systemy_dialogowe/utils/script.py

def remove_empty_lines(text):
    lines = text.split('\n')

    non_empty_lines = []
    for line in lines:
        if line.strip():
            non_empty_lines.append(line)

    return '\n'.join(non_empty_lines)


def join_lines(text):
    lines = text.split("\n")
    joined_lines = [lines[0]]

    for line in lines[1:]:
        if line.startswith("["):
            joined_lines.append(line)
        else:
            joined_lines[-1] += " " + line

    return "\n".join(joined_lines)


def process_text(text, nazwisko_user, nazwisko_system):
    lines = text.split('\n')
    result = []
    for line in lines:
        if nazwisko_system in line:
            result.append("system " + line.split(nazwisko_system, 1)[1].strip())
        elif nazwisko_user in line:
            result.append("user " + line.split(nazwisko_user, 1)[1].strip())
        else:
            result.append(line)
    return '\n'.join(result)


def read_tsv(file_path):
    with open(file_path, 'r', encoding='utf-8') as tsv_file:
        content = tsv_file.read()
        spaces=0
        pos_of_names = []
        for letter in range(15,len(content)):
            if content[letter] =="\n"or  content[letter] ==" ":
                spaces = spaces+1
                pos_of_names.append(letter)
                if spaces == 2:
                    break
        for letter in range(pos_of_names[1],len(content)):
            if content[letter] =="]":
                pos_of_names.append(letter+2)
                for letter2 in range(pos_of_names[2],len(content)):
                    if content[letter2] =="\n":
                        pos_of_names.append(letter2)
                        spaces=spaces+1
                        break
                if spaces==3:
                    break
        user = content[14:pos_of_names[1]]
        operator = content[pos_of_names[2]:pos_of_names[3]]
    return content,user,operator


def write_tsv(file_path, content):
    with open(file_path, 'w', encoding='utf-8') as tsv_file:
        tsv_file.write(content)


def replace_first_space_with_tab(text):
    lines = text.split("\n")
    new_lines = []
    for line in lines:
        first_space_index = line.find(" ")
        if first_space_index != -1:
            new_line = line[:first_space_index] + "\t" + line[first_space_index + 1:]
            new_lines.append(new_line)
        else:
            new_lines.append(line)
    return "\n".join(new_lines)

def do_all(filename):
    path="C:/Users/macty/OneDrive/Pulpit/conversations"
    file_path = path+"/"+filename
    text,user,operator = read_tsv(file_path)
    text = remove_empty_lines(text)
    text = join_lines(text)
    text = process_text(text, user, operator)
    text = replace_first_space_with_tab(text)
    write_tsv(file_path, text)


if __name__ == '__main__':
    import os
    
    for filename in os.listdir("C:/Users/macty/OneDrive/Pulpit/conversations"):
        do_all(filename)
    #nie działa dla sytuacji gdzie ktoś na początku pisze 2 razy zanim dostanie odpowiedź, ale nie chce mi sie kombinować z poprawianiem tego dla jednej sytuacji