валидатор говорит что считает больше чем есть слов на самом деле

public static void main(String[] args) throws IOException {
    BufferedReader reader = new BufferedReader(new InputStreamReader(System.in));
    FileReader file = new FileReader(reader.readLine());
    int counter = 0;
    char[] world = new char[]{'w','o','r','l','d'};
    for (int i = 0; file.ready(); i++) {
        int c = file.read();
        if (c == world[i]){
            if (i == world.length-1) {
                int next = file.read();
                if (!(next>=97 && next<=122)) counter++;
                i=-1;
            }
        }
        else i=-1;
    }
    System.out.println(counter);
    file.close();
    reader.close();

}

world,orld

worldworldworldworld

sworld

try (BufferedReader filenameReader = new BufferedReader(new InputStreamReader(System.in));
     BufferedReader fileReader = new BufferedReader(new FileReader(filenameReader.readLine()))) {
    String content = fileReader.lines().collect(Collectors.joining(System.lineSeparator()));

String[] words = content.split("\\W+");
int countWithSplitStream = (int) Arrays.stream(words).filter("world"::equals).count();

// если стримы непонятны, то можно сделать подлиннее, но проще для понимания:
int countWithSplitCycle = 0;
for (String word : words)
    if ("world".equals(word))
        countWithSplitCycle++;

Matcher matcher = Pattern.compile("\\bworld\\b").matcher(content);
int countWithRegexCycle = 0;
while (matcher.find())
    countWithRegexCycle++;

// или так:
int countWithRegexStream = (int) Pattern.compile("\\bworld\\b")
                                        .splitAsStream(content)
                                        .count();

System.out.println("regex + stream: " + countWithRegexStream);
System.out.println("regex + cycle:  " + countWithRegexCycle);
System.out.println("split + stream: " + countWithSplitStream);
System.out.println("split + cycle:  " + countWithSplitCycle);