Funções de manipulação de arquivos csv em shell script.

Funções para abrir, criar, pegar número de linhas, pegar quantidade de campos, pegar nomes dos campos, adicionar campos, adicionar linha, deletar linha, atualizar linha e outras ainda em desenvolvimento.

Quem quiser contribuir... desde já deixo meus agradecimentos.

Valeu
Gabriel Fernandes

• Download fernax.libcsv
• Enviar nova versão

Esconder código-fonte

#!/bin/bash
############################################################################
#    Fernax - Biblioteca de funcoes de apoio a shell script                #
#    libcsv 0.1 - Funcoes de manipulacao de arquivos CSV                   #
#    Copyright (C) 2010 Gabriel Fernandes                                  #
#    Autor: Gabriel Fernandes <gabriel@duel.com.br>                        #
#                                                                          #
#    This program is free software; you can redistribute it and#or modify  #
#    it under the terms of the GNU General Public License as published by  #
#    the Free Software Foundation; either version 2 of the License, or     #
#    (at your option) any later version.                                   #
#                                                                          #
#    This program is distributed in the hope that it will be useful,       #
#    but WITHOUT ANY WARRANTY; without even the implied warranty of        #
#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the         #
#    GNU General Public License for more details.                          #
#                                                                          #
#    You should have received a copy of the GNU General Public License     #
#    along with this program; if not, write to the                         #
#    Free Software Foundation, Inc.,                                       #
#    59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.             #
############################################################################

csv_openfile () 
{
  CSV_FILENAME=$1
  CSV_TITLE=$2
  CSV_IFS=$3
  
  [ -z "$1" ] && exit 1
  [ -z "$2" ] && TITLE="FALSE"
  [ -z "$3" ] && CSV_IFS=";"
  
  if [ -e ${CSV_FILENAME} ] ; then
    export CSV_FILENAME=${CSV_FILENAME}
    export CSV_TITLE=${CSV_TITLE}
    export CSV_IFS=${CSV_IFS}
    echo ${CSV_FILENAME}
    return 0
  else 
    return 1
  fi
}

csv_createfile ()
{
  CSV_FILENAME=$1
  CSV_TITLE=$2
  CSV_IFS=$3
  
  [ -z "$1" ] && exit 1
  [ -z "$2" ] && TITLE="FALSE"
  [ -z "$3" ] && CSV_IFS=";"
  
  touch ${CSV_FILENAME}
  
  if [ -e ${CSV_FILENAME} ] ; then
    export CSV_FILENAME=${CSV_FILENAME}
    export CSV_TITLE=${CSV_TITLE}
    export CSV_IFS=${CSV_IFS}
    echo ${CSV_FILENAME}
    return 0
  else 
    return 1
  fi
}

csv_getnumrows () 
{
  CSV_NUM_ROWS=$(cat ${CSV_FILENAME} |  wc -l)
  if [ "${CSV_TITLE}" = "TRUE" ] ; then
    CSV_NUM_ROWS=$(let CSV_NUM_ROWS-=1)
  fi

  if [ "$?" = "0" ] ; then
    export CSV_NUM_ROWS=${CSV_NUM_ROWS}
    echo ${CSV_NUM_ROWS};
    return 0
  else
    return 1
  fi
}

csv_getnumfields () 
{

  [ -z "$CSV_FILENAME" ] && exit 1
  [ -z "$CSV_TITLE" ] && TITLE="FALSE"
  [ -z "$CSV_IFS" ] && CSV_IFS=";"

  CSV_FIRST_LINE=$(head -n1 ${CSV_FILENAME})
  echo ${CSV_FIRST_LINE} > /tmp/character_count.$$

  LENGHT=$(echo $CSV_FIRST_LINE | wc -c)
  let LENGHT-=1
 
  COUNTER=1
  CSV_NUM_FIELDS=0
  while [ ${COUNTER} -le ${LENGHT} ] ; do
    CSV_FIRST_LINE=$(cut -c${COUNTER} /tmp/character_count.$$)
    if [ "${CSV_FIRST_LINE}" == "${CSV_IFS}" ] ; then
      let CSV_NUM_FIELDS+=1
    fi
    let COUNTER+=1
  done

  if [ "$?" = "0" ] ; then
    export CSV_NUM_FIELDS=${CSV_NUM_FIELDS}
    echo ${CSV_NUM_FIELDS};
    return 0
  else
    return 1
  fi

}

csv_getfieldtitle () 
{
  CSV_FIELD=$1

  [ -z "$CSV_FILENAME" ] && exit 1
  [ -z "$CSV_TITLE" ] && TITLE="FALSE"
  [ -z "$CSV_IFS" ] && CSV_IFS=";"

  if [ "${CSV_TITLE}" = "TRUE" ] ; then
    
    CSV_FIRST_LINE=$(head -n1 ${CSV_FILENAME})
    NUM_FIELDS=$(csv_getnumfields)
    FIELDS=""
    if [ -z "$CSV_FIELD" ] ; then
      for COUNTER in $(seq ${NUM_FIELDS}); do 
        if [ "${FIELDS}" = "" ] ; then
          FIELDS=${COUNTER}
        else
          FIELDS="${FIELDS},${COUNTER}"
        fi
      done

    else
      FIELDS="${CSV_FIELD}"
      
    fi
      
    CSV_TITLE=$(echo ${CSV_FIRST_LINE} | cut -d "${CSV_IFS}" -f${FIELDS})

    if [ "$?" = "0" ] ; then
      export CSV_TITLE=${CSV_TITLE}
      echo ${CSV_TITLE}
      return 0
    else
      return 1
    fi
  else
    return 1
  fi
  



}

csv_addfield ()
{
  NUM_FIELD=$1
  [ -z "$1" ] && NUM_FIELD=$(csv_getnumfields) 
}

#csv_delfield ()
#{

#}

csv_addrow () 
{
  [ -z "$1" ] && exit 1
  [ -z "$CSV_FILENAME" ] && exit 1
  [ -z "$CSV_TITLE" ] && TITLE="FALSE"
  [ -z "$CSV_IFS" ] && CSV_IFS=";"

  CSV_ROW=""
  
  while [ "$1" != "" ]; do
    if [ ${CSV_ROW} = "" ] ; then
      CSV_ROW="${1}${CSV_IFS}"
    else
      CSV_ROW="${CSV_ROW}${1}${CSV_IFS}"
    fi
    shift
  done
  
  echo "${CSV_ROW}" >> "${CSV_FILENAME}"
  
  return $?

}

csv_delrow () 
{

  # params: chave1 chave2
  # problema: chaves sempre em primeiro no arquivo

  [ -z "$1" ] && exit 1
  [ -z "$CSV_FILENAME" ] && exit 1
  [ -z "$CSV_TITLE" ] && TITLE="FALSE"
  [ -z "$CSV_IFS" ] && CSV_IFS=";"

  CSV_ROW_KEYS=""
  
  while [ "$1" != "" ]; do
    if [ ${CSV_ROW_KEYS} = "" ] ; then
      CSV_ROW_KEYS="${1}${CSV_IFS}"
    else
      CSV_ROW_KEYS="${CSV_ROW_KEYS}${1}${CSV_IFS}"
    fi
    shift
  done
  
  CSV_ID_ROW_DELETE=$(grep -n "^${CSV_ROW_KEYS}" "${CSV_FILENAME}" | cut -d : -f 1)

  sed -i~ ${CSV_ID_ROW_DELETE}d "${CSV_FILENAME}"

  return $?

}

csv_updaterow () 
{
  # params: chave1 chave2 ... SET 5 novoValor
  # problema: chaves sempre em primeiro no arquivo

  [ -z "$1" ] && exit 1
  [ -z "$CSV_FILENAME" ] && exit 1
  [ -z "$CSV_TITLE" ] && TITLE="FALSE"
  [ -z "$CSV_IFS" ] && CSV_IFS=";"

  CSV_ROW_KEYS=""
  
  while [ "$1" != "SET" ]; do
    if [ "${CSV_ROW_KEYS}" = "" ] ; then
      CSV_ROW_KEYS="${1}${CSV_IFS}"
    else
      CSV_ROW_KEYS="${CSV_ROW_KEYS}${1}${CSV_IFS}"
    fi
    shift
  done

  CSV_ID_FIELD_UPDATE="$2"
  CSV_VAL_FIELD_UPDATE="$3"

  CSV_ID_ROW_FOR_UPDATE=$(grep -n "^${CSV_ROW_KEYS}" "${CSV_FILENAME}" | cut -d : -f 1)
  CSV_ROW_FOR_UPDATE=$(grep "^${CSV_ROW_KEYS}" "${CSV_FILENAME}")

  CSV_NUM_FIELDS=$(csv_getnumfields)
  
  CSV_ROW_UPDATED=""

  for ((I=1;I<=${CSV_NUM_FIELDS};I++)) ; do
    if [ "${CSV_ID_FIELD_UPDATE}" != "${I}" ] ; then
      CSV_ROW_FIELD_TEMP=$(echo "${CSV_ROW_FOR_UPDATE}" | cut -d "${CSV_IFS}" -f ${I})
      
      if [ "${CSV_ROW_UPDATED}" = "" ] ; then
        CSV_ROW_UPDATED="${CSV_ROW_FIELD_TEMP}${CSV_IFS}"
      else
        CSV_ROW_UPDATED="${CSV_ROW_UPDATED}${CSV_ROW_FIELD_TEMP}${CSV_IFS}"
      fi
      
    else  
      CSV_ROW_UPDATED="${CSV_ROW_UPDATED}${CSV_VAL_FIELD_UPDATE}${CSV_IFS}"
    fi
  done; 

  sed -i ${CSV_ID_ROW_FOR_UPDATE}d "${CSV_FILENAME}"

  echo "${CSV_ROW_UPDATED}" >> "${CSV_FILENAME}"

  return $?

}

#csv_sortby () 
#{
#}

#csv_locate () 
#{
#}

Popular Mysql

Filtrando processos facilmente com psaux ( ps aux + grep + wc -l)

Claro 3G no Ubuntu 8.04

df (simplificado)

Cronometro em shell

Nenhum comentário foi encontrado.