替换所有字符串，但如果哈希表中存在文本，则替换为文本的第一个

Question

The text is stored in String variable and it is processed by some API to give me a hashmap which stores the keys and values. The keys are some specific words in text and value is a new word which will replace the key in the text. I have to process the text such that it will replace the keys with the value from hashmap but I have to leave the first instance of the key in the text as it is.

Problem: I am able to replace all the instances which I am doing by iterating the hashmap and replacing the keys in the text. I want to leave the first matched key as it is.

In string functions what I see is replace, replaceAll, replaceFirst.

How should I handle this case.

for eg:

input: Example [2] This is a sample text. This is a sample text [69-3]. This is a sample [69-3] text.

hashmap: {sample=sImple,text=text2,[69-3]=somenum}

output: Example [2] This is a sample text. This is a sImple text2 [69-3]. This is a sImple text2 somenum .

Also the key match is for the whole word and not a subString. like if name is the key and surname is the string in the text then it should not be matched and sur"name" should not be changed. I m using replaceAll instead of replace to do the replacement.

Thanks in advance.

Answer 1

The following

String input="Example [2] This is a sample text. This is a sample xtexty text [69-3]. This is a sample [69-3] textME text.";

        Map<String,String> map = new HashMap<String,String>();
        map.put("sample","sImple");
        map.put("text","text2");
        map.put("[69-3]","somenum");

        for(Map.Entry<String, String> entry : map.entrySet()){
            input =input.replace(entry.getKey(),entry.getValue());
        input = input.replaceFirst(entry.getValue(), entry.getKey());
        Pattern p = Pattern.compile("(\\w+)*"+entry.getValue()+"(\\w+)|(\\w+)"+entry.getValue()+"(\\w+)*");
        Matcher matcher =  p.matcher(input);
       while( matcher.find()){
          int r =  matcher.group().indexOf(entry.getValue());
          int s =r+input.indexOf(matcher.group());
     input = input.substring(0,s)+entry.getKey()+input.substring(s+entry.getValue().length());
       }    
        }

        System.out.println(input);

    }

will print:

  Example [2] This is a sample text. This is a sImple xtexty text2 [69-3]. This is a sImple somenum textME text2.

The above piece of code will not replace substrings, works as you want .

Answer 2

You can find the index of the first appearance, and replace all after this index. Because I didn't find a replaceAll taking an offset parameter, I can suggest you this handmade solution using StringBuilder#replace :

public static void replaceAllButFirst(StringBuilder modifiedString, String match, String replacement) {
    int index = modifiedString.indexOf(match);
    int matchLength = match.length(), replacementLength = replacement.length();
    if (index == -1) return;
    index += matchLength;
    index = modifiedString.indexOf(match, index);
    while (index != -1) {
        modifiedString.replace(index, index + matchLength, replacement);
        index += replacementLength;
        index = modifiedString.indexOf(match, index);
    }
}

Example

Answer 3

You can do this using regex.

The answer to your problem was already addressed in this post:

In Java how do you replace all instances of a character except the first one?

Answer 4

I kind of solved this one using replaceFirst and replaceAll.

Create a dictionary2 which will contain Key same as in dictionary and the values in dictionary2 will be modified version of key

foreg:

dictionary: {sample=simple}

dictionary2: {sample=sample---A--}

Then replace the first instance of the string in the text with the value in dictionary2 using replaceFirst.

Next replaceAll the instances of the strings in the text which will leave the first instance as it is, then replace the modified first instance with the key in dictionary2.