Parses portfolio related IL outputs to Excel
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
PortfolioParser/ILParser.py

105 lines
4.1 KiB

from pandas import DataFrame
import re
from re import Match, Pattern
from logging import getLogger, basicConfig
from json import load, dump
logger = getLogger(__name__)
logger.setLevel("DEBUG")
COLUMN_NAME_REGEX = re.compile(r"(?P<column_name>(\w|\.|#|\/)+)", re.IGNORECASE)
def replace_bad_cols(line: str, cols: list[str]) -> str:
"""
Replaces bad column names in a string with modified names that have spaces replaced with dots.
Args:
line (str): The string containing the column names to modify.
cols (list[str]): A list of column names to modify.
Returns:
str: The modified string with bad column names replaced.
"""
logger.debug(f"Line: {line} | Cols: {cols}")
for c in cols:
# Create a regex for the col
col_regex: Pattern = re.compile(c.replace(' ', r'(?:\s|\.)'))
logger.debug(f"Col_regex: {col_regex}")
# Get all columns that match that pattern
col_matches: list[str|tuple[str]] = re.findall(col_regex, line)
logger.debug(f"Col_matches: {col_matches}")
# Match the substition for all matches if any
col_name: str
for col_name in col_matches:
logger.debug(f"col_name: {col_name}")
# Replace the bad column name with the modified column name in the string
# Adding the '.' instead of a space helps the parser tell what the continous
# column are
line = line.replace(col_name, col_name.replace(' ', '.'))
return line
def extract_data(input_doc: str, column_list: list[str]) -> DataFrame|None:
"""
Extracts data from a string in a table-like format, where columns are identified by a list of column names, and
returns the data as a Pandas DataFrame.
Args:
input_doc (str): The string containing the table-like data to extract.
column_list (list[str]): A list of column names to identify the columns in the table-like data.
Returns:
pandas.DataFrame: A DataFrame containing the extracted data from the input string.
"""
line: str
columns = {}
data = {}
for line in input_doc.splitlines():
if len(columns) == 0 :
logger.debug(f"Columns = 0: {line}")
# Find the line that contains the column names and replace bad column names
if re.search("^\w", line):
logger.debug("Found word on first line.")
line = replace_bad_cols(line, column_list)
logger.debug(f"Column replacements made: {line}")
# Find the start and end positions of each column name and store them in a dictionary
columns_names = re.finditer(COLUMN_NAME_REGEX, line)
logger.debug(f"Found column names: {columns_names}")
for c in columns_names:
columns[c.group("column_name")] = {"start": c.start(), "end": c.end()}
logger.debug(f"Column section: {columns[c.group('column_name')]}")
data[c.group("column_name")] = []
continue
elif len(line) < 2:
logger.debug(f"Line len less than 2.")
continue
# Check if we've reached the end of the table and return the data
if re.search("\d+ records listed", line):
logger.debug(f"End of document: {line}")
logger.debug(f"Extracted data: {data}")
return DataFrame(data)
# Extract the data from each column based on the start and end positions
for key, span in columns.items():
data[key].append(line[span["start"]:span["end"]].strip())
if __name__ == "__main__":
basicConfig(filename='ILParser.log', encoding='utf-8',
level="DEBUG", filemode='w', force=True)
def test_replace_bad_cols():
with open("Inputs\CUST_ISSUE") as c:
input: str = c.read()
with open("config.json") as configFile:
config: dict = load(configFile)
columns: list[str] = config["COLS"]
replace_bad_cols(input.splitlines()[1], columns)
test_replace_bad_cols()