sourashtra / tools /split_terms.py
orsenthil's picture
Upload folder using huggingface_hub
1e321b4 verified
Raw
History Blame Contribute Delete
9.06 kB
#!/usr/bin/env python3
"""
Term Splitter for Sourashtra Dictionary
This script splits composite terms in processed CSV files where columns c1, c2, c3
contain multiple terms separated by '/' or ',' characters.
For composite terms like:
a1/a2, b1/b2, c1/c2, d1, e1
It creates separate lines:
a1, b1, c1, d1, e1
a2, b2, c2, d1, e1
Usage: python3 split_terms.py <processed_directory> [output_directory]
"""
import argparse
import csv
import os
import sys
from pathlib import Path
import re
def split_field_by_separators(field):
"""
Split a field by '/' or ',' separators, handling quoted content properly.
Args:
field (str): Field content that may contain composite terms
Returns:
list: List of individual terms
"""
if not field or field.strip() == '':
return ['']
# Handle cases where the entire field is quoted
field = field.strip()
if field.startswith('"') and field.endswith('"'):
# Remove outer quotes and split the inner content
inner_content = field[1:-1]
# Split by comma first, then by forward slash
parts = []
for part in inner_content.split(','):
part = part.strip()
if '/' in part:
parts.extend([p.strip() for p in part.split('/')])
else:
parts.append(part)
return parts
else:
# Split by forward slash, then by comma
parts = []
for part in field.split('/'):
part = part.strip()
if ',' in part:
parts.extend([p.strip() for p in part.split(',')])
else:
parts.append(part)
return parts
def process_csv_line(row):
"""
Process a single CSV line and split composite terms in c1, c2, c3.
Args:
row (list): CSV row with 5 columns
Returns:
list: List of expanded rows with individual terms
"""
if len(row) != 5:
return [row] # Return as-is if not 5 columns
c1, c2, c3, c4, c5 = row
# Split the first three columns
c1_terms = split_field_by_separators(c1)
c2_terms = split_field_by_separators(c2)
c3_terms = split_field_by_separators(c3)
# Find the maximum number of terms across c1, c2, c3
max_terms = max(len(c1_terms), len(c2_terms), len(c3_terms))
# If all columns have only one term, return the original row
if max_terms == 1:
return [row]
# Pad shorter lists with their last element (or empty string if empty)
def pad_list(lst, target_length):
if not lst:
return [''] * target_length
while len(lst) < target_length:
lst.append(lst[-1] if lst else '')
return lst
c1_terms = pad_list(c1_terms, max_terms)
c2_terms = pad_list(c2_terms, max_terms)
c3_terms = pad_list(c3_terms, max_terms)
# Create individual rows for each term combination
result_rows = []
for i in range(max_terms):
new_row = [c1_terms[i], c2_terms[i], c3_terms[i], c4, c5]
result_rows.append(new_row)
return result_rows
def process_csv_file(input_file, output_file):
"""
Process a single CSV file by splitting composite terms.
Args:
input_file (Path): Path to input CSV file
output_file (Path): Path to output CSV file
Returns:
tuple: (success, error_message, original_lines, expanded_lines)
"""
try:
original_lines = 0
expanded_lines = 0
all_expanded_rows = []
# First, read all data
with open(input_file, 'r', encoding='utf-8') as infile:
csv_reader = csv.reader(infile)
for row in csv_reader:
# Skip completely empty rows
if not row or all(field.strip() == '' for field in row):
continue
original_lines += 1
# Process the row to split composite terms
expanded_rows = process_csv_line(row)
# Add all expanded rows to our collection
for expanded_row in expanded_rows:
all_expanded_rows.append(expanded_row)
expanded_lines += 1
# Then write all data
with open(output_file, 'w', encoding='utf-8', newline='') as outfile:
csv_writer = csv.writer(outfile, lineterminator='\n')
for row in all_expanded_rows:
# Clean any carriage returns from the row data
cleaned_row = [field.replace('\r', '').replace('\n', ' ') if isinstance(field, str) else field for field in row]
csv_writer.writerow(cleaned_row)
return True, None, original_lines, expanded_lines
except Exception as e:
return False, f"Error processing file: {str(e)}", 0, 0
def main():
"""Main function to process all CSV files in the specified directory."""
# Set up argument parser
parser = argparse.ArgumentParser(
description='Split composite terms in processed CSV files',
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog='''
Examples:
python3 split_terms.py processed/
python3 split_terms.py processed/ split_output/
python3 split_terms.py /path/to/processed/ /path/to/output/
'''
)
parser.add_argument('processed_directory',
help='Path to the directory containing processed CSV files')
parser.add_argument('output_directory',
nargs='?',
default=None,
help='Path to the output directory (default: overwrite input files)')
args = parser.parse_args()
# Convert to Path objects and resolve
processed_dir = Path(args.processed_directory).resolve()
if args.output_directory:
output_dir = Path(args.output_directory).resolve()
overwrite_mode = False
else:
output_dir = processed_dir
overwrite_mode = True
# Validate input directory
if not processed_dir.exists():
print(f"Error: Processed directory not found at {processed_dir}")
sys.exit(1)
if not processed_dir.is_dir():
print(f"Error: {processed_dir} is not a directory")
sys.exit(1)
# Create output directory if it doesn't exist and we're not overwriting
if not overwrite_mode:
output_dir.mkdir(parents=True, exist_ok=True)
print(f"Processing CSV files from: {processed_dir}")
print(f"Output directory: {output_dir}")
if overwrite_mode:
print("⚠️ WARNING: Files will be overwritten in place!")
print("=" * 70)
# Find all CSV files
csv_files = sorted(processed_dir.glob('*.csv'))
if not csv_files:
print("No CSV files found in the directory.")
return
total_files = len(csv_files)
processed_files = 0
total_original_lines = 0
total_expanded_lines = 0
for csv_file in csv_files:
# Create output filename
output_file = output_dir / csv_file.name
print(f"\nProcessing: {csv_file.name}")
if not overwrite_mode:
print(f" → {output_file.name}")
# Process the file
success, error_msg, original_lines, expanded_lines = process_csv_file(csv_file, output_file)
if success:
expansion_info = ""
if expanded_lines > original_lines:
expansion_info = f" (expanded from {original_lines} to {expanded_lines} lines)"
elif expanded_lines == original_lines:
expansion_info = f" (no composite terms found)"
print(f" ✓ Success - {expanded_lines} lines written{expansion_info}")
processed_files += 1
total_original_lines += original_lines
total_expanded_lines += expanded_lines
else:
print(f" ✗ Failed - {error_msg}")
print(f" Stopping processing due to error.")
# Print summary before exiting
print(f"\nSummary:")
print(f"Files processed: {processed_files} out of {total_files}")
print(f"Total original lines: {total_original_lines}")
print(f"Total expanded lines: {total_expanded_lines}")
sys.exit(1)
# All files processed successfully
print("\n" + "=" * 70)
print("✓ All CSV files processed successfully!")
print(f"Summary:")
print(f"Total files processed: {processed_files}")
print(f"Total original lines: {total_original_lines}")
print(f"Total expanded lines: {total_expanded_lines}")
if total_expanded_lines > total_original_lines:
print(f"Lines added due to term splitting: {total_expanded_lines - total_original_lines}")
print(f"Output directory: {output_dir}")
if __name__ == "__main__":
main()