Extracting info music file

[skelosModerator]

Hello.
I’m making a script for autoidentify music files. I recovered music data from a webservice and I obtains something like this:

{“status”: “ok”, “results”: [{“recordings”: [{“artists”: [{“id”: “4dd6b6cd-7e05-44f2-9d58-7b523a161643”, “name”: “Piccolo coro dell’Antoniano”}], “duration”: 182, “releasegroups”: [{“secondarytypes”: [“Compilation”], “type”: “Album”, “id”: “1dd1ddf3-2cb1-3f07-945b-93490f850a7d”, “title”: “Grandi successi dello Zecchino d’Oro”}], “title”: “Quarantaquattro gatti”, “id”: “9de37de1-925a-47e9-b776-27f6d6cb5730″}, {“artists”: [{“id”: “04ba9285-19d9-4b6f-bd82-357c96130507”, “name”: “Coccodrillo Band”}], “duration”: 187, “releasegroups”: [{“secondarytypes”: [“Compilation”], “type”: “Album”, “id”: “ee0dc7af-8278-493f-a5fb-cdd0a1e15a39″, “title”: “Il coccodrillo come fa??”}], “title”: “44 gatti”, “id”: “ee6e1d2c-b7c3-4fbf-960e-eeefc5e38068″}], “score”: 0.938291, “id”: “a65416d2-2c6f-4ff4-9827-cf55e149e21d”}]}’

to be clear I indent it:

{"status": "ok", "results": 
   [{"recordings": 
      [
       {"artists": 
         [{"id": "4dd6b6cd-7e05-44f2-9d58-7b523a161643", "name": "Piccolo coro dell'Antoniano"}]
         , "duration": 182
         , "releasegroups": 
            [
	     {"secondarytypes": ["Compilation"], 
	      "type": "Album", 
              "id": "1dd1ddf3-2cb1-3f07-945b-93490f850a7d", 
              "title": "Grandi successi dello Zecchino d'Oro"
             }
             ]
         , "title": "Quarantaquattro gatti"
	 , "id": "9de37de1-925a-47e9-b776-27f6d6cb5730"
	}
        , 
       {"artists": 
         [{"id": "04ba9285-19d9-4b6f-bd82-357c96130507", "name": "Coccodrillo Band"}]
         , "duration": 187
         , "releasegroups": 
            [
	     {"secondarytypes": ["Compilation"],
	      "type": "Album", 
              "id": "ee0dc7af-8278-493f-a5fb-cdd0a1e15a39", 
              "title": "Il coccodrillo come fa??"
             }
            ]
         , "title": "44 gatti"
         , "id": "ee6e1d2c-b7c3-4fbf-960e-eeefc5e38068"
        }
       ]
      , "score": 0.938291
      , "id": "a65416d2-2c6f-4ff4-9827-cf55e149e21d"
   }]
}'

and put it into a variable named “txt”. It shows 2 different recordings each of them identified from :
artist, duration , release group (with his name), title and id.

So, I need to extract :
– status (here is “ok” => the song is identified)
– score (here is 0.938291 => there is 0.938291 probability to examined music is this one)

and (only for the first recording if there are more then one)
– Title (here “Quarantaquattro gatti”)
– artist (here “Piccolo coro dell’Antoniano”)
– Album (here “Grandi successi dello Zecchino d’Oro”)

I’m thinking to regular expression but I’m not sure this will work.
Fore recover status value I tried using :

echo $txt | sed 's/^status": ".*",$/1/'

but this dosn’t return nothing…

Anyone has ideas?

• This topic was modified 1 week ago by skelos.
• This topic was modified 1 week ago by skelos.
• This topic was modified 1 week ago by skelos.

[tomlohaveModerator]

can you send all your results on a simple txt file ?
It will be more easy to test.
Cheers,

[skelosModerator]

This is the script for MP3s. Video, zip and other kind files are here not considered. This is my first script and I’m sure that you find many and many not-intelligent instructions: it is done adapting code samples find on the web. I know that same things may be done with code more more smart… please don’t laugh of me. ^_^

#!/bin/bash
APIKey=ALFK2ODc
tollerance=0.90
#file to analyze: in future will be a parameter
FileName=/root/44gatti.mp3
clear
#cvalculate fpcalc
out=$(fpcalc "$FileName")
#adapting result
out2=echo "$out" | tr "n" "&"
out3="${out2/FINGERPRINT/fingerprint}"
out4="${out3/DURATION/duration}"
init="client=$APIKey&meta=recordings+releasegroups+compress&"
#write file to pass to curl
echo "$init$out4">da
#recover webserver response
txt=$(curl --data @da "http://api.acoustid.org/v2/lookup")
#remove da temp file
rm da
#getting the status
status=${txt:12:2}
if [ "$status" = "ok" ]; then
    #getting score
    pos=$(echo "$txt" | awk '{print match($0,"score")}')
    let pos=pos+7
    score=${txt:$pos:7}
    echo "SCORE : $score  THRESHOLD: $tollerance"
    var=$(awk 'BEGIN{ print "'$tollerance'"<"'$score'" }')
    #if score is greather than treshold go on..
    if [ "$var" -eq 1 ]; then
       echo "The file $FileName has been recocngized"
       echo "$txt"
       echo ""
       string=${txt//"/}
       #try to extract title
       expr match "$string" '.*(title.*,)'
    else
       echo "The file $FilenNme it is recognized but i'm not sure"
    fi
else
    echo "The file $FileName has not been recognised."
fi

This is the output:

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  4558  100   810  100  3748   1864   8627 --:--:-- --:--:-- --:--:-- 11676
SCORE : 0.93829  THRESHOLD: 0.90
The file /root/44gatti.mp3 has been recocngized
{"status": "ok", "results": [{"recordings": [{"artists": [{"id": "4dd6b6cd-7e05-44f2-9d58-7b523a161643", "name": "Piccolo coro dell'Antoniano"}], "duration": 182, "releasegroups": [{"secondarytypes": ["Compilation"], "type": "Album", "id": "1dd1ddf3-2cb1-3f07-945b-93490f850a7d", "title": "Grandi successi dello Zecchino d'Oro"}], "title": "Quarantaquattro gatti", "id": "9de37de1-925a-47e9-b776-27f6d6cb5730"}, {"artists": [{"id": "04ba9285-19d9-4b6f-bd82-357c96130507", "name": "Coccodrillo Band"}], "duration": 187, "releasegroups": [{"secondarytypes": ["Compilation"], "type": "Album", "id": "ee0dc7af-8278-493f-a5fb-cdd0a1e15a39", "title": "Il coccodrillo come fa??"}], "title": "44 gatti", "id": "ee6e1d2c-b7c3-4fbf-960e-eeefc5e38068"}], "score": 0.938291, "id": "a65416d2-2c6f-4ff4-9827-cf55e149e21d"}]}
title: 44 gatti, id: ee6e1d2c-b7c3-4fbf-960e-eeefc5e38068}], score: 0.938291,
#

I would find a regular espression that return
title: 44 gatti,
and use it also for return other parts of this string like :
“name”: “Coccodrillo Band”
“score”: 0.938291
and so on…

[Author]

Posts