fix: fails to decode PDFDocEncoding-encoded Strings in metadata

[TPKato]

Related to: #459.
May also be related to #110.

Problem

The Document::load_metadata() function does not decode the string correctly if it is encoded using PDFDocEncoding.

According to PDF 1.7, Section 14.3.3, the type of title, author, etc., in the document information dictionary shall be a text string. That is,

bytes that represent characters that shall be encoded using either PDFDocEncoding or UTF-16BE with a leading byte-order marker
(PDF 1.7, Section 7.9.1)

Fix

use common_data_structures::decode_text_string() to decode strings (#297)

Example

Example PDF with a PDFDocEncoding-encoded document information dictionary:
http://mirrors.ctan.org/info/l2tabu/german/l2tabu.pdf

Code for test:

use std::path::Path;
use lopdf::Document;

fn main() {
    let file = "l2tabu.pdf";
    let m = Document::load_metadata(Path::new(file)).expect("load_metadata failed");
    println!("Title:  {:?}", m.title.unwrap());
    // println!("Author: {:?}", m.author.unwrap());

Result: (lopdf = "0.39.0")

Title:  "Das LaTeX2e-S�ndenregister oder Veraltete Befehle, Pakete und andere Fehler"

Expected/Fixed:

Title:  "Das LaTeX2e-Sündenregister oder Veraltete Befehle, Pakete und andere Fehler"

Disclaimer

I am a complete beginner to Rust, so I'm unsure if this patch is appropriate. Please feel free to rewrite it if necessary.

[J-F-Liu]

Thanks, please change Some(common_data_structures::decode_text_string(obj).unwrap()) to common_data_structures::decode_text_string(obj).ok()