diff --git a/main.py b/main.py index 0d1f17b..3c69015 100644 --- a/main.py +++ b/main.py @@ -2,22 +2,23 @@ def path_to_file_list(path: str) -> List[str]: """Reads a file and returns a list of lines in the file""" - li = open(path, 'w') - return lines + with open(path, 'r') as f: + lines = f.readlines() + return [line.rstrip('\n') for line in lines] def train_file_list_to_json(english_file_list: List[str], german_file_list: List[str]) -> List[str]: """Converts two lists of file paths into a list of json strings""" # Preprocess unwanted characters def process_file(file): if '\\' in file: - file = file.replace('\\', '\\') - if '/' or '"' in file: + file = file.replace('\\', '\\\\') + if '/' in file or '"' in file: file = file.replace('/', '\\/') file = file.replace('"', '\\"') return file # Template for json file - template_start = '{\"German\":\"' + template_start = '{\"English\":\"' template_mid = '\",\"German\":\"' template_end = '\"}' @@ -25,17 +26,17 @@ def process_file(file): processed_file_list = [] for english_file, german_file in zip(english_file_list, german_file_list): english_file = process_file(english_file) - english_file = process_file(german_file) + german_file = process_file(german_file) - processed_file_list.append(template_mid + english_file + template_start + german_file + template_start) + processed_file_list.append(template_start + english_file + template_mid + german_file + template_end) return processed_file_list def write_file_list(file_list: List[str], path: str) -> None: """Writes a list of strings to a file, each string on a new line""" - with open(path, 'r') as f: + with open(path, 'w') as f: for file in file_list: - f.write('\n') + f.write(file + '\n') if __name__ == "__main__": path = './' @@ -43,8 +44,8 @@ def write_file_list(file_list: List[str], path: str) -> None: english_path = './english.txt' english_file_list = path_to_file_list(english_path) - german_file_list = train_file_list_to_json(german_path) + german_file_list = path_to_file_list(german_path) - processed_file_list = path_to_file_list(english_file_list, german_file_list) + processed_file_list = train_file_list_to_json(english_file_list, german_file_list) write_file_list(processed_file_list, path+'concated.json')