perl get-gfs.pl data 中间老中断

雨竹晓风

在windows下用perl  C:\perl\bin\get_gfs.pl data 2015062512 0 192 3 RH:UGRD:VGRD:VVEL 1000_mb:975_mb:950_mb:925_mb:900_mb:850_mb:800_mb:750_mb:700_mb:650_mb:600_mb:550_mb:500_mb:450_mb:400_mb:350_mb:250_mb:200_mb D:\gfs\2015062412 下载资料，中间总是中断，中断的提示是：
error code=6,  problem reading http://nomads.ncep.noaa.gov/pub/data/nccf/com/gfs
/prod/gfs.2015062612/gfs.t12z.pgrb2.0p25.f015.idx  ，但是网站里有gfs.t12z.pgrb2.0p25.f015.idx ，日期也没错。
还有一种是* Recv failure: Connection was reset
* Closing connection 0
error in getting file 56 http://nomads.ncep.noaa.gov/pub/data/nccf/com/gfs/prod/
gfs.2015062512/gfs.t12z.pgrb2.0p25.f012，直接失去接收了。


这到底该怎么解决啊？？

andrewsoong

你的脚本放到其他的linux系统可以么？

雨竹晓风

@andrewsoong  我是在windows下cmd里运行的，因为Grads是在windows下的安装的。脚本就是用的get_gfs.pl 这个脚本，太长了：#!/usr/bin/perl -w
# w. ebisuzaki CPC/NCEP/NWS/NOAA 10/2006
#
# simple script to download gfs files
# inspired by Dan Swank's get-narr.pl script
# this script uses the tecnique described in
#     http://www.cpc.ncep.noaa.gov/products/wesley/fast_downloading_grib.html
#
# arguments:  action YYYYMMDDHH HR0 HR1 DHR VAR_LIST LEV_LIST DIRECTORY
#
#   action = inv  (display inventory of 1st file)
#            data (get data)
#
#   HR0, HR1, DHR: forecast hour parameters
#            do f_hour = HR0, HR1, DHR                                  (fortran)
#            for (f_hour = HR1; f_hour <= HR1; f_hour = f_hour + DHR)   (C)
#
#   VAR_LIST:   list of variable separated by colons, blanks replaced by underscore
#            ex.  HGT:TMP:OZONE
#   LEV_LIST:   list of levesl separated by colons, blanks replaced by underscore
#            ex.  500_mb:sfc
#
#   DIRECTORY:  name of the directory in which to place the files
#
# v1.0  10/2006  who is going to find the first bug?
# v1.0a 10/2006  better help page
# v2.0beta  10/2006  no need for egrep, get_inv.pl and get_grib.pl
# v2.0beta2  10/2006  no need for egrep, get_inv.pl and get_grib.pl
# v2.0beta3  10/2006  update messages, ignore case on matches
# v2.0 1/2009 J.M. Berg: no need for OUTDIR for inventory
# v2.0.1 1/2011 change URL to http://nomads.ncep.noaa.gov
# v2.0.2 10/2011 change bad year check to > 2030
#
#------------ user customization section -----------------------------------------

# location of curl
$curl="/usr/bin/curl";

# the URLs of the inventory and grib must be defined by $URL$inv and $URL$grb
# symbolic variables supported YYYY MM DD HH FHR (forecast hour)
#
# grib2 files from operational nomads server
#
# 1x1 degree GFS
#$URL='http://nomads.ncep.noaa.gov/pub/data/nccf/com/gfs/prod/gfs.$YYYY$MM$DD$HH/gfs.t${HH}z.pgrb2f${FHR}';
# 0.5x0.5 degree GFS
# $URL='http://nomads.ncep.noaa.gov/pub/data/nccf/com/gfs/prod/gfs.$YYYY$MM$DD$HH/master/gfs.t${HH}z.mastergrb2f${FHR}';
# 0.25x0.25 degree GFS
$URL='http://nomads.ncep.noaa.gov/pub/data/nccf/com/gfs/prod/gfs.$YYYY$MM$DD$HH/gfs.t${HH}z.pgrb2.0p25.f${FHR}';
#
#
# grib2 files from www.ftp.ncep.noaa.gov (not operational)
# $URL='http://www.ftp.ncep.noaa.gov/data/nccf/com/gfs/prod/gfs.$YYYY$MM$DD$HH/gfs.t${HH}z.pgrb2f${FHR}';
#
#
$inv='.idx';
$grb='';

$windows='no';
# $windows='yes';
#------------- guts of script ---------------------------------------------------
$version="2.0";
if ($#ARGV != 7) {
  print "get_gfs.pl $version\n";
  print "\nget_gfs.pl action YYYYMMDDHH HR0 HR1 DHR VAR_LIST LEV_LIST DIRECTORY\n\n";
  print "   action = inv  (display inventory of first file)\n";
  print "            data (get data)\n";
  print "   HR0, HR1, DHR: forecast hour parameters\n";
  print "      do f_hour = HR0, HR1, DHR\n";
  print "   VAR_LIST: list of variable separated by colons, blanks replaced by underscore or all\n";
  print "      ex.  HGT:TMP:OZONE, all\n";
  print "   LEV_LIST:   list of levels separated by colons, blanks replaced by underscore or all\n";
  print "      ex.  500_mb:sfc, all\n";
  print "   DIRECTORY:  name of the directory in which to place the files\n";
  print "\n                   EXAMPLES\n\n Reading an inventory\n\n";
  print "       get_gfs.pl inv 2006101800 0 0 0 all all .\n\n";
  print " Downloading 500 hPa Height and Temp 0/3/6 hours forecasts to current directory\n\n";
  print "       get_gfs.pl data 2006101800 0 6 3 HGT:TMP 500_mb .\n\n";
  print " Downloading PRATE 12/18 hours forecasts to current directory\n\n";
  print "       get_gfs.pl data 2006101800 12 18 6 PRATE all .\n\n";
  print " Of course the date code will have to be current.\n";
  exit(8);
}

$action = $ARGV[0];
$time = $ARGV[1];
$hr0=$ARGV[2];
$hr1=$ARGV[3];
$dhr=$ARGV[4];
$VARS=$ARGV[5];
$LEVS=$ARGV[6];
$OUTDIR = $ARGV[7];

$YYYY = substr($time,0,4);
$MM = substr($time,4,2);
$DD = substr($time,6,2);
$HH = substr($time,8,2);

# check values

if ($action ne 'data' && $action ne 'inv') {
   print "action must be inv or data, not $action\n";
   exit(8);
}

if ($YYYY < 2006 || $YYYY > 2030) {
   print "bad date (year) code $time\n";
   exit(8);
}
if ($MM < 1 || $MM > 12) {
   print "bad date (month) code $time\n";
   exit(8);
}
if ($DD < 1 || $DD > 31) {
   print "bad date (day) code $time\n";
   exit(8);
}
if ($HH < 0 || $HH > 23) {
   print "bad date (hour) code $time\n";
   exit(8);
}
if ($hr0 == $hr1) {
   $dhr = 3;
}

if ($dhr != 3 && $dhr != 6 && $dhr != 12 && $dhr != 24) {
   print "dhr must be 3, 6, 12 or 24, not $dhr\n";
   exit(8);
}

if ($hr0 > $hr1) {
   print "hr0 needs to be <= hr1\n";
   exit(8);
}
if ($dhr <= 0) {
   print "dhr needs to be > 0\n";
   exit(8);
}
if (! -d $OUTDIR && $action ne 'inv') {
   print "Directory $OUTDIR does not exist\n";
   exit(8);
}

$VARS =~ tr/:_/| /;
if( $VARS =~ m/ALL/ig ) { $VARS = "."; }
else { $VARS = ":($VARS):"; }

$LEVS =~ tr/:_/| /;
if( $LEVS =~ m/ALL/ig ) { $LEVS = "."; }
else { $LEVS = ":($LEVS)" ; }

$URL =~ s/\$YYYY/$YYYY/g;
$URL =~ s/\${YYYY}/$YYYY/g;
$URL =~ s/\$MM/$MM/g;
$URL =~ s/\${MM}/$MM/g;
$URL =~ s/\$DD/$DD/g;
$URL =~ s/\${DD}/$DD/g;
$URL =~ s/\$HH/$HH/g;
$URL =~ s/\${HH}/$HH/g;

$output = '';

$fhr=$hr0;
while ($fhr <= $hr1) {
   if ($fhr <= 9) { $fhr="00$fhr"; }
    if (($fhr <= 99) && ($fhr >= 10)) { $fhr="0$fhr"; }
   $url = $URL;
   $url =~ s/\$FHR/$fhr/g;
   $url =~ s/\${FHR}/$fhr/g;
   $file = $url;
   $file =~ s/^.*\///;

   #
   # read the inventory
   #    $line[] = wgrib inventory,  $start[] = start of record (column two of $line[])
   #

   if ($windows eq 'yes') {
      $err = system("$curl -f -s $url$inv -o $OUTDIR/$file.tmp");
      $err = $err >> 8;
      if ($err) {
         print STDERR "error code=$err,  problem reading $url$inv\n";
         sleep(10);
         exit(8);
      }
      open (In, "$OUTDIR/$file.tmp");
   }
   else {
      open (In, "$curl -f -s $url$inv |");
   }

   $n=0;
   while (<In>) {
      chomp;
      $line[$n] = $_;
      s/^[^:]*://;
      s/:.*//;
      $start[$n] = $_;
      $n++;
   }
   close(In);
   if ($n == 0) {
       print STDERR "Problem reading file $url$inv\n";
       sleep(10);
       exit(8);
   }

   #
   # find end of record: $last[]
   #

   $lastnum = $start[$n-1];
   for ($i = 0; $i < $n; $i++) {
      $num = $start[$i];
      if ($num < $lastnum) {
         $j = $i + 1;
         while ($start[$j] == $num) { $j++; }
         $last[$i] = $start[$j] - 1;
      }
      else {      
         $last[$i] = '';
      }
   }
   
   if ($action eq 'inv') {
      for ($i = 0; $i < $n; $i++) {
         print "$line[$i]:range=$start[$i]-$last[$i]\n";
      }
      exit(0);
   }

   #
   # make the range field for Curl
   #

   $range = '';
   $lastfrom = '';
   $lastto = '-100';
   for ($i = 0; $i < $n; $i++) {
      $_ = $line[$i];
      if (/$LEVS/i && /$VARS/i) {
         $from=$start[$i];
         $to=$last[$i];

         if ($lastto + 1 == $from) {
            $lastto = $to;
         }
         elsif ($lastto ne $to) {
            if ($lastfrom ne '') {
               if ($range eq '') { $range = "$lastfrom-$lastto"; }
               else { $range = "$range,$lastfrom-$lastto"; }
            }
            $lastfrom = $from;
            $lastto = $to;
        }
      }
   }
   if ($lastfrom ne '') {
      if ($range eq '') { $range="$lastfrom-$lastto"; }
      else { $range="$range,$lastfrom-$lastto"; }
   }

   if ($range ne '') {
      $err = system("$curl -f -v -s -r \"$range\" $url$grb -o $OUTDIR/$file.tmp");
      $err = $err >> 8;
      if ($err != 0) {
         print STDERR "error in getting file $err $url$grb\n";
         sleep(20);
         exit $err;
      }
      rename "$OUTDIR/$file.tmp", "$OUTDIR/$file";
      $output = "$output $OUTDIR/$file";
   }
   else {
      print "no matches (no download) for $file\n";
   }
   $fhr += $dhr;
}
print "\n\nfinished download\n\n$output\n";
exit(0);

schliezer

你好，问题解决了吗？我也遇到了相似问题

雨竹晓风

最后改成遇错不让程序EXIT，但是还是有一些数据不全，这样只能暂时解决下载不中断，还希望PERl高手指点！