如何优化下面的程序Text 转csv

schrosourcexuey

sed 's/,//g' $1 > test.txt
sed 's/D E S C R I P T I O N/D E S C R I P  T I O N /g' test.txt > testA.txt
sed 's/LAST YEAR/LAST-YEAR/g' testA.txt > testB.txt
rm -f $1.csv
touch $1.csv
while read i
do
HOMEOWNER=`echo $i | grep HOMEOWNER | wc -l`; export HOMEOWNER
FEES=`echo $i | grep FEES | wc -l`; export FEES
RENTAL=`echo $i | grep RENTAL | wc -l`; export RENTAL
YEAR=`echo $i | grep YEAR | wc -l`; export YEAR
ACTUALS=`echo $i | grep ACTUALS | wc -l`; export ACTUALS
if [ $HOMEOWNER -gt 0 -o $FEES -gt 0 -o $RENTAL -gt 0 -o $ACTUALS -gt 0 ]; then
   DESCRIPT=`echo "$i" | cut -c1-36`; export DESCRIPT
   RESTST=`echo "$i" | cut -c37- | awk -v FIELDWIDTHS='13 14 13 14 17 13 13' -v OFS=',' '{$1=$1;print }'` ; export RESTST
   echo "$DESCRIPT"",""$RESTST">> testC.txt
elif [ $YEAR -gt 0 ]; then
   RESTST=`echo "$i" | awk -v FIELDWIDTHS='13 14 13 14 17 13 13' -v OFS=',' '   {$1=$1;print }'` ; export RESTST
   echo ",""$RESTST">> testC.txt
else
   echo "$i" >> testC.txt
fi
done < testB.txt
tr -d '\r' < testC.txt > $1.csv

waker

建议给出源数据例子/目的数据例子/转化规律

rdcwayx

schrosourcexuey 发表于 2012-06-10 19:55
sed 's/,//g' $1 > test.txt
sed 's/D E S C R I P T I O N/D E S C R I P  T I O N /g' test.txt > testA ...

大概看懂了你的意思，但是代码写得实在太烂。

其实只要提供源文件，和期望的输出，大家就可以直接帮你。

下面的代码是按照你现有的思路改了，无需多个临时文件，无需while 循环。我代码中的testB其实是可以不需要的，但是没看到原文件， 先这样用吧。

1. rm -f $1.csv
   
2. 
   
3. awk '{gsub(/,/,"");gsub(/D E S C R I P T I O N/,"D E S C R I P  T I O N ");gsub(/LAST YEAR/,"LAST-YEAR")}1' $1 > testB.txt
   
4. 
   
5. awk '/HOMEOWNER/||/FEES/||/RENTAL/||/YEAR/||/ACTUALS/{ DESCRIPT=substr($0,0,35); RESTST=substr($0,37) }       # You need explain how to get RESTST value
   
6.        /YEAR/{ RESTST="HERE is not clear enough." }                # You need explain how to get RESTST value
   
7.      
   
8.        print DESCRIPT, RESTST
   
9.       }' OFS="," testB.txt > $1.csv

复制代码

xiaopan3322

text的内容呢？