(2023-01-23, 04:46)rmtonin Wrote: Hello, thanks for the awesome plugin.

I'm a bit confused, I have installed the Arctic Zephyr Reloaded skin but I'm getting only list and thumbnail views. Am I doing something wrong?

(2023-01-23, 12:43)mcarni Wrote: hi @chrisism ,

regarding 1)
skin issue is really not a problem, it was just for info, do not worry about it...

regarding 2)
this is a link to the gamelist.xml
link
i didn't modify the frontend parameters so Skyscraper should have used the standard output suitable for EmulationStation
i saw that with Skyscraper i could change the output to better fit to a different frontend, but i have not played with it (Skyscraper Frontend)

my coding skills are limited, but if you want i could try to see if i can generate a bash script that takes the above gamelist and converts it into something that is ok for AKL (i never used xml as output, i would feel more comfortable with json... but i could try ;-) no promises though...)
can you maybe give me an example of the nfo structure that needs to be respected?

thanks a lot

M

• Added new 'View ROM' details page. 
  
• Updated dependencies.
  
• Separated overwriting of existing metadata and assets with scraping.
  
• Fixes for deleting categories.
  
• Small database fixes.
  

bash:
#! /bin/bash



# for the moment we hardcode the name of the gamelist

#gameList=$PWD"testGamelist.xml"

gameList="testGamelist.xml";



# Skyscraper standard output fields, initialise them as empty arrays

declare -a path;

declare -a name;

declare -a thumbnail;

declare -a image;

declare -a marquee;

declare -a video;

declare -a rating;

declare -a desc;

declare -a releasedate;

declare -a developer;

declare -a publisher;

declare -a genre;

declare -a players;

declare -a kidgame;



# based on https://stackoverflow.com/questions/893585/how-to-parse-xml-in-bash

# let's define a function that will parse the xml file and

# assign ENTITY and CONTENT to anything between > and <

# i.e. in <tag>value</tag> --> ENTITY=tag and CONTENT=value

read_dom () {

    local IFS=\>

    read -d \< ENTITY CONTENT

}





while read_dom; do

    case "$ENTITY" in

        "path")         path+=("$CONTENT") ;;

        "name")         name+=("$CONTENT") ;;

        "thumbnail")    thumbnail+=("$CONTENT") ;;

        "image")        image+=("$CONTENT") ;;

        "marquee")      marquee+=("$CONTENT") ;;

        "video")        video+=("$CONTENT") ;;

        "rating")       rating+=("$CONTENT") ;;

        "desc")         desc+=("$CONTENT") ;;

        "releasedate")  releasedate+=("$CONTENT") ;;

        "developer")    developer+=("$CONTENT") ;;

        "publisher")    publisher+=("$CONTENT") ;;

        "genre")        genre+=("$CONTENT") ;;

        "players")      players+=("$CONTENT") ;;

        "kidsgame")     kidsgame+=("$CONTENT") ;;

    esac

done < $gameList



gameQty=${#path[@]};

echo "processed gamelist contains: ""$gameQty"" game(s)"



# change from full release date (in the original xml) to only year (as in the .nfo file)

declare -a year;

for (( i=0 ; i<=($gameQty-1) ; i++ )); do

    temp=${releasedate[$i]};

    year=${temp:0:4};

done

# something similar needs to be done for the rating...

# not sure if skyscraper uses x/10 or /20 or whatever





# this is currently using the structure of the dr_mario.nfo example, can be suited to what is needed



# implement some checks not to overwrite existing files

# if "${name[$i]}".nfo already exists (2 game with the same name) don't overwrite them



# maybe get the platform from the folder location i.e. atari2600/roms --> platform = atari2600



# marquees saved under clearlogos?

# image (screenshot) saved under fanart?

# better change the asset location in AKL ?



for (( i=0 ; i<=($gameQty-1) ; i++ )); do

    echo "game"$(($i+1))": "${name[$i]}" generating nfo file";

    echo "<?xml version="'"1.0"'" encoding="'"utf-8"'" standalone="'"yes"?>' $'\n'\

'<game>' $'\n'\

  '<title>'"${name[$i]}"'</title>' $'\n'\

  '<year>'"${year[$i]}"'</year>' $'\n'\

  '<genre>'"${genre[$i]}"'</genre>' $'\n'\

  '<publisher>'"${publisher[$i]}"'</publisher>' $'\n'\

  '<rating>'"${rating[$i]}"'</rating>' $'\n'\

  '<plot>'"${desc[$i]}"'</plot>' $'\n'\

'</game>' $'\n'\

> "${name[$i]}".nfo

done

(2023-01-26, 13:20)mcarni Wrote: @chrisism ,

just a quick message to let you know that the script worked...
i am amazed it did, but it actually worked...
i took the generated nfo files and moved them where the roms are ( can make this automatic... shouldn't be difficult)
then asked AKL to scan local file and by magic atari2600 roms have some metadata
for the moment I have:   <title>; <year>; <genre>; <publisher>; <rating>; <plot>

I need to convert ratings from whatever format screensaver is using to AKL standard

and I need to understand how to do the same with assets.
(i probably need also to understand what is what, skyscraper uses covers, marquees, screenshots and wheels...)
but if all fails i can always point AKL to the right folder

anyway, that's it for this quick update

thanks a lot for your work

m

(2023-01-23, 11:44)chrisism Wrote:

(2023-01-21, 14:25)Juppstein Wrote:  

And with existing games it did get and download the files etc?
Can you try again with both AKL and MobyGames plugin set on Debug level? Then share the logs again with me. Double check the logs to see if now it is still the HTTP 401 issue or simply not finding the game at MobyGames. Sometimes it is strict with the search input. I haven't encountered this before and I am not aware of maybe some special permissions settings at MobyGames. In the past I have encountered this when I didnt enter my API key completely, missed out a '='. Sensitive data can be shared through private message of course, but the logs itself should be clean.

Quote:"Just because you can write your own parser, doesn't mean you should. – Stephen Niedzielski"

(2023-02-12, 07:20)ippytick Wrote: Hi, I'm setting up a new kodi instance running version 20 (Nexus). In the past I have used advanced emulator launcher for games and browsers but being that there isn't a release for the latest version of Kodi I've decided to try AKL.

Today I have managed to setup all my games to run with the Arctic Zephyr reloaded skin and the retroplayer emulator which I am happy to report all works well. However when trying to launch the firefox browser to a specific website I am at a bit of a loss.

How does one setup a launcher to acheive this? I am sure it is the Rom Collection (standalone option) but I can't work it out. Once upon a time I used to create a shortcut file to the browsers .exe file then include the url in the shortcuts target and then link AEL to that file.

1. Add new ROM (Standalone)
   
2. Select the appropriate category
   
3. Answer “No” to is it a file based ROM/executable
   
4. Write the name for the launcher “Firefox”
   
5. Go into the context menu of the newly created launcher
   
6. Edit ROM
   
7. Add new launcher to ROM
   
8. Select Application launcher
   
9. Find and select the firefox.exe
   
10. In the Application arguments enter the website I wanted firefox to open to “www.example.com”
    
11. Edit metadata/assets as appropriate
    

(2023-02-03, 10:15)Juppstein Wrote:

(2023-01-23, 11:44)chrisism Wrote:

(2023-01-21, 14:25)Juppstein Wrote:  

And with existing games it did get and download the files etc?
Can you try again with both AKL and MobyGames plugin set on Debug level? Then share the logs again with me. Double check the logs to see if now it is still the HTTP 401 issue or simply not finding the game at MobyGames. Sometimes it is strict with the search input. I haven't encountered this before and I am not aware of maybe some special permissions settings at MobyGames. In the past I have encountered this when I didnt enter my API key completely, missed out a '='. Sensitive data can be shared through private message of course, but the logs itself should be clean.

Not sure if you did something in a recent update but the error message is gone now when scraping with MobyGames. Looking at the log it just says that it did not find a match.So now when I try to scrape a new game it will always fail to find a match without a message or a comment from the scraper. Not sure if MobyGames is extra picky there or whatever. What I have to do then is go to Edit ROM and then select either Edit Metadata or Edit Assets and try to scrape from there. Then the scraper will give me a form where I can edit the search term for MobyGames. Given a good name it will find the game and then will ask for the different entries if there are multiple files, in the case for assets, for it.  Same for metadata. The tedious part here is that you basically have to scrape a game twice before a data set is complete. Would be more convenient if there would be a separate search where you can alter the search terms before the search is triggered. If someone has a few hundred games to scrape like that, the above method would take quite the time to complete.

Edit: Oh, and is there a setting that would allow deleting an entry from the database, including the actual game files on disk?

(2023-02-07, 20:06)mcarni Wrote: @chrisism  and guys,

in case it can be of some help, i can bring some feedback on my script to parse xml gamelist and AKL in general.

I stumbled upon a couple of cases where my script would struggle.
I guess this was mainly when in the name or in the path of the games there were special characters like *./|\ etc...
I tried to fight this but in the end i changed strategy and followed what one of the guys on the original SO thread said:

Quote:"Just because you can write your own parser, doesn't mean you should. – Stephen Niedzielski"

So I installed xmlstarlet and rewrote my script so that it uses xmlstartlet instead of trying to read the xml gamelist with IFS and read.

I tested it with atari2600, atari 5200, megadrive, gameboy (original + color + advance), nes, snes, nintendo 64, mame and fba and it works.
Generating nfo files became much slower but it works definitvely better, if anyone want to take a look you can find it here

thanks a lot to @chrisism , the addon works very well.
I have an issue only with c64, for some reason it cannot scan existing roms. (d64 files)
I says critical error and it stops, it is quite peculiar since it seems to me that i have done exactly the same that I have done for the other roms...

I also noticed that on my raspberrypi the scraping of my local nfo files failed for some platforms.
I could not find any special difference with the other platforms where it worked, apart from being the platforms with the highest quantity of roms (mame, nes etc)
So i proceeded to split the roms in 2 (in order to have approx. max. 2000 roms per folder) and after that the scraping worked fine.
I am not sure if this is a limitation of the raspberrypi or of the addon, anyhow i thought it might have helped to highlight it.

thanks a lot

M