def datadocs_to_es(datadoc, session=None):
title = datadoc.title
cells_as_text = []
for cell in datadoc.cells:
if cell.cell_type == DataCellType.text:
cells_as_text.append(richtext_to_plaintext(cell.context))
elif cell.cell_type == DataCellType.query:
cell_title = cell.meta.get("title", "")
cell_text = (cell.context if not cell_title else
f"{cell_title}\n{cell.context}")
cells_as_text.append(cell_text)
else:
cells_as_text.append("[... additional unparsable content ...]")
joined_cells = escape("\n".join(cells_as_text))
# There is no need to compute the list of editors
# for public datadoc since everyone is able to see it
editors = ([
editor.uid
for editor in get_data_doc_editors_by_doc_id(data_doc_id=datadoc.id,
session=session)
] if not datadoc.public else [])
expand_datadoc = {
"id": datadoc.id,
"environment_id": datadoc.environment_id,
"owner_uid": datadoc.owner_uid,
"created_at": DATETIME_TO_UTC(datadoc.created_at),
"cells": joined_cells,
"title": title,
"public": datadoc.public,
"readable_user_ids": editors,
}
return expand_datadoc
def table_to_es(table, session=None):
schema = table.data_schema
column_names = [c.name for c in table.columns]
schema_name = schema.name
table_name = table.name
description = (
richtext_to_plaintext(table.information.description, escape=True)
if table.information
else ""
)
full_name = "{}.{}".format(schema_name, table_name)
weight = get_table_weight(table.id, session=session)
expand_table = {
"id": table.id,
"metastore_id": schema.metastore_id,
"schema": schema_name,
"name": table_name,
"full_name": full_name,
"full_name_ngram": full_name,
"completion_name": {
"input": [full_name, table_name,],
"weight": weight,
"contexts": {"metastore_id": schema.metastore_id,},
},
"description": description,
"created_at": DATETIME_TO_UTC(table.created_at),
"columns": column_names,
"golden": table.golden,
"importance_score": weight,
"tags": [tag.tag_name for tag in table.tags],
}
return expand_table
def board_to_es(board, fields=None, session=None):
def get_board_doc_titles():
return [doc.title for doc in board.docs if doc.title]
def get_board_table_names():
table_names = []
for table in board.tables:
table_names.append("{}.{}".format(table.data_schema.name,
table.name))
return table_names
def get_readable_user_ids():
uids = [board.owner_uid]
for board_editor in board.editors:
uids.append(board_editor.uid)
return uids
field_to_getter = {
"id": board.id,
"title": board.name,
"environment_id": board.environment_id,
"description": lambda: richtext_to_plaintext(board.description),
"public": board.public,
"owner_uid": board.owner_uid,
"full_table_name": get_board_table_names,
"doc_name": get_board_doc_titles,
"readable_user_ids": get_readable_user_ids,
}
return _get_dict_by_field(field_to_getter, fields=fields)
def get_joined_cells(datadoc):
cells_as_text = []
for cell in datadoc.cells:
if cell.cell_type == DataCellType.text:
cells_as_text.append(richtext_to_plaintext(cell.context))
elif cell.cell_type == DataCellType.query:
cell_title = cell.meta.get("title", "")
cell_text = (cell.context if not cell_title else
f"{cell_title}\n{cell.context}")
cells_as_text.append(cell_text)
else:
cells_as_text.append("[... additional unparsable content ...]")
joined_cells = escape("\n".join(cells_as_text))
return joined_cells
def get_table_description():
return (richtext_to_plaintext(table.information.description,
escape=True)
if table.information else "")
