http://qs321.pair.com?node_id=1227703


in reply to Re: Handling utf-8 characters when scraping
in thread Handling utf-8 characters when scraping

The code I'm having trouble with works just like the sample code above. I don't want the ellipsis getting to get encoded into \x{2026}.

$PM = "Perl Monk's";
$MCF = "Most Clueless Friar Abbot Bishop Pontiff Deacon Curate Priest Vicar";
$nysus = $PM . ' ' . $MCF;
Click here if you love Perl Monks

  • Comment on Re^2: Handling utf-8 characters when scraping

Replies are listed 'Best First'.
Re^3: Handling utf-8 characters when scraping
by haukex (Bishop) on Dec 26, 2018 at 10:17 UTC
    I don't want the ellipsis getting to get encoded into \x{2026}.

    That's Data::Dumper doing that, and I don't think there's a way to turn it off. Data::Dump seems to be similar. Data::Printer does seem to do what you want:

    use warnings; use strict; use open qw/:std :utf8/; use Data::Printer { print_escapes=>1 }; my $str = "(\N{U+2026}\n)"; p $str;

    Gives me: "(\n)"

    ... although on the other hand, these modules are all debugging tools, and not really tools for generating consistently formatted output. For that, other formats are better - if you could explain the application, then perhaps we could make other suggestions.

      OK, I thought I had ruled out the possibility that it was Dumper doing the encoding but looking at it again, I think you are right.

      For now, I'm just storing the data in a file with Storable. The data will probably end up in a simple database like sql lite eventually. I've been bit in the past by utf8 encoding and my objective for now is to nip the problem in the bud by ensuring the data stays in utf8 from start to finish to avoid any problems.

      Thanks for looking at this and providing assistance.

      $PM = "Perl Monk's";
      $MCF = "Most Clueless Friar Abbot Bishop Pontiff Deacon Curate Priest Vicar";
      $nysus = $PM . ' ' . $MCF;
      Click here if you love Perl Monks

        One way to verify what Perl is really storing internally is Devel::Peek. What I would look for is that the UTF8 flag is on, and the string when shown as UTF-8 is correct:

        use Devel::Peek; my $str = "\x{20AC}"; Dump($str); __END__ SV = PV(0x15c0d70) at 0x15e0440 REFCNT = 1 FLAGS = (POK,IsCOW,pPOK,UTF8) PV = 0x15e4e10 "\342\202\254"\0 [UTF8 "\x{20ac}"] CUR = 3 LEN = 10 COW_REFCNT = 1

        Here, [UTF8 "\x{20ac}"] is correct. There's also utf8::is_utf8($str) to check for the UTF8 flag, although I'd recommend only using that for debugging as well. If you don't want all the extra output, you might just say:

        use Data::Dump; my $str = "\x{20AC}"; dd $str, utf8::is_utf8($str); __END__ ("\x{20AC}", 1)