Commande qui fonctionne en ligne de commande mais pas en script ((standard_in) 1: syntax error)

Bonjour,
J'ai un script extrait des valeurs d'un fichier (speedtest) dont le contenu est du type

Code:

---

1 2 3 4 5 6 7 8 9 10 11 12 13 14

##"ORANGE FRANCE - Puteaux","23884","3.5467","0.32625","0","26577418","8390410","364999650","91138521","https://www.speedtest.net/result/c/c643aa2b-5046-4013-aefe-3fdf7044977f","1","17.9035","26.4743","2.991","257.569","4.27329","2.50788","2.273","57.746","3.194","3.789" ## #!/bin/bash var=$(cat $1) Server_Name=$(echo $var | cut -d"," -f1-1) Server_ID=$(echo $var | cut -d"," -f2-2) Server_Ping=$(echo $var | cut -d"," -f3-3) Download=$(echo $var | cut -d"," -f6-6) Upload=$(echo $var | cut -d"," -f7-7) LienResultat=$(echo $var | cut -d"," -f10-10) ## problème sur cette ligne ! Download=$(echo "scale=3; $Download/ 125000" | bc | awk '{printf "%.2f\n", $0}') #...

---

si j'exécute le script j'obtiens

Code:

---

1 2 3

bash trancheHoraire.sh file.cvs (standard_in) 1: syntax error

---

Si je lance cette commande dans terminal, ça fonctionne !
Avez-vous une idée sur le problème ?

Et hop! Un p'tit coup de debug (trace) pour voir ce qu'il fait:

Code:

---

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

% bash -x trancheHoraire.sh file.cvs ++ cat file.cvs + var='##"ORANGE FRANCE - Puteaux","23884","3.5467","0.32625","0","26577418","8390410","364999650","91138521","https://www.speedtest.net/result/c/c643aa2b-5046-4013-aefe-3fdf7044977f","1","17.9035","26.4743","2.991","257.569","4.27329","2.50788","2.273","57.746","3.194","3.789"' ++ echo '##"ORANGE' FRANCE - 'Puteaux","23884","3.5467","0.32625","0","26577418","8390410","364999650","91138521","https://www.speedtest.net/result/c/c643aa2b-5046-4013-aefe-3fdf7044977f","1","17.9035","26.4743","2.991","257.569","4.27329","2.50788","2.273","57.746","3.194","3.789"' ++ cut -d, -f1-1 + Server_Name='##"ORANGE FRANCE - Puteaux"' [je snippe un peu] ++ echo '##"ORANGE' FRANCE - 'Puteaux","23884","3.5467","0.32625","0","26577418","8390410","364999650","91138521","https://www.speedtest.net/result/c/c643aa2b-5046-4013-aefe-3fdf7044977f","1","17.9035","26.4743","2.991","257.569","4.27329","2.50788","2.273","57.746","3.194","3.789"' ++ cut -d, -f6-6 + Download='"26577418"' [je snippe encore un peu] ++ echo 'scale=3; "26577418"/ 125000' ++ bc ++ awk '{printf "%.2f\n", $0}' (standard_in) 1: parse error + Download=

---

Je teste un peu à la mano pour voir:

Code:

---

1 2	% echo 'scale=3; "26577418"/ 125000' | bc (standard_in) 1: parse error

---

Ça se confirme ! bc n'aime pas qu'on divise une chaine par un nombre...

PS: Par ailleurs, quelle drôle d'idée d'appeler "file.cvs" un fichier qui semble plus "comma separated values" que "Canvas format" :koi:

Merci Jack,
Pourtant ça marche en ligne de commande.
&Download converti mais pas en script.
Comment convertir "1234567" en 1234567 dans le script ?

Bonjour :coucou:

Code:

---

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

$ awk -F '","' '{sub(/^./,"");sub(/.$/,"");} {for (i=1;i<=NF;i++) print $i;}' <<eof > "ORANGE FRANCE - Puteaux","23884","3.5467","0.32625","0","26577418","8390410","364999650","91138521","https://www.speedtest.net/result/c/c643aa2b-5046-4013-aefe-3fdf7044977f","1","17.9035","26.4743","2.991","257.569","4.27329","2.50788","2.273","57.746","3.194","3.789" > eof ORANGE FRANCE - Puteaux 23884 3.5467 0.32625 0 26577418 8390410 364999650 91138521 https://www.speedtest.net/result/c/c643aa2b-5046-4013-aefe-3fdf7044977f 1 17.9035 26.4743 2.991 257.569 4.27329 2.50788 2.273 57.746 3.194 3.789

---

Code:

---

1 2 3 4

$ awk -F '","' '{sub(/^./,"");sub(/.$/,"");} {print $6/12500;}' <<eof "ORANGE FRANCE - Puteaux","23884","3.5467","0.32625","0","26577418","8390410","364999650","91138521","https://www.speedtest.net/result/c/c643aa2b-5046-4013-aefe-3fdf7044977f","1","17.9035","26.4743","2.991","257.569","4.27329","2.50788","2.273","57.746","3.194","3.789" eof 2126.19

---

PS : l'erreur pointée précédemment vient de la différence entre le vide et zéro.

Citation:

---

Envoyé par Flodelarab

Code:

---

$ awk ...

---

---

:ptdr:

Code:

---

1 2 3 4 5

$ IFS=',' read -ra array <<<'"ORANGE FRANCE - Puteaux","23884","3.5467","0.32625","0","26577418","8390410","364999650","91138521","https://www.speedtest.net/result/c/c643aa2b-5046-4013-aefe-3fdf7044977f","1","17.9035","26.4743","2.991","257.569","4.27329","2.50788","2.273","57.746","3.194","3.789"' $ strip() { pat=$1; shift; expr=$@; out=${expr/#$pat/}; echo ${out/%$pat/};} $ strip '"' "${array[0]}" #par exemple, avec le premier élément du tableau ORANGE FRANCE - Puteaux $

---

Finalement grâce à vos conseils j'ai résolu le problème, peut-être d'une manière peu élégante, mais ça fait le job.
Je commence par supprimer les guillemets, ce qui permet d'avoir directement des variables numériques lors des affectations.

Code:

---

1 2	# Jeux de travail "ORANGE FRANCE - Paris","24215","4.2742","0.334","0","24931854","9970410","300317462","138365570","https://www.speedtest.net/result/c/44ec3b6d-130a-49f6-90d1-e71335636d96","1","22.0343","29.4288","4.449","241.625","5.28687","1.84831","3.203","13.79","3.441","4.634"

---

Code:

---

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

#!/bin/bash #supprime les " sed -i 's/"\+//g' $1 var=$(cat $1) #Création des variables à partir du fichier Server_Name=$(echo $var | cut -d, -f1-1) Server_ID=$(echo $var | cut -d, -f2-2) Server_Ping=$(echo $var | cut -d, -f3-3) Download=$(echo $var | cut -d, -f6-6) Upload=$(echo $var | cut -d, -f7-7) LienResultat=$(echo $var | cut -d, -f10-10) #Conversion Mbp et mise au format entier Server_Name=$(echo $Server_Name| cut -c 1-23) Server_Ping=$(echo "scale=3; $Server_Ping" | bc | awk '{printf "%.0f\n", $0}') Download=$(echo "scale=3; $Download/ 125000" | bc | awk '{printf "%.0f\n", $0}') Upload=$(echo "scale=3; $Upload/ 125000" | bc | awk '{printf "%.0f\n", $0}') #Création TimeStamp Jour=$(date +%Y-%m-%d) Heure=$(date +%H:%M:%S) #Calcul tranche horaire HeureTranche=$(echo $Heure | cut -d":" -f1) #Heure premier champs délimité par " " Minutes=$(echo $Heure | cut -d":" -f2 | cut -d":" -f1) #Les minutes sont entre les : if [ $Minutes -ge 45 ] then MinTranche="45" elif [ $Minutes -ge 30 ] then MinTranche="30" elif [ $Minutes -ge 15 ] then MinTranche="15" else MinTranche="00" fi Tranche=$(echo "$HeureTranche:$MinTranche:00") #Formatage du résultat format="%s\t%-s\t%-s\t%-s\t%-s\t%-s-\t%-s\t%-s\t%-s\t%\n" echo $Jour,$Heure,$Tranche,$Server_Name,$Server_ID,$Server_Ping,$Download,$Upload,$LienResultat>$1 sed -i 's/,\+/\t/g' $1 #remplace les virgules par tab

---

Merci à tous pour vos conseils

sais-tu que awk peut faire des calculs arithmétiques ?

Quelques astuces pour les futur scripts :

Code:

---

Download=$(echo "scale=3; $Download/ 125000" | bc | awk '{printf "%.2f\n", $0}')

---

Tu demandes à bc de te donner une précision de 3 décimaux après la virgule,
puis tu convertis en 2 décimaux avec awk.
suffit de demande à bc de te donner 2 décimaux à la base.

Code:

---

1 2 3 4 5 6 7 8

$ bc -q scale=2 10/3 3.33 scale=29 10/3 3.33333333333333333333333333333 $

---

Autre remarque,
tu utilises awk pour printf,
tu peux utiliser printf pour printf

Code:

---

1 2 3 4

$ echo 3.3333333 | xargs printf "%.2f\n" 3.33 $

---

Pour les modos : est-il possible d'enlever le coloriage des balises CODE ?

Citation:

---

est-il possible d'enlever le coloriage des balises CODE ?

---

oui, il faut paramétrer la balise CODE en indiquant x : [ CODE = x ] (sans espaces)

Code:

---

avec du "code" sans coloration syntaxique

---

ce qui peut être pratique pour du pseudo-code uniquement.

Merci l'artiste !

Méthode sauvage:

Code:

---

1 2 3 4

$ echo 'scale=3; "26577418"/ 125000' | bc (standard_in) 1: syntax error $ echo 'scale=3; "26577418"/ 125000' | xargs echo | bc 212.619

---

:mouarf: