hestia-earth-utils 0.16.9__py3-none-any.whl → 0.16.10__py3-none-any.whl

hestia_earth/utils/pipeline.py

@@ -5,14 +5,19 @@ import numpy as np
 from .tools import current_time_ms, non_empty_list, flatten
 from .api import find_related
 from .storage._s3_client import (
-    _load_from_bucket, _upload_to_bucket, _last_modified, _read_metadata, _update_metadata, _exists_in_bucket
+    _load_from_bucket,
+    _upload_to_bucket,
+    _last_modified,
+    _read_metadata,
+    _update_metadata,
+    _exists_in_bucket,
 )
 from .storage._sns_client import _get_sns_client
 
-PROGRESS_EXT = '.progress'
-CALC_FOLDER = 'recalculated'
-METADATA_STAGE_KEY = 'stage'
-METADATA_PROGRESS_KEY = 'calculating'
+PROGRESS_EXT = ".progress"
+CALC_FOLDER = "recalculated"
+METADATA_STAGE_KEY = "stage"
+METADATA_PROGRESS_KEY = "calculating"
 
 
 # fix error "Object of type int64 is not JSON serializable"
@@ -27,10 +32,12 @@ class NpEncoder(json.JSONEncoder):
         return super(NpEncoder, self).default(obj)
 
 
-def to_string(data: dict, indent: int = None): return json.dumps(data, indent=indent, ensure_ascii=False, cls=NpEncoder)
+def to_string(data: dict, indent: int = None):
+    return json.dumps(data, indent=indent, ensure_ascii=False, cls=NpEncoder)
 
 
-def to_bytes(data: dict): return to_string(data).encode('utf8')
+def to_bytes(data: dict):
+    return to_string(data).encode("utf8")
 
 
 def upload_json(bucket_name: str, file_key: str, body: dict):
@@ -38,15 +45,21 @@ def upload_json(bucket_name: str, file_key: str, body: dict):
         bucket=bucket_name,
         key=file_key,
         body=to_bytes(body),
-        content_type='application/json'
+        content_type="application/json",
     )
 
 
 def _to_file_progress(filepath: str):
-    return filepath.replace('.csv', PROGRESS_EXT).replace('.json', PROGRESS_EXT).replace('.hestia', PROGRESS_EXT)
+    return (
+        filepath.replace(".csv", PROGRESS_EXT)
+        .replace(".json", PROGRESS_EXT)
+        .replace(".hestia", PROGRESS_EXT)
+    )
 
 
-def handle_result(bucket_name: str, file_key: str, step: str, start: int, content: dict):
+def handle_result(
+    bucket_name: str, file_key: str, step: str, start: int, content: dict
+):
     filepath = _to_file_progress(file_key)
 
     # try to read existing progress to update the time per step
@@ -55,89 +68,122 @@ def handle_result(bucket_name: str, file_key: str, step: str, start: int, conten
     except Exception:
         data = {}
 
-    return upload_json(bucket_name, filepath, {
-        **data,
-        'step': step,
-        'time': {
-            **(data.get('time', {}) if isinstance(data.get('time', {}), dict) else {}),
-            step: current_time_ms() - start
+    return upload_json(
+        bucket_name,
+        filepath,
+        {
+            **data,
+            "step": step,
+            "time": {
+                **(
+                    data.get("time", {})
+                    if isinstance(data.get("time", {}), dict)
+                    else {}
+                ),
+                step: current_time_ms() - start,
+            },
+            **content,
         },
-        **content
-    })
+    )
 
 
 def handle_error(
-    bucket_name: str, file_key: str, step: str, start: int,
-    err: str = '', stack: str = '', errors=[], warnings=[],
-    extras: dict = {}
+    bucket_name: str,
+    file_key: str,
+    step: str,
+    start: int,
+    err: str = "",
+    stack: str = "",
+    errors=[],
+    warnings=[],
+    extras: dict = {},
 ):
-    return handle_result(bucket_name, file_key, step, start, extras | {
-        'success': False,
-        'error': {
-            'message': err,
-            'stack': stack,
-            'errors': errors,
-            'warnings': warnings
+    return handle_result(
+        bucket_name,
+        file_key,
+        step,
+        start,
+        extras
+        | {
+            "success": False,
+            "error": {
+                "message": err,
+                "stack": stack,
+                "errors": errors,
+                "warnings": warnings,
+            },
+            "warning": {"warnings": warnings},
         },
-        'warning': {
-            'warnings': warnings
-        }
-    })
+    )
 
 
-def handle_success(bucket_name: str, file_key: str, step: str, start: int, extras: dict = {}):
-    return handle_result(bucket_name, file_key, step, start, extras | {'success': True})
+def handle_success(
+    bucket_name: str, file_key: str, step: str, start: int, extras: dict = {}
+):
+    return handle_result(bucket_name, file_key, step, start, extras | {"success": True})
 
 
-def publish_result(topic_arn: str, bucket_name: str, file_key: str, filepath: str, step: str, success: bool):
+def publish_result(
+    topic_arn: str,
+    bucket_name: str,
+    file_key: str,
+    filepath: str,
+    step: str,
+    success: bool,
+):
     return _get_sns_client().publish(
         TopicArn=topic_arn,
-        Message=to_string({
-            'bucket': bucket_name,
-            'key': file_key,
-            'filepath': filepath
-        }),
+        Message=to_string(
+            {"bucket": bucket_name, "key": file_key, "filepath": filepath}
+        ),
         MessageAttributes={
-            'functionName': {
-                'DataType': 'String',
-                'StringValue': step + ('Done' if success else 'Error')
+            "functionName": {
+                "DataType": "String",
+                "StringValue": step + ("Done" if success else "Error"),
             }
-        }
+        },
     )
 
 
 def _parse_event_s3(event: dict):
-    return {'bucket': event['s3']['bucket']['name'], 'key': event['s3']['object']['key']}
+    return {
+        "bucket": event["s3"]["bucket"]["name"],
+        "key": event["s3"]["object"]["key"],
+    }
 
 
 def _parse_event_SNS(event: dict):
-    event = event.get('Sns', {})
-    data = json.loads(event.get('Message', '{}'))
-    attributes: dict = event.get('MessageAttributes', {})
-    data['attributes'] = {key: value.get('Value') for key, value in attributes.items()}
+    event = event.get("Sns", {})
+    data = json.loads(event.get("Message", "{}"))
+    attributes: dict = event.get("MessageAttributes", {})
+    data["attributes"] = {key: value.get("Value") for key, value in attributes.items()}
     return data
 
 
 def _parse_event_SQS(event: dict):
-    condition = event.get('requestContext', {}).get('condition')
-    return _get_data_from_event(event.get('requestPayload', {})) if condition != 'RetriesExhausted' else None
+    condition = event.get("requestContext", {}).get("condition")
+    return (
+        _get_data_from_event(event.get("requestPayload", {}))
+        if condition != "RetriesExhausted"
+        else None
+    )
 
 
 def _get_data_from_event(event):  # noqa: C901
     if isinstance(event, dict):
-        if 's3' in event:
+        if "s3" in event:
             return _parse_event_s3(event)
         # invoked when running asynchronously
-        if 'Sns' in event:
+        if "Sns" in event:
             return _parse_event_SNS(event)
         # invoked through http event
-        if 'body' in event:
-            return _get_data_from_event(json.loads(event.get('body', '{}')))
+        if "body" in event:
+            return _get_data_from_event(json.loads(event.get("body", "{}")))
         # invoked through s3 put object
-        if 'Records' in event:
-            return flatten(map(_get_data_from_event, event.get('Records', [])))
+        if "Records" in event:
+            return flatten(map(_get_data_from_event, event.get("Records", [])))
         # invoked when calculation timedout or failed
-        if 'requestPayload' in event:
+        if "requestPayload" in event:
             return _parse_event_SQS(event)
         return event
     if isinstance(event, str):
@@ -149,34 +195,51 @@ def parse_event(event: dict):
     return non_empty_list(flatten(data) if isinstance(data, list) else [data])
 
 
-def _node_type(node: dict): return node.get('@type', node.get('type'))
+def _node_type(node: dict):
+    return node.get("@type", node.get("type"))
 
 
-def _node_id(node: dict): return node.get('@id', node.get('id'))
+def _node_id(node: dict):
+    return node.get("@id", node.get("id"))
 
 
-def _node_path(node: dict, folder: str = ''): return join(folder, _node_type(node), f"{_node_id(node)}.jsonld")
+def _node_path(node: dict, folder: str = ""):
+    return join(folder, _node_type(node), f"{_node_id(node)}.jsonld")
 
 
-def _load_node(bucket: str, file_key: str): return json.loads(_load_from_bucket(bucket, file_key))
+def _load_node(bucket: str, file_key: str):
+    return json.loads(_load_from_bucket(bucket, file_key))
 
 
-def _cache_path(node: dict): return join(_node_type(node), f"{_node_id(node)}.cache")
+def _cache_path(node: dict):
+    return join(_node_type(node), f"{_node_id(node)}.cache")
 
 
-def _has_cache(bucket: str, node: dict): return _exists_in_bucket(bucket, _cache_path(node))
+def _has_cache(bucket: str, node: dict):
+    return _exists_in_bucket(bucket, _cache_path(node))
 
 
-def is_calculating(bucket: str, node: dict, folder: str = ''):
-    return _read_metadata(bucket, _node_path(node, folder)).get(METADATA_PROGRESS_KEY, 'false') == 'true'
+def is_calculating(bucket: str, node: dict, folder: str = ""):
+    return (
+        _read_metadata(bucket, _node_path(node, folder)).get(
+            METADATA_PROGRESS_KEY, "false"
+        )
+        == "true"
+    )
 
 
-def set_calculating(bucket: str, node: dict, in_progress: bool, folder: str = ''):
-    return _update_metadata(bucket, _node_path(node, folder), {METADATA_PROGRESS_KEY: str(in_progress).lower()})
+def set_calculating(bucket: str, node: dict, in_progress: bool, folder: str = ""):
+    return _update_metadata(
+        bucket,
+        _node_path(node, folder),
+        {METADATA_PROGRESS_KEY: str(in_progress).lower()},
+    )
 
 
 def get_stage(bucket: str, node: dict, folder: str = CALC_FOLDER):
-    stage = _read_metadata(bucket, _node_path(node, folder=CALC_FOLDER)).get(METADATA_STAGE_KEY)
+    stage = _read_metadata(bucket, _node_path(node, folder=CALC_FOLDER)).get(
+        METADATA_STAGE_KEY
+    )
     return int(stage) if stage else stage
 
 
@@ -196,29 +259,43 @@ def load_cache(bucket: str, node: dict):
     dict
         The cached data.
     """
-    cache_path = join(node['@type'], f"{node['@id']}.cache")
+    cache_path = join(node["@type"], f"{node['@id']}.cache")
     try:
         return json.loads(_load_from_bucket(bucket, cache_path))
     except Exception:
-        print('No cache found for', cache_path)
+        print("No cache found for", cache_path)
         return {}
 
 
-def _filter_by_type(nodes: list, type: str): return [n for n in nodes if n.get('@type', n.get('type')) == type]
+def _filter_by_type(nodes: list, type: str):
+    return [n for n in nodes if n.get("@type", n.get("type")) == type]
 
 
-def _find_related_nodes(from_type: str, from_id: str, related_type: str, related_key: str):
-    should_find_related = related_key == 'related'
-    print('Find related nodes from API', from_type, from_id, related_key, related_type)
-    return find_related(from_type, from_id, related_type, limit=10000) if should_find_related else []
+def _find_related_nodes(
+    from_type: str, from_id: str, related_type: str, related_key: str
+):
+    should_find_related = related_key == "related"
+    print("Find related nodes from API", from_type, from_id, related_key, related_type)
+    return (
+        find_related(from_type, from_id, related_type, limit=10000)
+        if should_find_related
+        else []
+    )
 
 
-def _get_cached_nodes(cache: dict, related_key: str, from_type: str, from_id: str, to_type: str):
+def _get_cached_nodes(
+    cache: dict, related_key: str, from_type: str, from_id: str, to_type: str
+):
     # if key is in cache, use nodes in cache, otherwise use API
     if related_key in cache:
         nodes = _filter_by_type(cache.get(related_key, []), to_type)
-        print('Using cached data to', related_key, to_type, nodes)
-        return list(map(lambda node: {'@type': to_type, '@id': node.get('@id', node.get('id'))}, nodes))
+        print("Using cached data to", related_key, to_type, nodes)
+        return list(
+            map(
+                lambda node: {"@type": to_type, "@id": node.get("@id", node.get("id"))},
+                nodes,
+            )
+        )
     else:
         return _find_related_nodes(from_type, from_id, to_type, related_key)
 
@@ -244,15 +321,22 @@ def get_related_nodes(node: dict, related_key: str, related_type: str, cache: di
     List[dict]
         The related nodes.
     """
-    from_type = node.get('@type', node.get('type'))
-    from_id = node.get('@id', node.get('id'))
+    from_type = node.get("@type", node.get("type"))
+    from_id = node.get("@id", node.get("id"))
 
-    related_nodes = _get_cached_nodes(cache or {}, related_key, from_type, from_id, related_type) or []
+    related_nodes = (
+        _get_cached_nodes(cache or {}, related_key, from_type, from_id, related_type)
+        or []
+    )
 
-    return list({f"{node['@type']}/{node['@id']}": node for node in related_nodes}.values())
+    return list(
+        {f"{node['@type']}/{node['@id']}": node for node in related_nodes}.values()
+    )
 
 
-def get_related_nodes_data(bucket_name: str, node: dict, related_key: str, related_type: str, cache: dict):
+def get_related_nodes_data(
+    bucket_name: str, node: dict, related_key: str, related_type: str, cache: dict
+):
     """
     Given a node, return all related nodes with extra data.
 
@@ -275,14 +359,19 @@ def get_related_nodes_data(bucket_name: str, node: dict, related_key: str, relat
     List[dict]
         The related nodes with extra data: `indexed_at`, `recalculated_at` and `recalculated_stage`.
     """
-    related_nodes = get_related_nodes(node=node, related_key=related_key, related_type=related_type, cache=cache)
+    related_nodes = get_related_nodes(
+        node=node, related_key=related_key, related_type=related_type, cache=cache
+    )
 
     return [
-        node | {
-            'indexed_at': _last_modified(bucket=bucket_name, key=_node_path(node)),
-            'recalculated_at': _last_modified(bucket=bucket_name, key=_node_path(node, folder=CALC_FOLDER)),
-            'recalculated_stage': get_stage(bucket_name, node),
-            'is_calculating': is_calculating(bucket_name, node)
+        node
+        | {
+            "indexed_at": _last_modified(bucket=bucket_name, key=_node_path(node)),
+            "recalculated_at": _last_modified(
+                bucket=bucket_name, key=_node_path(node, folder=CALC_FOLDER)
+            ),
+            "recalculated_stage": get_stage(bucket_name, node),
+            "is_calculating": is_calculating(bucket_name, node),
         }
         for node in related_nodes
     ]

hestia_earth/utils/pivot/_shared.py

@@ -5,10 +5,12 @@ from hestia_earth.schema import EmissionMethodTier
 
 EXCLUDE_FIELDS = ["@type", "type", "@context"]
 EXCLUDE_PRIVATE_FIELDS = [
-    "added", "addedVersion",
-    "updated", "updatedVersion",
+    "added",
+    "addedVersion",
+    "updated",
+    "updatedVersion",
     "aggregatedVersion",
-    "_cache"
+    "_cache",
 ]
 
 
@@ -17,11 +19,11 @@ def _with_csv_formatting(dct):
     Use as object_hook when parsing a JSON node: json.loads(node, object_hook=_with_csv_formatting).
     Ensures parsed JSON has field values formatted according to hestia csv conventions.
     """
-    if 'boundary' in dct:
-        dct['boundary'] = json.dumps(dct['boundary'])
+    if "boundary" in dct:
+        dct["boundary"] = json.dumps(dct["boundary"])
     for key, value in dct.items():
         if _is_scalar_list(value):
-            dct[key] = ';'.join([str(el) for el in value])
+            dct[key] = ";".join([str(el) for el in value])
     return dct
 
 
@@ -37,15 +39,17 @@ def _is_scalar_list(value):
 
 
 def _filter_not_relevant(blank_node: dict):
-    return blank_node.get('methodTier') != EmissionMethodTier.NOT_RELEVANT.value
+    return blank_node.get("methodTier") != EmissionMethodTier.NOT_RELEVANT.value
 
 
 def _filter_emissions_not_relevant(node: dict):
     """
     Ignore all emissions where `methodTier=not relevant` so save space.
     """
-    return node | ({
-        key: list(filter(_filter_not_relevant, node[key]))
-        for key in ['emissions', 'emissionsResourceUse']
-        if key in node
-    })
+    return node | (
+        {
+            key: list(filter(_filter_not_relevant, node[key]))
+            for key in ["emissions", "emissionsResourceUse"]
+            if key in node
+        }
+    )

hestia_earth/utils/pivot/pivot_csv.py

@@ -9,7 +9,12 @@ from flatten_json import flatten as flatten_json
 
 # __package__ = "hestia_earth.utils" # required to run interactively in vscode
 from ..api import find_term_ids_by_names
-from ._shared import EXCLUDE_FIELDS, EXCLUDE_PRIVATE_FIELDS, _with_csv_formatting, _filter_emissions_not_relevant
+from ._shared import (
+    EXCLUDE_FIELDS,
+    EXCLUDE_PRIVATE_FIELDS,
+    _with_csv_formatting,
+    _filter_emissions_not_relevant,
+)
 
 
 # We only want to pivot array items containing blank nodes
@@ -18,11 +23,13 @@ def _get_blank_node_uniqueness_fields():
     filtered_uniqueness_fields = copy.deepcopy(UNIQUENESS_FIELDS)
     for node_type, array_fields in UNIQUENESS_FIELDS.items():
         for array_field in array_fields.keys():
-            if SORT_CONFIG[node_type][array_field]['type'] in NODE_TYPES:
+            if SORT_CONFIG[node_type][array_field]["type"] in NODE_TYPES:
                 del filtered_uniqueness_fields[node_type][array_field]
             # include `impactAssessment.@id` since it is not part of original uniqueness
-            if 'impactAssessment.id' in array_fields[array_field]:
-                filtered_uniqueness_fields[node_type][array_field].append('impactAssessment.@id')
+            if "impactAssessment.id" in array_fields[array_field]:
+                filtered_uniqueness_fields[node_type][array_field].append(
+                    "impactAssessment.@id"
+                )
     return filtered_uniqueness_fields
 
 
@@ -226,15 +233,17 @@ def _do_pivot(df_in, name_id_dict):
                         deep_pivoted, left_index=True, right_index=True, how="outer"
                     )
 
-            field_cols.dropna(axis=0, how='all', inplace=True)
+            field_cols.dropna(axis=0, how="all", inplace=True)
 
-            with_grouped_cols = field_cols.T.groupby(
-                _get_term_index, group_keys=True
-            ).apply(
-                _group_by_term,
-                name_id_dict=name_id_dict,
-                uniqueness_fields=uniqueness_fields
-            ).T
+            with_grouped_cols = (
+                field_cols.T.groupby(_get_term_index, group_keys=True)
+                .apply(
+                    _group_by_term,
+                    name_id_dict=name_id_dict,
+                    uniqueness_fields=uniqueness_fields,
+                )
+                .T
+            )
 
             pivoted_terms = with_grouped_cols.apply(
                 _pivot_row, axis=1, uniqueness_fields=uniqueness_fields
@@ -242,9 +251,13 @@ def _do_pivot(df_in, name_id_dict):
 
             # merge any duplicated columns caused by shuffled term positions
             # this operation coincidentally sorts the columns alphabetically
-            pivoted_terms = pivoted_terms.T.groupby(
-                level=pivoted_terms.columns.nlevels - 1, group_keys=False
-            ).apply(lambda term: term.bfill().iloc[0, :]).T
+            pivoted_terms = (
+                pivoted_terms.T.groupby(
+                    level=pivoted_terms.columns.nlevels - 1, group_keys=False
+                )
+                .apply(lambda term: term.bfill().iloc[0, :])
+                .T
+            )
 
             pivoted_terms.columns = map(
                 lambda col: f"{nt_label}.{field}.{col}", pivoted_terms.columns
@@ -265,14 +278,18 @@ def _format_and_pivot(df_in):
     df_out = _do_pivot(df_in.copy(), name_id_dict)
 
     _sort_inplace(df_out)
-    df_out = df_out.astype('object')
+    df_out = df_out.astype("object")
     df_out.fillna("-", inplace=True)
     return df_out
 
 
 def nodes_to_df(nodes: list[dict]):
     nodes_flattened = [
-        flatten_json(dict([(_get_node_type_label(node.get("@type", node.get('type'))), node)]), '.') for node in nodes
+        flatten_json(
+            dict([(_get_node_type_label(node.get("@type", node.get("type"))), node)]),
+            ".",
+        )
+        for node in nodes
     ]
 
     return pd.json_normalize(nodes_flattened)
@@ -306,7 +323,7 @@ def pivot_hestia_file(hestia_file: str):
         Pandas dataframe with pivoted array terms
     """
     parsed = json.loads(hestia_file, object_hook=_with_csv_formatting)
-    nodes = parsed.get('nodes', [parsed])
+    nodes = parsed.get("nodes", [parsed])
     return pivot_nodes(nodes)
 
 

hestia_earth/utils/pivot/pivot_json.py

@@ -6,10 +6,17 @@ from collections import defaultdict
 from copy import deepcopy
 
 from hestia_earth.utils.pipeline import _node_type
-from ._shared import EXCLUDE_FIELDS, EXCLUDE_PRIVATE_FIELDS, _with_csv_formatting, _filter_emissions_not_relevant
+from ._shared import (
+    EXCLUDE_FIELDS,
+    EXCLUDE_PRIVATE_FIELDS,
+    _with_csv_formatting,
+    _filter_emissions_not_relevant,
+)
 
 pivot_exclude_fields = Term().fields
-pivot_exclude_fields.update({k: "" for k in EXCLUDE_FIELDS} | {k: "" for k in EXCLUDE_PRIVATE_FIELDS})
+pivot_exclude_fields.update(
+    {k: "" for k in EXCLUDE_FIELDS} | {k: "" for k in EXCLUDE_PRIVATE_FIELDS}
+)
 
 term_exclude_fields = Term().fields
 del term_exclude_fields["name"]
@@ -31,8 +38,10 @@ for node_type, array_fields in UNIQUENESS_FIELDS.items():
                 if f not in ("properties.term.@id", "properties.value")
             ]
         # include `impactAssessment.@id` since it is not part of original uniqueness
-        if 'impactAssessment.id' in array_fields[array_field]:
-            ADAPTED_UNIQUENESS_FIELDS[node_type][array_field].append('impactAssessment.@id')
+        if "impactAssessment.id" in array_fields[array_field]:
+            ADAPTED_UNIQUENESS_FIELDS[node_type][array_field].append(
+                "impactAssessment.@id"
+            )
 
 
 def _combine_node_ids(nodes: list):
@@ -42,13 +51,18 @@ def _combine_node_ids(nodes: list):
 def _base_pivoted_value(key: str, value, is_top_level: bool):
     # handle list of Nodes
     return (
-        _combine_node_ids(value) if isinstance(value[0], dict) and value[0].get('@type') in NODE_TYPES else
-        json.dumps(value, separators=(',', ':')) if any([
-            is_top_level,
-            key in ['distribution']
-        ]) else
-        value
-    ) if isinstance(value, list) else value
+        (
+            _combine_node_ids(value)
+            if isinstance(value[0], dict) and value[0].get("@type") in NODE_TYPES
+            else (
+                json.dumps(value, separators=(",", ":"))
+                if any([is_top_level, key in ["distribution"]])
+                else value
+            )
+        )
+        if isinstance(value, list)
+        else value
+    )
 
 
 def _do_pivot(node, parent_node_type=None, parent_field=None, level=0):  # noqa: C901
@@ -57,13 +71,15 @@ def _do_pivot(node, parent_node_type=None, parent_field=None, level=0):  # noqa:
     if node_type not in ADAPTED_UNIQUENESS_FIELDS:
         return node
     pivoted_node = {
-        field: _base_pivoted_value(field, value, level==0)
+        field: _base_pivoted_value(field, value, level == 0)
         for field, value in node.items()
-        if all([
-            field not in ADAPTED_UNIQUENESS_FIELDS[node_type],
-            node_type != 'Term' or field not in term_exclude_fields,
-            field not in EXCLUDE_PRIVATE_FIELDS
-        ])
+        if all(
+            [
+                field not in ADAPTED_UNIQUENESS_FIELDS[node_type],
+                node_type != "Term" or field not in term_exclude_fields,
+                field not in EXCLUDE_PRIVATE_FIELDS,
+            ]
+        )
     }
 
     fields_to_pivot = [
@@ -227,7 +243,7 @@ def pivot_hestia_file(hestia_file: str):
     Pivot json array of schema-compliant nodes on 'nodes' key of unparsed json string
     """
     parsed = json.loads(hestia_file, object_hook=_with_csv_formatting)
-    return pivot_nodes(parsed.get('nodes', []))
+    return pivot_nodes(parsed.get("nodes", []))
 
 
 def pivot_nodes(nodes: list[dict]):