item_ids + links
meta[obj['source_id']] = dict(
tag_ids=obj.pop('tag_ids', obj.pop('tags', [])),
content_item_ids=obj.pop('content_item_ids', []),
links=obj.pop('links', [])
)
# split out meta fields
obj = _split_meta(obj, get_table_columns(Event))
# add to lookup
events[obj['source_id']] = obj
def validate_fields(obj, fields=[], incl=[], suffix='to select by'):
"""
check a list of fields against column names.
"""
columns = get_table_columns(obj, incl)
bad_fields = []
for field in fields:
if field not in columns:
bad_fields.append(field)
if len(bad_fields):
if len(bad_fields) == 1:
msg = 'is not a valid field name'
else:
msg = 'are not valid field names'
raise RequestError("'{}' {} {}. Choose from: {}.".format(
', '.join(bad_fields), msg, suffix, ", ".join(columns)))
def _clean():
__prepare = partial(_prepare,
requires=requires,
recipe=recipe,
org_id=org_id,
type='event')
for obj in p.imap_unordered(__prepare, data):
# split out tags_ids + content_item_ids + links
meta[obj['source_id']] = dict(
tag_ids=obj.pop('tag_ids', obj.pop('tags', [])),
content_item_ids=obj.pop('content_item_ids', []),
links=obj.pop('links', []))
# split out meta fields
obj = _split_meta(obj, get_table_columns(Event))
# add to lookup
events[obj['source_id']] = obj
def _clean():
__prepare = partial(_prepare, requires=requires, recipe=recipe,
org_id=org_id, type='content_item',
extract=kw.get('extract', True))
for obj in p.imap_unordered(__prepare, data):
# determine unique id.
uniqkey = "{}||{}".format(obj['url'], obj['type'])
# set metadata.
meta[uniqkey] = dict(
author_ids=obj.pop('author_ids', obj.pop('authors', [])),
tag_ids=obj.pop('tag_ids', obj.pop('tags', [])),
links=obj.pop('links', []),
)
# split out meta fields
obj = _split_meta(obj, get_table_columns(ContentItem))
cis[uniqkey] = obj
def validate_fields(obj, fields=[], incl=[], suffix='to select by'):
"""
check a list of fields against column names.
"""
columns = get_table_columns(obj, incl)
bad_fields = []
for field in fields:
if field not in columns:
bad_fields.append(field)
if len(bad_fields):
if len(bad_fields) == 1:
msg = 'is not a valid field name'
else:
msg = 'are not valid field names'
raise RequestError(
"'{}' {} {}. Choose from: {}."
.format(', '.join(bad_fields), msg, suffix, ", ".join(columns)))
def _clean():
__prepare = partial(_prepare,
requires=requires,
recipe=recipe,
org_id=org_id,
type='content_item',
extract=kw.get('extract', True))
for obj in p.imap_unordered(__prepare, data):
# determine unique id.
uniqkey = "{}||{}".format(obj['url'], obj['type'])
# set metadata.
meta[uniqkey] = dict(
author_ids=obj.pop('author_ids', obj.pop('authors', [])),
tag_ids=obj.pop('tag_ids', obj.pop('tags', [])),
links=obj.pop('links', []),
)
# split out meta fields
obj = _split_meta(obj, get_table_columns(ContentItem))
cis[uniqkey] = obj
def ingest(obj,
org_id,
url_fields=['body'],
requires=['url', 'type'],
extract=True,
kill_session=True):
"""
Ingest an Event.
"""
# distinct session for this eventlet.
session = gen_session()
# check required fields
ingest_util.check_requires(obj, requires, type='Content Item')
# validate type
validate_content_item_types(obj['type'])
# check if the org_id is in the body
# TODO: I don't think this is necessary.
org_id = obj.pop('org_id', org_id)
# get rid of ``id`` if it somehow got in here.
obj.pop('id', None)
# normalize the url
obj['url'] = ingest_util.prepare_url(obj, 'url')
# run article extraction.
if extract:
cache_response = extract_cache.get(url=obj['url'], type=obj['type'])
if not cache_response:
# make sure to kill this key.
extract_cache.invalidate(url=obj['url'], type=obj['type'])
raise RequestError(
'Extraction failed on {type} - {url}'.format(**obj))
# extraction succeeded
else:
data = cache_response.value
obj.update(data)
else:
obj['title'] = ingest_util.prepare_str(obj, 'title')
obj['description'] = ingest_util.prepare_str(obj, 'description')
obj['body'] = ingest_util.prepare_str(obj, 'body')
obj['created'] = ingest_util.prepare_str(obj, 'created')
if not obj['created']:
obj.pop('created')
# get thumbnail
obj['thumbnail'] = ingest_util.prepare_thumbnail(obj, 'img_url')
# split out tags_ids + authors + links
tag_ids = obj.pop('tag_ids', [])
authors = obj.pop('author_ids', [])
authors.extend(obj.pop('authors', [])) # accept names too
# links = obj.pop('links', {})
# determine event provenance
obj = _content_item_provenance(obj, org_id)
# split out meta fields
obj = ingest_util.split_meta(obj, get_table_columns(ContentItem))
# see if the event already exists.
c = session.query(ContentItem)\
.filter_by(org_id=org_id, type=obj['type'], url=obj['url'])\
.first()
# if not, create it
if not c:
# create event
c = ContentItem(org_id=org_id, **obj)
# else, update it
else:
for k, v in obj.items():
setattr(c, k, v)
# extract urls and normalize urls asynchronously.
# urls = ingest_util.extract_urls(
# obj,
# url_fields,
# source=data.get('url'),
# links=_links)
# detect content_items
# if len(_links):
# c = _associate_content_items(c, org_id, _links)
# associate tags
if len(tag_ids):
c = _associate_tags(c, org_id, tag_ids, session)
# associate tags
if len(authors):
_authors = _associate_authors(c, org_id, authors, session)
for a in _authors:
if a.id not in c.author_ids:
c.authors.append(a)
session.add(c)
session.commit()
if kill_session:
session.close()
return c
def event_update(user, org, event_id):
"""
Modify an individual event.
"""
e = Event.query\
.filter_by(id=event_id, org_id=org.id)\
.first()
if not e:
raise NotFoundError(
'An Event with ID {} does not exist.'
.format(event_id))
# get request data
req_data = request_data()
# fetch tag and thing
tag_ids = listify_data_arg('tag_ids')
content_item_ids = listify_data_arg('content_item_ids')
# a list of content items to apply impact tags to.
if len(tag_ids):
tags = Tag.query\
.filter_by(org_id=org.id)\
.filter(Tag.id.in_(tag_ids))\
.all()
if not len(tags):
raise RequestError(
'Tag(s) with ID(s) {} do(es) not exist.'
.format(tag_ids))
for tag in tags:
# validate tag
if tag.type != 'impact':
raise RequestError('Events can only be assigned Impact Tags.')
# add it
if tag.id not in e.tag_ids:
e.tags.append(tag)
if len(content_item_ids):
content_items = ContentItem.query\
.filter_by(org_id=org.id)\
.filter(ContentItem.id.in_(content_item_ids))\
.all()
if not len(content_items):
raise RequestError(
'ContentItem(s) with ID(s) {} do(es) not exist.'
.format(tag_ids))
# add content items
for c in content_items:
if c.id not in e.content_item_ids:
e.content_items.append(c)
# filter out any non-columns
columns = get_table_columns(Event)
for k in req_data.keys():
if k not in columns:
req_data.pop(k)
# update fields
for k, v in req_data.items():
setattr(e, k, v)
# ensure no one sneakily/accidentally
# updates an organization id
e.org_id = org.id
# update event metrics associated with these content item ids
if len(e.content_item_ids):
rollup_metric.content_summary_from_events(org, e.content_item_ids)
# commit changes
db.session.add(e)
db.session.commit()
# return modified event
return jsonify(e)
def ingest(
obj,
org_id,
org_domains,
url_fields=['title', 'body', 'description'],
requires=['title'],
must_link=False,
kill_session=True):
"""
Ingest an Event.
"""
# distinct session for this eventlet.
session = gen_session()
has_content_items = False
# check required fields
ingest_util.check_requires(obj, requires, type='Event')
# validate status
if 'status' in obj:
validate_event_status(obj['status'])
if obj['status'] == 'deleted':
raise RequestError(
'You cannot create an Event with status "deleted."')
# check if the org_id is in the body
# TODO: I don't think this is necessary.
org_id = obj.pop('org_id', org_id)
# get rid of ``id`` if it somehow got in here.
obj.pop('id', None)
# normalize the url
obj['url'] = ingest_util.prepare_url(obj, 'url')
# sanitize creation date
obj['created'] = ingest_util.prepare_date(obj, 'created')
if not obj['created']:
obj.pop('created')
# sanitize text/html fields
obj['title'] = ingest_util.prepare_str(obj, 'title', obj['url'])
obj['description'] = ingest_util.prepare_str(
obj, 'description', obj['url'])
obj['body'] = ingest_util.prepare_str(obj, 'body', obj['url'])
# get thumbnail
obj['thumbnail'] = ingest_util.prepare_thumbnail(obj, 'img_url')
# split out tags_ids + content_item_ids
tag_ids = obj.pop('tag_ids', [])
content_item_ids = obj.pop('content_item_ids', [])
links = obj.pop('links', [])
# determine event provenance
obj = _event_provenance(obj, org_id, session)
# split out meta fields
obj = ingest_util.split_meta(obj, get_table_columns(Event))
# see if the event already exists.
e = session.query(Event)\
.filter_by(org_id=org_id)\
.filter_by(source_id=obj['source_id'])\
.first()
# if not, create it
if not e:
# create event
e = Event(org_id=org_id, **obj)
# else, update it
else:
# if it's deleted, issue a message.
if e.status == 'deleted':
raise UnprocessableEntityError(
'Event {} already exists and has been previously deleted.'
.format(e.id))
for k, v in obj.items():
setattr(e, k, v)
# extract urls and normalize urls asynchronously.
links = ingest_util.prepare_links(links, org_domains)
# detect content_items
if len(links):
e, has_content_items = _associate_content_items(
e, org_id, links, content_item_ids, session)
# associate tags
if len(tag_ids):
e = _associate_tags(e, org_id, tag_ids, session)
# dont commit event if we're only looking
# for events that link to content_items
if not has_content_items and must_link:
return None
session.add(e)
session.commit()
if kill_session:
session.close()
return e
def event_update(user, org, event_id):
"""
Modify an individual event.
"""
e = Event.query\
.filter_by(id=event_id, org_id=org.id)\
.first()
if not e:
raise NotFoundError(
'An Event with ID {} does not exist.'.format(event_id))
# get request data
req_data = request_data()
# fetch tag and thing
tag_ids = listify_data_arg('tag_ids')
content_item_ids = listify_data_arg('content_item_ids')
# a list of content items to apply impact tags to.
if len(tag_ids):
tags = Tag.query\
.filter_by(org_id=org.id)\
.filter(Tag.id.in_(tag_ids))\
.all()
if not len(tags):
raise RequestError(
'Tag(s) with ID(s) {} do(es) not exist.'.format(tag_ids))
for tag in tags:
# validate tag
if tag.type != 'impact':
raise RequestError('Events can only be assigned Impact Tags.')
# add it
if tag.id not in e.tag_ids:
e.tags.append(tag)
if len(content_item_ids):
content_items = ContentItem.query\
.filter_by(org_id=org.id)\
.filter(ContentItem.id.in_(content_item_ids))\
.all()
if not len(content_items):
raise RequestError(
'ContentItem(s) with ID(s) {} do(es) not exist.'.format(
tag_ids))
# add content items
for c in content_items:
if c.id not in e.content_item_ids:
e.content_items.append(c)
# filter out any non-columns
columns = get_table_columns(Event)
for k in req_data.keys():
if k not in columns:
req_data.pop(k)
# update fields
for k, v in req_data.items():
setattr(e, k, v)
# ensure no one sneakily/accidentally
# updates an organization id
e.org_id = org.id
# commit changes
db.session.add(e)
db.session.commit()
# return modified event
return jsonify(e)
def ingest(
obj,
org_id,
url_fields=['body'],
requires=['url', 'type'],
extract=True,
kill_session=True):
"""
Ingest an Event.
"""
# distinct session for this eventlet.
session = gen_session()
# check required fields
ingest_util.check_requires(obj, requires, type='Content Item')
# validate type
validate_content_item_types(obj['type'])
# check if the org_id is in the body
# TODO: I don't think this is necessary.
org_id = obj.pop('org_id', org_id)
# get rid of ``id`` if it somehow got in here.
obj.pop('id', None)
# normalize the url
obj['url'] = ingest_util.prepare_url(obj, 'url')
# run article extraction.
if extract:
cache_response = extract_cache.get(url=obj['url'], type=obj['type'])
if not cache_response:
# make sure to kill this key.
extract_cache.invalidate(url=obj['url'], type=obj['type'])
raise RequestError(
'Extraction failed on {type} - {url}'
.format(**obj))
# extraction succeeded
else:
data = cache_response.value
obj.update(data)
else:
obj['title'] = ingest_util.prepare_str(obj, 'title')
obj['description'] = ingest_util.prepare_str(obj, 'description')
obj['body'] = ingest_util.prepare_str(obj, 'body')
obj['created'] = ingest_util.prepare_str(obj, 'created')
if not obj['created']:
obj.pop('created')
# get thumbnail
obj['thumbnail'] = ingest_util.prepare_thumbnail(obj, 'img_url')
# split out tags_ids + authors + links
tag_ids = obj.pop('tag_ids', [])
authors = obj.pop('author_ids', [])
authors.extend(obj.pop('authors', [])) # accept names too
# links = obj.pop('links', {})
# determine event provenance
obj = _content_item_provenance(obj, org_id)
# split out meta fields
obj = ingest_util.split_meta(obj, get_table_columns(ContentItem))
# see if the event already exists.
c = session.query(ContentItem)\
.filter_by(org_id=org_id, type=obj['type'], url=obj['url'])\
.first()
# if not, create it
if not c:
# create event
c = ContentItem(org_id=org_id, **obj)
# else, update it
else:
for k, v in obj.items():
setattr(c, k, v)
# extract urls and normalize urls asynchronously.
# urls = ingest_util.extract_urls(
# obj,
# url_fields,
# source=data.get('url'),
# links=_links)
# detect content_items
# if len(_links):
# c = _associate_content_items(c, org_id, _links)
# associate tags
if len(tag_ids):
c = _associate_tags(c, org_id, tag_ids, session)
# associate tags
if len(authors):
_authors = _associate_authors(c, org_id, authors, session)
for a in _authors:
if a.id not in c.author_ids:
c.authors.append(a)
session.add(c)
session.commit()
if kill_session:
session.close()
return c
def ingest(obj,
org_id,
org_domains,
url_fields=['title', 'body', 'description'],
requires=['title'],
must_link=False,
kill_session=True):
"""
Ingest an Event.
"""
# distinct session for this eventlet.
session = gen_session()
has_content_items = False
# check required fields
ingest_util.check_requires(obj, requires, type='Event')
# validate status
if 'status' in obj:
validate_event_status(obj['status'])
if obj['status'] == 'deleted':
raise RequestError(
'You cannot create an Event with status "deleted."')
# check if the org_id is in the body
# TODO: I don't think this is necessary.
org_id = obj.pop('org_id', org_id)
# get rid of ``id`` if it somehow got in here.
obj.pop('id', None)
# normalize the url
obj['url'] = ingest_util.prepare_url(obj, 'url')
# sanitize creation date
obj['created'] = ingest_util.prepare_date(obj, 'created')
if not obj['created']:
obj.pop('created')
# sanitize text/html fields
obj['title'] = ingest_util.prepare_str(obj, 'title', obj['url'])
obj['description'] = ingest_util.prepare_str(obj, 'description',
obj['url'])
obj['body'] = ingest_util.prepare_str(obj, 'body', obj['url'])
# get thumbnail
obj['thumbnail'] = ingest_util.prepare_thumbnail(obj, 'img_url')
# split out tags_ids + content_item_ids
tag_ids = obj.pop('tag_ids', [])
content_item_ids = obj.pop('content_item_ids', [])
links = obj.pop('links', [])
# determine event provenance
obj = _event_provenance(obj, org_id, session)
# split out meta fields
obj = ingest_util.split_meta(obj, get_table_columns(Event))
# see if the event already exists.
e = session.query(Event)\
.filter_by(org_id=org_id)\
.filter_by(source_id=obj['source_id'])\
.first()
# if not, create it
if not e:
# create event
e = Event(org_id=org_id, **obj)
# else, update it
else:
# if it's deleted, issue a message.
if e.status == 'deleted':
raise UnprocessableEntityError(
'Event {} already exists and has been previously deleted.'.
format(e.id))
for k, v in obj.items():
setattr(e, k, v)
# extract urls and normalize urls asynchronously.
links = ingest_util.prepare_links(links, org_domains)
# detect content_items
if len(links):
e, has_content_items = _associate_content_items(
e, org_id, links, content_item_ids, session)
# associate tags
if len(tag_ids):
e = _associate_tags(e, org_id, tag_ids, session)
# dont commit event if we're only looking
# for events that link to content_items
if not has_content_items and must_link:
return None
session.add(e)
session.commit()
if kill_session:
session.close()
return e