how to extract metadata from AFP file?

Question

i have a sample AFP file.. want to read TLE(Tag Logical Element) from it to get certain predefined tags such as account no or bill number so that i can convert that to PDF for particular bill no. Also is there any public Java api for performing the same action.

Answer 1

There are two different types of TLE: group level and page level. Group level TLE appear right after a BNG (Begin Named Group), while page level TLEs appear after the environment group of a page (BPG). The TLE itself uses two triplets to carry the meta data: FullyQualifiedName and AttributeValue. The first contains the key name, and the other one the value.

With the assumption that your AFP is well-formed following code helps you extracting the meta data from TLEs:

import java.io.BufferedInputStream;
import java.io.FileInputStream;
import java.io.IOException;
import java.util.Optional;

import org.afplib.afplib.*;
import org.afplib.base.*;
import org.afplib.io.AfpInputStream;

public class DumpTLE {

public static void main(String[] args) {
    try (AfpInputStream in = new AfpInputStream(
            new BufferedInputStream(new FileInputStream(args[0])))) {

        int page = 0;
        int group = 0;
        boolean pageLevel = false;
        SF sf;
        while((sf = in.readStructuredField()) != null) {

            if(sf instanceof BPG) {
                page++;
                pageLevel = true;
            }
            if(sf instanceof EPG) {
                pageLevel = false;
            }
            if(sf instanceof BNG) {
                group++;
            }

            if(sf instanceof TLE) {
                TLE tle = (TLE) sf;

                Optional<FullyQualifiedName> fqn = tle.getTriplets().stream()
                        .filter(FullyQualifiedName.class::isInstance).map(FullyQualifiedName.class::cast)
                        .findFirst();
                Optional<AttributeValue> value = tle.getTriplets().stream()
                        .filter(AttributeValue.class::isInstance).map(AttributeValue.class::cast)
                        .findFirst();

                if(fqn.isPresent() && value.isPresent()) {
                    if(pageLevel) {
                        System.out.println("page "+page+" - "+fqn.get().getFQName()+":"+value.get().getAttVal());
                    } else {
                        System.out.println("group "+group+" - "+fqn.get().getFQName()+":"+value.get().getAttVal());
                    }
                }
            }
        }
    } catch (IOException e) {
        e.printStackTrace();
    }

}

}

The code uses afplib (https://github.com/yan74/afplib). Converting your AFP to PDF is a totally different story and not that easy.