i am working on a project to take a file, and save its sections. sections can be
1.
2.
3.
etc., but can also be
1.1
2.3.1.II.
etc.
now the basic of how to read i know. i need to know if there is a good way to detect and separate the text into sections.
i thought about regex, but i don't know enough regex to do it. any suggestions?
example:
1. Position
1.1. Position.
1.2. Scope
1.3. Location.
2. Compensation
2.1. Schedule
2.2.
3. Term
3.1. Term.
3.1.i. bla
3.1.ii. bla bla
You can use this regex to divide and capture the numbered section in group1 and paragraph section in group2.
^((?:[a-zA-Z\d]{1,2}\.)+)\s+(.*)
Here, ^((?:[a-zA-Z\d]{1,2}\.)+) captures the numbered section which starts with one to two alphanumeric characters followed by a literal dot whole of it one or more times. Then followed by a space hence \s+ then (.*) captures the remaining text which is assumed to be a paragraph. With your given sample data, this is what I have come up with. In case you need more cases covered differently, please add more samples and I will give you further refined solution.
Here is a sample Java code,
List<String> list = Arrays.asList("1. Position", "1.1. Position.", "1.2. Scope", "1.3. Location. ",
"2. Compensation", "2.1. Schedule", "2.2. ", "3. Term", "3.1. Term.", "3.1.i. bla", "3.1.ii. bla bla",
"12.a. some para", "13.a. some para", "1.a. some para", "A.1.a. another para", "B.1.a. some para");
Pattern p = Pattern.compile("^((?:[a-zA-Z\\d]+\\.)+)\\s+(.*)");
list.stream().forEach(x -> {
Matcher m = p.matcher(x);
if (m.matches()) {
System.out.println(x + " --> " + "number section: ("+m.group(1)+")" + " para section: ("+m.group(2)+")");
}
});
Prints,
1. Position --> number section: (1.) para section: (Position)
1.1. Position. --> number section: (1.1.) para section: (Position.)
1.2. Scope --> number section: (1.2.) para section: (Scope)
1.3. Location. --> number section: (1.3.) para section: (Location. )
2. Compensation --> number section: (2.) para section: (Compensation)
2.1. Schedule --> number section: (2.1.) para section: (Schedule)
2.2. --> number section: (2.2.) para section: ()
3. Term --> number section: (3.) para section: (Term)
3.1. Term. --> number section: (3.1.) para section: (Term.)
3.1.i. bla --> number section: (3.1.i.) para section: (bla)
3.1.ii. bla bla --> number section: (3.1.ii.) para section: (bla bla)
12.a. some para --> number section: (12.a.) para section: (some para)
13.a. some para --> number section: (13.a.) para section: (some para)
1.a. some para --> number section: (1.a.) para section: (some para)
A.1.a. another para --> number section: (A.1.a.) para section: (another para)
B.1.a. some para --> number section: (B.1.a.) para section: (some para)
You can match headings with regex like this one (assuming that Roman numerals are u to X):
^((?:(?:\d+|I{1,3}|IV|VI{0,3}|IX|X)\.)+)
Explanation:
^ beginning of the line\d+|I{1,3}|IV|VI{0,3}|IX|X - matches a numeral:
\d+ digitsI{1,3}|IV|VI{0,3}|IX|X Roman numerals(?:...) non capturing groups
\. dot separating the numeralsNUMERAL DOT groups once or moreEDIT:
In java you need to escape the pattern (so that java interprets it correctly) and use Pattern.MULTILINE (so that ^ marks beginning of the line not beginning of the string):
Pattern.compile("^((?:(?:\\d+|I{1,3}|IV|VI{0,3}|IX|X)\\.)+)", Pattern.MULTILINE)
