Hello, thanks for reading.
My request should be very straight-forward & fairly simple.
As it's just really an installation of the correct/best Server Hosted Browser, with some basic settings (Load Page + Output result).
It's something I can probably implement myself, but it's not my area of experience, so I do not know what is the best setup I should go with.
I've been reading about: Puppeteer, Node.js, Xvfb, PhantomJS, Selenium, cheerio, jsdom, htmlparser2, etc...
but I do not yet know what is the best (proper functionality, easiest and most efficient) setup for my needs.
This is what I'd like to have set up:
(I'm mostly trying to be very specific (so there shouldn't be any unexpected additional requests), but there are not very many steps/complexity to this.)
1a) Have a good Server Hosted Browser (Chrome?), installed on my Dedicated Server.
1b) The Browser/Setup should be of a maintained project, and modern (unlike PhantomJS - now suspended).
1c) The Browser should have both Headless (default) and GUI option, in case I would like to observe/interact.
2a) The Browser auto-runs on Server Boot (with option to disable this).
2b) Ideally if possible, I can have the Browser configs by default be read from a cPanel directory (for easier access/changes), eg: home/username/public_html/
2c) With a Fallback config file in server "root" or wherever appropriate (in case default isn't found).
2d) Allow configuration of Cookies and Header(?)
3b) The browser waits x seconds (or load completion), before moving to the next step.
3c) This may be IMPORTANT - I believe the page in specific, may be using Websocket to pass the variables into the output, so the Browser will need that functionality.
3d) Once assumed to be loaded, the Browser periodically outputs the resulting HTML/content (which should now contain the JS processed values) - Output to: home/username/public_html/folder/output-file-url-#.txt
"#" = Current Timestamp (date)
"url" = Just the Domain + Ending Path of the visited webpage
eg: URL visited: "[login to view URL]"
Output File = "[login to view URL]"
4a) Once the Browser output (HTML) is stored on a file in my server, I expect that I will be able to fetch that output for further processing, using my own PHP code.
4b) I should be able to handle all tasks after output, on my own (reading output, cron-job to clean up files, etc..).
I just need you to d the browser installation for me, have it auto-run (or manual Launch & Stop via URL visit parameter, if possible?)
5a) Provide some way (URL path via my browser? Server port?) where I can run the server hosted browser with GUI option, and observe (& interact?) with the resulting output.
5b) Preferably, this should be a different instance than the default Headless Browser which auto-runs (so it does not stop/interfere with it).
6) If possible, it may be better for me if you simply produce the code I need (and instructions), and I can do the setup on my own.
I'm very familiar with SSH, and code in general, and should be able to complete the setup on my own, with simple guidance and the provided configurations.
My Server details:
CentOS 6 (may soon need upgrade to v7)
Linux 64-bit (I believe)
WHM & cPanel
Fairly Good/High RAM & CPU
I hope my explanation has been concise to you,
I believe there should be no further tasks apart from what is explained above.
And (I believe) there should not be too many steps in getting this done.
It should be a matter of:
- Choose the appropriate setup for my needs
- Provide needed information or files from my setup
- Write the configurations per my needs
- I will do the setup of the server browser myself, I just need: Instructions + Configuration