def test_change_existing_schema(self):
# First we'll actually create a dataset
rev = self.create_rev()
input_schema = self.create_input_schema(rev = rev)
output_schema = input_schema.get_latest_output_schema()
(ok, job) = rev.apply(output_schema = output_schema)
assert ok, job
done = False
attempts = 0
while not done and attempts < 20:
(ok, job) = job.show()
attempts += 1
assert ok, job
done = job.attributes['status'] == 'successful'
sleep(0.5)
assert done, "Polling job never resulted in a successful completion: %s" % job
(ok, rev) = self.view.revisions.create_replace_revision()
assert ok, rev
self.rev.discard()
self.rev = rev
# Ok, we've got a dataset. Let's create a revision on it and mess with its schema!
(ok, source) = rev.source_from_dataset()
assert ok, source
input_schema = source.get_latest_input_schema()
output_schema = input_schema.get_latest_output_schema()
(ok, new_output_schema) = output_schema.add_column('d', 'D', 'a + 5')\
.drop_column('b')\
.drop_column('c')\
.run()
assert ok, new_output_schema
(ok, job) = rev.apply(output_schema = new_output_schema)
assert ok, job
done = False
attempts = 0
while not done and attempts < 20:
(ok, job) = job.show()
attempts += 1
assert ok, job
done = job.attributes['status'] == 'successful'
sleep(0.5)
assert done, "Polling job never resulted in a successful completion: %s" % job
from socrata.http import get
(ok, result) = get("https://{domain}/id/{fourfour}".format(domain = auth.domain, fourfour = rev.attributes['fourfour']), auth)
assert ok, result
self.assertEqual(result, [{'a': '1', 'd': '6'}, {'a': '2', 'd': '7'}, {'a': '3', 'd': '8'}, {'a': '4', 'd': '9'}])
def validate_row_id(self, uri, field_name):
"""
Set the row id. Note you must call `validate_row_id` before doing this.
Args:
```
field_name (str): The column to validate as the row id
```
Returns:
```
result (bool, dict): Returns an API Result; where the response says if it can be used as a row id
```
"""
output_column = [
oc for oc in self.attributes['output_columns']
if oc['field_name'] == field_name
]
if len(output_column):
[output_column] = output_column
transform_id = output_column['transform']['id']
return get(self.path(uri.format(transform_id=transform_id)),
auth=self.auth)
else:
return (False, {
"reason": "No column with field_name = %s" % field_name
})
def validate_row_id(self, uri, field_name):
"""
Set the row id. Note you must call `validate_row_id` before doing this.
Args:
```
field_name (str): The column to validate as the row id
```
Returns:
```
boolean
```
"""
output_column = [
oc for oc in self.attributes['output_columns']
if oc['field_name'] == field_name
]
if len(output_column):
[output_column] = output_column
transform_id = output_column['transform']['id']
return get(self.path(uri.format(transform_id=transform_id)),
auth=self.auth)['valid']
else:
return False
def lookup(self, name):
"""
Obtain a single ImportConfig by name
"""
return self._subresource(Config, get(
self.path() + '/' + name,
auth = self.auth
))
def list(self):
"""
List all the ImportConfigs on this domain
"""
return self._subresources(Config, get(
self.path(),
auth = self.auth
))
def show_input_schema(self, uri, input_schema_id):
(ok, res) = result = get(self.path(
uri.format(input_schema_id=input_schema_id)),
auth=self.auth)
if ok:
return self._subresource(InputSchema, result)
return result
def _get_rows(self, uri, offset, limit):
resp = get(
self.path(uri),
params = {'limit': limit, 'offset': offset},
auth = self.auth
)
rows = resp[1:]
return [self._munge_row(row) for row in rows]
def from_uri(cls, auth, uri):
path = 'https://{domain}{uri}'.format(
domain = auth.domain,
uri = uri
)
(ok, resp) = result = get(path, auth = auth)
if ok:
return (ok, cls(auth, resp))
return result
def plan(self, uri):
"""
Return the list of operations this revision will make when it is applied
Returns:
```
dict
```
"""
return pluck_resource(get(self.path(uri), auth=self.auth))
def list(self):
"""
List all the revisions on the view
Returns:
```
list[Revision]
```
"""
return self._subresources(Revision, get(self.path(), auth=self.auth))
def latest_output(self, uri):
"""
Get the latest (most recently created) OutputSchema
which descends from this InputSchema
Returns:
OutputSchema
"""
return self._subresource(OutputSchema, get(
self.path(uri),
auth = self.auth,
))
def plan(self, uri):
"""
Return the list of operations this revision will make when it is applied
Returns:
```
result (bool, dict): The revision plan
```
"""
(ok, resource) = result = get(self.path(uri), auth = self.auth)
if ok:
return (ok, resource['resource'])
return result
def latest_output(self, uri):
"""
Get the latest (most recently created) OutputSchema
which descends from this InputSchema
Returns:
result (bool, OutputSchema | dict): Returns an API Result; the new OutputSchema or an error response
"""
return self._subresource(OutputSchema,
get(
self.path(uri),
auth=self.auth,
))
def schema_errors_csv(self):
"""
Get the errors that results in transforming into this output schema
as a CSV stream.
Note that this returns a Reponse, where Reponse
is a python requests Reponse object
"""
return get(
self.path(self.schema_errors_uri),
auth = self.auth,
headers = {'accept': 'text/csv', 'content-type': 'text/csv'},
stream = True
)
def lookup(self, fourfour):
"""
Lookup the view by ID
Args:
```
fourfour (str): The view's identifier, ex: abcd-1234
```
Returns:
```
result (bool, Revision | dict): Returns an API Result; the View or an error response
```
"""
return self._subresource(
View, get(self.path() + '/' + fourfour, auth=self.auth))
def lookup(self, source_id):
"""
Lookup a source
Args:
```
source_id (int): The id
```
Returns:
```
Source: Returns the new Source The Source resulting from this API call, or an error
```
"""
return self._subresource(
Source, get(self.path() + '/' + str(source_id), auth=self.auth))
def lookup(self, revision_seq):
"""
Lookup a revision within the view based on the sequence number
Args:
```
revision_seq (int): The sequence number of the revision to lookup
```
Returns:
```
Revision The Revision resulting from this API call, or an error
```
"""
return self._subresource(
Revision, get(self.path() + '/' + str(revision_seq),
auth=self.auth))
def lookup(self, fourfour):
"""
Lookup the view by ID
Args:
```
fourfour (str): The view's identifier, ex: abcd-1234
```
Returns:
```
View
```
"""
return self._subresource(View, get(
self.path() + '/' + fourfour,
auth = self.auth
))
def from_uri(cls, auth, uri):
path = 'https://{domain}{uri}'.format(domain=auth.domain, uri=uri)
resp = get(path, auth=auth)
return cls(auth, resp)
def show_input_schema(self, uri, input_schema_id):
res = get(self.path(uri.format(input_schema_id=input_schema_id)),
auth=self.auth)
return self._subresource(InputSchema, res)
def list_sources(self, uri):
return self._subresources(Source, get(
self.path(uri),
auth = self.auth
))
def show(self, uri):
return self._mutate(get(self.path(uri), auth=self.auth))