Why doesn't the parser parse links?

Question

Java

I merge content from the Lenfilm.tv website. I got to links to files that lead to a file hosting service, for example, here - idwl.cn/file/428924/_SOFTBOX__Menyayuschii_igru.mk... There is only one div with id = download-link in the page code. It has only one a with id = dwl. In fact, I only need to take the href value and that's it. But this way I get a link not to a file, but to the same page.
I want to draw your attention to the fact that the page code along with the link (correct, verified) is displayed in the Via browser. Like this:

<div class="arrow" id="download-link">
            <a style="text-decoration: none; color: green;" id="dwl" href="http://p5.idwl.cn/secure/S-O8cm7_esbK0yAva7uUvA/1493530973/4290/428923/_SOFTBOX__Menyayuschii_igru.mkv_320p_320x240.mp4">СКАЧАТЬ
            </a>
</div>

I decompiled this browser but did not understand at all what is used there, I use JSOUP. By the way, here is the code.

String download = downpage.getElementById("download-link").getElementById("dwl").absUrl("href");

I'm going crazy, what could be the reason? By the way, here is some suspicious JS script (in the same div):

<script>
    document.getElementById('dwl').href = obfuscate('.10816.13456.13456.12544.3364.2209.2209.12544.2809.2116.11025.10000.14161.11664.2116.9801.12100.2209.13225.10201.9801.13689.12996.10201.2209.6889.2025.6241.3136.9801.11881.3025.9025.10201.13225.9604.5625.2304.14641.4225.13924.9409.3025.13689.7225.13924.4225.2209.2401.2704.3249.2601.2809.2601.2304.3249.3025.2601.2209.2704.2500.3249.2304.2209.2704.2500.3136.3249.2500.2601.2209.9025.6889.6241.4900.7056.4356.6241.7744.9025.9025.5929.10201.12100.14641.9409.14641.13689.13225.9801.10816.11025.11025.9025.11025.10609.12996.13689.2116.11881.11449.13924.9025.2601.2500.2304.12544.9025.2601.2500.2304.14400.2500.2704.2304.2116.11881.12544.2704');
        </script>

Apparently the URL is written on the fly, but how? What are these numbers? And why does Via show the code perfectly, along with a link? Maybe you should study how Via works?

Answer 1

There the link is obfuscated quite funny. For deobfuscation, the string is divided by points, the square root of each number from the resulting array is taken and converted into the unicode character corresponding to this number. Something like this - https://jsfiddle.net/ycbxfh9e/. The code for deobfuscation is primitive, it will not be difficult to repeat it in java. Why in the via browser is given a non-fussified link - xs. Maybe the user agent determines who to show what. I tried the user agent from opera mini - the result is the same. You can copy it from via and try again.