String decompression (Huawei OD)

String decompression

There is a simple compression algorithm ：
For strings composed of all lowercase English letters , Compress the part of more than two consecutive identical letters into consecutive numbers plus the other parts of the letter to keep the same
unchanged , for example aaaab 4ab
White those decompression functions , According to the input string , Judge whether it is a legally compressed string
Input description ：
The input line , For one asccii character string
Length not exceeding 100 character
The use case ensures that the length of the output string will not exceed 100 character
Output description ：
If it is judged that the input is a legal compressed string
Output the string before compression
If the input is illegal , Output "!error"
4dff
ddddff

• Answer key
• The situation of this problem is relatively , I don't know how to organize filter conditions , It's not difficult to mention . At present, I can only do the following , such 4dddd This situation is not considered .
  The code is as follows ：

def decompression():
    string = input()
    legal_string = "abcdefghijklmnopqrstuvwxyz0123456789"
    index = 0
    res = ""
    #  Validation 
    if len(set(string)) == 1 and len(string) >= 3:
        print("!error")
        return

    while index < len(string):
        if string[index] not in legal_string:
            print("!error")
            break
        num = ""
        if string[index].isdigit():
            while string[index].isdigit(): #  consider  22d 11d 20a 10a The situation of 
                num += string[index]
                index += 1
           
            if num == "2" or num == "1": #  Exclusion number is 2  perhaps  1  The situation of 
                print("!error")
                break
             #  It must be a letter 
            res += int(num) * string[index]
        else:
            res += string[index]
        index += 1
    else:
        print(res)
if __name__ == '__main__':
    decompression()