def write_to_json(self, export_fp: str):
"""
get all QASet objects from self.pool
convert every QASet object into a dictionary
write the list of dictionaries into json
"""
export_fp: Path = Path(export_fp)
if export_fp.is_file():
overwrite = input(
f"{export_fp} already exist, overwrite it (y/n)? ")
if overwrite in ("Y", "y", "yes"):
logger.warning(f"Overwriting {export_fp}")
else:
logger.error(f"Export aborted")
return
lst = []
# gather all data from pool
for qa_set in self.pool:
lst.append(
{tag: getattr(qa_set, tag)
for tag in QASet.property_list})
with export_fp.open(mode="w") as f:
json.dump(lst, f, indent=4)
def export_excel_to_json(qa_bank_excel: str):
qa_bank_excel = Path(qa_bank_excel)
if not qa_bank_excel.is_file():
logger.error(f"Filepath '{str(qa_bank_excel)}' does not exist.")
return
if str(qa_bank_excel)[-5:] != ".xlsx":
logger.error(
f"QA Bank Filepath requires an Excel file. '{str(qa_bank_excel)}' doesn't match with '*.xlsx'."
)
return
pool = QASetPool()
pool.load_from_excel(qa_bank_excel)
export_fp = qa_bank_excel.parent / f"{qa_bank_excel.stem}.json"
pool.write_to_json(export_fp)
return
def load_from_json(self, json_fp: str):
""" 1. read from json
2. create QASet objects from json data
3. add QASet objects into self.pool
"""
assert Path(json_fp).is_file()
try:
qadict_lst = json.load(open(json_fp))
logger.debug(f"{json_fp} has been loaded")
except:
logger.error(f"Failed loading {json_fp}")
return
for qa_dict in qadict_lst:
self.add_to_pool(QASet(**qa_dict))
def load_from_excel(self, excel_fp: str):
""" 1. read from excel
2. create QASet objects from excel data
3. add QASet objects into self.pool
"""
qa_sheet = pandas.read_excel(excel_fp)
rows_num, cols_num = qa_sheet.shape
logger.debug(f"rows_num: {rows_num}")
logger.debug(f"cols_num: {cols_num}\n")
for i in range(0, rows_num):
try:
qa_set_id = int(qa_sheet.iloc[i][2])
logger.debug(f"ID: {qa_set_id}")
except Exception as e:
logger.error("Failed getting QASet ID from excel row")
continue
qa_set_qn: str = qa_sheet.iloc[i][3]
logger.debug(f"qa_set_qn: {qa_set_qn}")
if not self.get_by_id(qa_set_id):
qa_set_type: str = qa_sheet.iloc[i][0]
logger.debug(f"qa_set_type: {qa_set_type}")
qa_set_subtype: str = qa_sheet.iloc[i][1]
logger.debug(f"qa_set_subtype: {qa_set_subtype}")
qa_set_options: OrderedSet = OrderedSet(
str(qa_sheet.iloc[i][j]) for j in range(4, cols_num))
# remove emtpy option, which is 'nan' in pandas
qa_set_options.discard("nan")
logger.debug(f"qa_set_options: {qa_set_options}")
qa_set: QASet = QASet(
id=qa_set_id,
type=qa_set_type,
subtype=qa_set_subtype,
qns=qa_set_qn,
options=qa_set_options,
)
self.add_to_pool(qa_set)
logger.debug(">>> Added new QASet into pool\n")
# logger.debug(qa_set)
else:
self.get_by_id(qa_set_id).append_rephrase(qa_set_qn)
logger.debug(
"### Added new rephrased version into existing QASet\n")
def subtype(self, value):
logger.error(
f"QASet subtype is immutable, and it remains as {self._subtype}")
return
def id(self, value):
logger.error(f"QASet ID is immutable, and it remains as {self._id}")
return
def options(self, value):
logger.error(
f"QASet options is immutable, and it remains as {self._options}")
return