Last active
November 7, 2017 20:53
-
-
Save phluid61/a5abbdb18295b8369477a5843c0a71d5 to your computer and use it in GitHub Desktop.
EPrints 'reencode' subroutine to heuristically detect certain non-UTF-8 encodings
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# | |
# A percent-encoded URI does not necessarily have to encode UTF-8 | |
# sequences. For example: | |
# | |
# <https://zhidao.baidu.com/question/210190177.html?qbl=relate_question_0&word=%C4%DA%B2%BF%C9%F3%BC%C6%CD%E2%CE%C4%B2%CE%BF%BC% | |
# | |
# The percent-encoded sequence is actually EUC-CN (a Chinese character | |
# encoding). We have no problem in perl when the 'word' parameter is | |
# naively converted to octets, however the database will definitely | |
# complain when we attempt to insert those octets into a field that | |
# expects valid UTF-8 character sequences. | |
# | |
# This function attempts to detect EUC-CN or Big5 character sequences | |
# and transcode them to UTF-8-encoded Unicode. | |
# | |
# As a final fallback, any octets that aren't successfully converted | |
# to UTF-8 are transcoded from Windows-1252. | |
# | |
my $ascii = qr/[\x00-\x7F]/; | |
my $euc = qr/[\xA1-\xF7][\xA1-\xFE]/; | |
my $big5 = qr/[\x81-\xFE]([\x40-\x7E]|[\xA1-\xFE])/; | |
my $euc_string = qr/^($ascii)*(($euc)+($ascii)*)+$/; | |
my $big5_string = qr/^($ascii)*(($big5)+($ascii)*)+$/; | |
my $utf8_2 = qr/[\xC0-\xDF]/; | |
my $utf8_3 = qr/[\xE0-\xEF]/; | |
my $utf8_4 = qr/[\xF0-\xF7]/; | |
my $utf8_n = qr/[\x80-\xBF]/; | |
my $utf8_string = qr/^(($ascii)|($utf8_2)($utf8_n)|($utf8_3)($utf8_n){2}|($utf8_4)($utf8_n){3}$)+$/; | |
sub _reencode { | |
use bytes; | |
STRING: for ( @_ ) | |
{ | |
next STRING unless $_; | |
eval { | |
# Detect EUC-CN/GBK encoding | |
if( $_ =~ $euc_string ) | |
{ | |
$_ = decode('euc-cn', $_, Encode::FB_CROAK); | |
} | |
# Detect Big5 | |
elsif( $_ =~ $big5_string ) | |
{ | |
$_ = decode('big5-eten', $_, Encode::FB_CROAK); | |
} | |
1; | |
}; | |
eval { | |
# Fall back to Windows-1252 (~= ISO-8859-1) for non-UTF-8 | |
if( $_ !~ $utf8_string ) | |
{ | |
$_ = decode('windows-1252', $_, Encode::FB_CROAK); | |
} | |
1; | |
}; | |
} | |
return @_; | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment