Fix GoogleParser first-child selector problem.

[kcy1019]

For now, the GoogleParser can't parse results of files like these:

[PDF]david singer (LoA).pdf home.sogang.ac.kr/.../david%20singer%20(LoA).pd...

In this case, since <span class="_ogd b w xsm">[PDF]</span> is the first child and <a href="..."> is the second child, the parser can't parse the link. However, w/o first-child works very well in both cases(file and non-file cases). I think this is somewhat redundant and should be removed.