JavaScript (spidermonkey) version
Way to carried away with this… here’s the SpiderMonkey version (as opposed to the rhino) one. Biggest complaint is the File object is a bit of a pain to use…
1fd = File('p.test');
2fd.open("read",'text');
3
4mre = /^__MULTI_TOKEN__\s+(\S+)\s+(.*)\t?\s*(\d+)\s*$/;
5sre = /^__SINGLE_TOKEN__\s+(\S+)\s*\t?\s*(\d+)\s*$/;
6
7var ofd = File('full.txt');
8ofd.open('write,create,replace','text');
9first = {}
10
11while (line = fd.readln()) {
12 if (m = line.match(mre)) {
13 first[m[1]] = (first[m[1]] ? first[m[1]] : 0) + parseInt(m[3]);
14 ofd.writeln(m[1] + " " + m[2] + "\t" + m[3]);
15 } else if (m = line.match(sre)) {
16 first[m[1]] = (first[m[1]] ? first[m[1]] : 0) + parseInt(m[2]);
17 } else {
18 print("Unknown: " + line);
19 }
20}
21
22ofd.close()
23
24ofd = File('first.txt');
25ofd.open('write,create,replace','text');
26for (key in first) {
27 ofd.writeln(key + "\t" + first[key]);
28}
29ofd.close()
This is all based on the PHP/Perl/Python performance code.