如何自动在文本文件中插入字段?

时间:2012-01-09 04:32:34

标签: perl automation

假设您有一堆包含以下行的文件:

{yellow_forest_ant|monsters_insects:2|Yellow forest ant|forestant||5|||10|100|||2|2|15||insect|||||||||};
{small_rabid_dog|monsters_dogs:1|Small rabid dog|forestdog||6|||10|90|||2|2|||canine|||||||||};

并且您希望在第5个和第6个字段之间插入三个字段,其中一些新内容取决于现有字段的内容。

你会如何以自动化的方式做到这一点? 在现有文本文件的行内插入一些动态内容。

我的解决方案(在Perl中):

while(<>) {
  if (/\{(.+?)\};/) {
    my @v= $1 =~ /([^\|\{\}]*?|\{\{.*?\}\})\|/g;
    my @output= (@v[0..4], guessMonsterClass($v[1]), $uniques{$v[0]}, '',@v[5..24]);
    print '{'.join('|',@output)."|};\n";
  } else { print; }
}

虽然我的解决方案有效,但效果不佳。 请改进!

3 个答案:

答案 0 :(得分:1)

如果您的输入不包含转义垂直条,则可以使用splitsplice

while (<>) {
    if (/\{(.+?)\};/) {
        my @v = split /\|/, $1, -1;
        splice @v, 5, 0, guessMonsterClass($v[1]), $uniques{$v[0]}, '';
        print '{', join('|', @v), "};\n";
    } else {
        print;
    }
}

注意使用-1作为split的LIMIT来保持空字段的结尾。捕获所有空字段,因此您无需在print中添加额外的竖线。

答案 1 :(得分:0)

摆脱卷曲的眨眼({};),为你的字段命名(完整,合理):

F1|F2|F3|F4|F5|F6
yellow_forest_ant|monsters_insects:2|"Yellow forest ant"|forestant||5
small_rabid_dog|monsters_dogs:1|"Small rabid dog"|forestdog||6

使用DBIDBD::CSV

  my $dbh = DBI->connect( 'dbi:CSV:', "", "", {
        f_dir        => "../data"
      , csv_sep_char => '|'
      , PrintError   => 0
      , RaiseError   => 1
    });

  my $sth = $dbh->prepare('SELECT * FROM monsters.txt');
  $sth->execute;
  while(my @row = $sth->fetchrow_array()) {
    print '|', join( '|', @row ), "|\n";
  }

  $sth = $dbh->prepare("UPDATE monsters.txt SET F5 = F6 * 2 WHERE F4 = 'forestant'");
  $sth->execute;

  $sth = $dbh->prepare('SELECT * FROM monsters.txt');
  $sth->execute;
  while(my @row = $sth->fetchrow_array()) {
    print '|', join( '|', @row ), "|\n";
  }

输出:

|yellow_forest_ant|monsters_insects:2|Yellow forest ant|forestant||5|
|small_rabid_dog|monsters_dogs:1|Small rabid dog|forestdog||6|
|yellow_forest_ant|monsters_insects:2|Yellow forest ant|forestant|10|5|
|small_rabid_dog|monsters_dogs:1|Small rabid dog|forestdog||6|

答案 2 :(得分:-1)

我建议你用awk处理它。

http://search.cpan.org/search?query=awk&mode=all