使用Java从html文件中提取数据

Question

I have an HTML file full of addresses I need to extract. Looks like this, but with about 60 streets, and multiple numbers on each street

    <BR>
    <Font Color=#FF0000 Size=7>MACARTNEY STREET (L)</Font>
    <BR>
    <BR>
    10........<Font Color=#FFFFFF> CM </Font>
    <BR>
    <BR>
    15........<Font Color=#FF0000> SH </Font>
    <BR>
    <BR>
    43A.......<Font Color=#FFFFFF> CM </Font>
    <BR>

I've been using regex to extract data, which works a treat for getting the street names

    final Pattern STREETNAME = Pattern.compile("<Font Color=#FF0000 Size=7>(.+?)</Font>");
    Matcher stMatcher = STREETNAME.matcher("");
    while ((line = reader.readLine()) != null) {
       stMatcher = STREETNAME.matcher(line);
        if (stMatcher.find()) {
            String street = stMatcher.group(1);
            customerList.add(new Customer(street));}
    //customerList is an array of Customer Objects, defined elsewhere in the program

but no matter what, I just can't get it to read the house numbers (10, 15 and 43A in the example).

Ideally I would store the street name string, extract the house number and concat them together before creating the customer object. I'll need to check for the CM or SH line as well, but that can wait.

Anyone got an idea that might help? I'm pretty stumped right now.

Thanks!

Answer 1

It's easy (I change also your existing regexp to match any color definition), this should work and get numbers and additional CM or SH

final Pattern STREETNAME = Pattern.compile("<Font Color=.* Size=7>(.+?)</Font>");
final Pattern STREETNUMBER = Pattern.compile("^(\\d[^\\.]*)<Font Color=.*>\\s*(.+?)\\s*</Font>");
Matcher stMatcher;
Customer lastCustomer = null;
while ((line = reader.readLine()) != null) {
    stMatcher = STREETNAME.matcher(line);
    if (stMatcher.find()) {
        String street = stMatcher.group(1);
        lastCustomer = new Customer(street);
    } else {
        stMatcher = STREETNUMBER.matcher(line);
        if (stMatcher.find()) {
            if (lastCustomer != null) {
                lastCustomer.setStreetNumber(stMatcher.group(1));
                lastCustomer.setStreetCmSh(stMatcher.group(2));
                customerList.add(lastCustomer);
            }
        }
    }
}

How does it work?

• The pattern looks for a decimal character \\\\d
• at the beginning ^ of a line,
• accept all * characters through the first dot \\\\. (all other than dot: [^\\\\.] )
• and put it in group 1.

The group 2 is filled without spaces \\\\s .

---

I guess you want to have separate attributes for the number and the additional info. If not just concatenate the matches, eg

String lastStreet; // instead of lastCustomer
... 1st if/then:
lastStreet = stMatcher.group(1);
... 2nd if/then:
customerList.add(new Customer(lastStreet
    + " " + stMatcher.group(1)
    + " " + stMatcher.group(2));