Java scanner delimiter detects ';' but not '\n'

Question

I have a formatted text file that I need to extract information from and put into corresponding member variables in a class Item (1 line of text = 1 Item). I am using Scanner and delimiter, but it can only recognize the ; separating information and not new lines.

I've tried a few different regex expressions and my latest one is down below where I specify only the delimiter I know is coming up. I've also tried the regex [;\\n] . My only conclusion is that scanner treats new line characters different than other (which kind of makes sense, I know it has functions based on new lines).

Here is the text file format

1000;Knock Bits;88;12.67;8015
1001;Widgets;10;35.50;8004
1002;Grommets;20;23.45;8001

and here is what my code looks like

while (scan.hasNext())
{           
Item item = new Item();
scan.useDelimiter("[;]");

item.setID(scan.nextInt());
item.setName(scan.next());
item.setQuantity(scan.nextInt());
item.setPriceInCents((int) scan.nextFloat()*100);

scan.useDelimiter("\\n");
item.setSupplierID(scan.nextInt()); 
}

All the above code works except the last line getting supplierID with nextInt(). I know I could just replace that line with

item.setSupplierID(Integer.parseInt(scan.nextLine()));

But that's kind of ugly and there should be a way to do it using regex without having to customize a line specifically for the last word. Preferably using only one delimiter for the whole loop.

It might not work if your end of line is something other than \n, which is possible. Try `"[;\\s]+"` instead. — markspace, Feb 06 '19 at 22:41

Milo Felipe · Answer 1 · 2019-02-06T23:18:24.697

2

Use scanner.useDelimiter("[;\n]"); or scanner.useDelimiter("[;\r\n]"); to work in both Windows and Linux systems.

Call it outside the while loop:

    scanner.useDelimiter("[;\r\n]");

    while (scanner.hasNext()) {
        Item item = new Item();
        item.setID(scanner.nextInt());
        item.setName(scanner.next());
        item.setQuantity(scanner.nextInt());
        item.setPriceInCents((int) scanner.nextFloat()*100);
        item.setSupplierID(scanner.nextInt());
    }

If using \R as suggested by mypetlion: scanner.useDelimiter(";|\\R");

edited Feb 06 '19 at 23:18

answered Feb 06 '19 at 22:45

Milo Felipe

41
2

1

To be fully agnostic, you can replace the `\r\n` with `\R`. Introduced to Java's pattern syntax in Java 8. – mypetlion Feb 06 '19 at 22:48
Using delimiter `[;\r\n]` for data like `1;2\r\n3;4` would fail since character class `[...]` describes potential set for *single* character it can match. In other words *each* `\r` and `\n` is treated as **separate** delimiter, meaning that scanner will think that there is one element (empty string) at `\r\n` (between `\r` and `\n`). Using `\R` instead solves this problem (but it must be used outside of character class since it doesn't represent single character but also `\r\n` *sequence*). – Pshemo Feb 06 '19 at 23:21
This solved the problem. I suppose the new line character was \r. To be fully agnostic without too much trouble, I suppose you could come up with a regex to include all numbers, spaces, tabs, and leave any other characters (punctuation, \n, \r etc) to be delimiters. This would allow for commas and other seperators to work. – Jedediah Feb 07 '19 at 01:14
Slightly unrelated question, \n\r and \\n\\r both work and do the same thing. How is that possible? Unless they coincidentally do the same thing in this scenario? – Jedediah Feb 07 '19 at 01:16
@Jedediah related [Understanding regex in Java: split(“\t”) vs split(“\\t”) - when do they both work, and when should they be used](https://stackoverflow.com/q/3762347) – Pshemo Feb 07 '19 at 08:19

mavriksc · Answer 2 · 2019-02-06T23:15:05.183

read the whole line then use string.split(";") to get the elements

List<Map<String,Object>> items = new ArrayList<>();
    try {
        Scanner scanner = new Scanner(new File(ClassLoader.getSystemResource("stats.txt").getFile()));
        while (scanner.hasNext()){
            String[] itemString = scanner.nextLine().split(";");
            Map<String,Object> item = new HashMap<>();
            item.put("id",Integer.parseInt(itemString[0]));
            item.put("name",itemString[1]);
            item.put("quant",Integer.parseInt(itemString[2]));
            item.put("price",Float.parseFloat(itemString[3]));
            item.put("suplierId",Integer.parseInt(itemString[4]));
            items.add(item);

        }
    } catch (FileNotFoundException e) {
        e.printStackTrace();
    }
    items.forEach(i->{
        i.forEach((key, value) -> System.out.println(key + " : " + value));
        System.out.println();
    });

Instead of Map you can just populate the object parsed values

Java scanner delimiter detects ';' but not '\n'

2 Answers2