import re
import os
def parse_ast(ast_filepath):
# check if file exists in folder
if not os.path.isfile(ast_filepath):
return {}
result = {}
with open(ast_filepath, "r") as f:
lines = f.readlines()
# remove endlines
lines = [line.strip() for line in lines]
# remove empty lines
lines = [line for line in lines if line.strip()]
# find regex expression: c="pain" 55:10 55:10\|\|t="problem"\|\|a="hypothetical"
regex = re.compile(r'c="(.*?)" (\d+):(\d+) (\d+):(\d+)\|\|t="(.*?)"\|\|a="(.*?)"')
matches = [regex.match(line) for line in lines if regex.match(line)]
result["concept_text"] = [match.group(1) for match in matches]
result["start_line"] = [int(match.group(2)) for match in matches]
result["start_word_number"] = [int(match.group(3)) for match in matches]
result["end_line"] = [int(match.group(4)) for match in matches]
result["end_word_number"] = [int(match.group(5)) for match in matches]
result["concept_type"] = [match.group(6) for match in matches]
result["assertion_type"] = [match.group(7) for match in matches]
return result
def parse_concept(concept_filepath):
# check if file exists in folder
if not os.path.isfile(concept_filepath):
return {}
result = {}
with open(concept_filepath, "r") as f:
lines = f.readlines()
# remove endlines
lines = [line.strip() for line in lines]
# remove empty lines
lines = [line for line in lines if line.strip()]
# find regex expression: c="a workup" 27:2 27:3\|\|t="test"
regex = re.compile(r'c="(.*?)" (\d+):(\d+) (\d+):(\d+)\|\|t="(.*?)"')
matches = [regex.match(line) for line in lines if regex.match(line)]
result["concept_text"] = [match.group(1) for match in matches]
result["start_line"] = [int(match.group(2)) for match in matches]
result["start_word_number"] = [int(match.group(3)) for match in matches]
result["end_line"] = [int(match.group(4)) for match in matches]
result["end_word_number"] = [int(match.group(5)) for match in matches]
result["concept_type"] = [match.group(6) for match in matches]
return result
def find_char_indexes(row, text):
""" function that can be applied to concept_df or assertion_df
Example:
assertions_df = assertions_df.apply(find_char_indexes, axis=1, args=(text,))
"""
text = text.lower()
line = text.split("\n")[row["start_line"] - 1] # NOTE: we assume that start_line == end_line
line = " ".join(line.split()) # remove multiple spaces
row["concept_text"] = " ".join(row["concept_text"].split()) # remove multiple spaces
# find character index start and end of concept
start_char_index = len(" ".join(line.split()[: row["start_word_number"]])) # number of chars before concept
if start_char_index > 0:
start_char_index += 1
end_char_index = start_char_index + len(row["concept_text"])
# assert (
# line[start_char_index:end_char_index] == row["concept_text"]
# ), f"concept_text doesn't match the found indexes. '{line[start_char_index:end_char_index]}' != '{row['concept_text']}'"
row["start_char_index"] = start_char_index
row["end_char_index"] = end_char_index
return row
def parse_relation(relation_filepath):
# check if file exists in folder
if not os.path.isfile(relation_filepath):
return {}
result = {}
with open(relation_filepath, "r") as f:
lines = f.readlines()
# remove endlines
lines = [line.strip() for line in lines]
# remove empty lines
lines = [line for line in lines if line.strip()]
# find regex expression: c="percocet" 55:1 55:1\|\|r="TrAP"\|\|c="pain" 55:10 55:10
regex = re.compile(r'c="(.*?)" (\d+):(\d+) (\d+):(\d+)\|\|r="(.*?)"\|\|c="(.*?)" (\d+):(\d+) (\d+):(\d+)')
matches = [regex.match(line) for line in lines if regex.match(line)]
result["concept_text_1"] = [match.group(1) for match in matches]
result["start_line_1"] = [int(match.group(2)) for match in matches]
result["start_word_number_1"] = [int(match.group(3)) for match in matches]
result["end_line_1"] = [int(match.group(4)) for match in matches]
result["end_word_number_1"] = [int(match.group(5)) for match in matches]
result["concept_text_2"] = [match.group(7) for match in matches]
result["start_line_2"] = [int(match.group(8)) for match in matches]
result["start_word_number_2"] = [int(match.group(9)) for match in matches]
result["end_line_2"] = [int(match.group(10)) for match in matches]
result["end_word_number_2"] = [int(match.group(11)) for match in matches]
result["relation_type"] = [match.group(6) for match in matches]
return result
