Upload remove_language_tags.py with huggingface_hub
Browse files- remove_language_tags.py +21 -0
remove_language_tags.py
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import re
|
| 2 |
+
|
| 3 |
+
def remove_language_tags(input_file, output_file):
|
| 4 |
+
language_tags = r'^(hi|en|mr|pa|bn|or|ur|as|ml|ta|te|kn|gu)\s*:'
|
| 5 |
+
|
| 6 |
+
with open(input_file, 'r', encoding='utf-8') as infile, open(output_file, 'w', encoding='utf-8') as outfile:
|
| 7 |
+
for line in infile:
|
| 8 |
+
# Remove language tags
|
| 9 |
+
cleaned_line = re.sub(language_tags, '', line.strip(), flags=re.IGNORECASE)
|
| 10 |
+
# Write the cleaned line, preserving original whitespace
|
| 11 |
+
outfile.write(cleaned_line.strip() + '\n')
|
| 12 |
+
|
| 13 |
+
print(f"Processed {input_file} and saved results to {output_file}")
|
| 14 |
+
|
| 15 |
+
def main():
|
| 16 |
+
# Process source file
|
| 17 |
+
remove_language_tags('NMT_V2/train.src', 'NMT_V2/train_cleaned.src')
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
if __name__ == "__main__":
|
| 21 |
+
main()
|