Java Regex编号多行列表

Question

I want to parse a document and extract each element of a numbered list for example I have this:

 1. I like to blah
    and blah
 2. But also to blah 
    and blah

I would like to extract each element from the list like [1. text for item1, 2. text from item2] I used a regular expression like this "[0-9].*;" before because I though each list item ended with ; but that is not always true. So I would like a regex to extract the text without it ending with ";". This is what I tried:

String regexLineNumber = "[0-9]..*;";
String[] splitted = inputData.split(regexLineNumber);

Answer 1

I would try to avoid regular expressions when you can in general. They are terribly memory inefficient and in most circumstances are just used as a shortcut. In this situation, you could easily create a BufferedReader and read each line looking for a certain value. something like

BufferedReader reader = new BufferedReader(/*instantiate here*/);
int nextNum = 2;
StringBuilder curRecord = new StringBuilder();
String line;
Collection<String> elements = new ArrayList<String>(/*Expected number ??*/);
while ((line = reader.readLine()) != null) {
   if (line.trim().startsWith(nextNum + ".")) {
       elements.add(curRecord.toString());
       curRecord = new StringBuilder();
       nextNum++;
   }
   curRecord.append(line);
}
if (!curRecord.trim().equals("")) {
   elements.add(curRecord.toString());
}

Answer 2

I suggest you use a regex that would allow numbers in the middle of or at the end of a sentence.

(?<=[\n\r\s]*|^)(\d\.[^\d]*)

Also remember to use matches and not findall .