Using Spaces in Java Regex

Question

I'm having a lot of difficulty with spaces in Java while using regular expressions. The assignment is to split a comma-separated input string like,

J,Project report,"F, G, I",1

into separate strings containing:
J
Project report
F, G, I
1
if that makes sense. I'm using a Scanner to split the string. My regex (and code) I'm using is.

while (t.hasNext("([a-zA-Z0-9]| )*(\".+\")*,?")) { 
    System.out.println("t.next is : " + t.next());

...where t is a scanner of the input string as described above. But this does not appear to ever resort to true as nothing is printed. The closest I can get to working is just using simply ".*" as my regex, but that will separate at spaces and I need to separate only at the commas NOT within quotation marks. Can anyone assist? Thank you.

must you use regular expressions? they are perhaps not the best tool for the job here — stew, Dec 10 '11 at 05:46
No it's not required, but simply using a "," as a delimiter won't work, as F, G, and I will be separated when they are to be part of one string. This can admittedly be simpler, but I am not sure how to approach it. — Stefan Arambasich, Dec 10 '11 at 05:48

score 1 · Answer 1 · answered Dec 10 '11 at 05:53

This CAN be done with a regular expression, but a regular expression is perhaps not the best tool for the job. The expression you are going to end up with is going to be hard to read/maintain, and isn't necessarily going to be any more efficient.

Without going into too much detail, as this is your homework, not mine, I'd think about this another way:

You need a stateful scanner. You have two states, "i'm in the middle of quotes" and "i'm not". Scan the string character by character and each character will cause you either to accumulate a future result, emit a result or change states.

If this needs to be more robust, it might need to have even more states, for example. if you need to also parse something like:

a,"b\"c",d

No, it doesn't need to be very robust. This is the entire file: http://pastebin.com/imej2AYM I don't need to worry about anything else besides these cases. I like the state idea too. — Stefan Arambasich, Dec 10 '11 at 05:56

score 1 · Accepted Answer · edited May 23 '17 at 11:55

1

Try this out:

(",(?=([^\"]*\"[^\"]*\")*[^\"]*$)")

Reference: Java: splitting a comma-separated string but ignoring commas in quotes

Also, http://regexpal.com/ is a really neat and useful tool when it comes to testing out regexes :)

edited May 23 '17 at 11:55

Community

1
1

answered Dec 10 '11 at 06:38

Hristo

42,002
60
155
224

Your help is greatly appreciated! I can't believe I didn't find that while searching before. That is a big help. – Stefan Arambasich Dec 10 '11 at 21:31

score 1 · Answer 3 · answered Dec 10 '11 at 09:16

1

I agree with the suggestion that a robust third parter CSV library is a way to go. However, here is how you can use Scanner.

Scanner t = new Scanner(new File("test.csv"));
t.useDelimiter(',(?=([^\"]*\"[^\"]*\")*[^\"]*$)');
while( t.hasNext() ) {
    System.out.println(t.next());
}

I used the regex from @Hristo answer.

answered Dec 10 '11 at 09:16

JRideout

1,525
10
16

Thank you JRideout! Third-party is good but it's always nice to be able to solve a problem yourself without handing it off to someone else to finish - I want to understand what's going on and be able to understand the gory details to make me a better programmer. But maybe I shouldn't be using Java in that case? ;) The assignment wasn't actually mine, I was helping a friend, but this will help a lot thank you! – Stefan Arambasich Dec 10 '11 at 21:37

Mark O'Connor · Answer 4 · 2011-12-10T13:51:06.003

CSV files are more complex than they at first appear. For example in German countries the file separator is normally the ";" character..... While I understand your assigment was to use regexp's, don't waste your time when solving this problem for real.

My tool of choice is opencsv. Here's a groovy script (I leave you to convet it to Java) that parses your string:

import au.com.bytecode.opencsv.CSVParser

@Grapes([
    @Grab(group='net.sf.opencsv', module='opencsv', version='2.3')
])

CSVParser csv = new CSVParser()
String[] result = csv.parseLine('J,Project report,"F, G, I",1')

assert result[0] == "J"
assert result[1] == "Project report"
assert result[2] == "F, G, I"
assert result[3] == "1"

The CSVReader object provides ways to iterate over the file contents:

new File("data.csv").withReader { reader ->
    CSVReader csv = new CSVReader(reader);

    csv.readAll().each {
        println it[0]
        println it[1]
        println it[2]
        println it[3]
    }
}

The assignment wasn't actually to use regex and it actually isn't even my assignment haha. My friend needed help for his software engineering class, and it was bothering me that I wasn't able to help solve the problem, hence my thread. But keeping in mind this library would probably be beneficial for both my friend and I. Thanks! I almost picked this response as the answer. Very informative! — Stefan Arambasich, Dec 10 '11 at 21:34

Using Spaces in Java Regex

4 Answers4