summaryrefslogtreecommitdiffstats
path: root/intl/uconv/tools/parse-mozilla-encoding-table.pl
diff options
context:
space:
mode:
authorMatt A. Tobin <mattatobin@localhost.localdomain>2018-02-02 04:16:08 -0500
committerMatt A. Tobin <mattatobin@localhost.localdomain>2018-02-02 04:16:08 -0500
commit5f8de423f190bbb79a62f804151bc24824fa32d8 (patch)
tree10027f336435511475e392454359edea8e25895d /intl/uconv/tools/parse-mozilla-encoding-table.pl
parent49ee0794b5d912db1f95dce6eb52d781dc210db5 (diff)
downloadUXP-5f8de423f190bbb79a62f804151bc24824fa32d8.tar
UXP-5f8de423f190bbb79a62f804151bc24824fa32d8.tar.gz
UXP-5f8de423f190bbb79a62f804151bc24824fa32d8.tar.lz
UXP-5f8de423f190bbb79a62f804151bc24824fa32d8.tar.xz
UXP-5f8de423f190bbb79a62f804151bc24824fa32d8.zip
Add m-esr52 at 52.6.0
Diffstat (limited to 'intl/uconv/tools/parse-mozilla-encoding-table.pl')
-rw-r--r--intl/uconv/tools/parse-mozilla-encoding-table.pl92
1 files changed, 92 insertions, 0 deletions
diff --git a/intl/uconv/tools/parse-mozilla-encoding-table.pl b/intl/uconv/tools/parse-mozilla-encoding-table.pl
new file mode 100644
index 000000000..d68b18e92
--- /dev/null
+++ b/intl/uconv/tools/parse-mozilla-encoding-table.pl
@@ -0,0 +1,92 @@
+#!/usr/bin/perl
+# parse-mozilla-encoding-table.pl, version 0.6
+#
+# Script to deassemble existing Mozilla *.uf or *.ut files
+# back to source conversion tables.
+# by Anthony Fok <anthony@thizlinux.com>, ThizLinux Laboratory Ltd., 2002/11/27
+# License: GNU General Public License, version 2 or newer
+#
+# Used for verifying HKSCS-1999 hkscs.uf and hkscs.ut so that I can make
+# new ones for HKSCS-2001. This script is quick-and-dirty and not very
+# robust, so if the debug output of fromu/tou ever changes, this script
+# will need to be modified too. :-)
+
+my %data = ();
+my $mappingPos = 0;
+my $filename = shift;
+my $mode;
+if ($filename =~ /\.(ut|uf)$/) {
+ print $filename, "\n";
+ $mode = $1;
+} else {
+ die;
+}
+
+open(INFILE, "<$filename") or die;
+
+# Quick-and-dirty routine to populate %data
+while (<INFILE>) {
+ if (/^Begin of Item ([[:xdigit:]]+)/) {
+ die if defined($itemId) and hex($itemId) + 1 != hex($1);
+ $itemId = $1;
+ <INFILE> =~ /Format ([012])/ or die;
+ $format = $1;
+ <INFILE> =~ /srcBegin = ([[:xdigit:]]+)/ or die;
+ $srcBegin = $1;
+
+ if ($format == 0) { # Range
+ <INFILE> =~ /srcEnd = ([[:xdigit:]]+)/ or die;
+ $srcEnd = $1;
+ <INFILE> =~ /destBegin = ([[:xdigit:]]+)/ or die;
+ $destBegin = $1;
+
+ for ($i = hex($srcBegin); $i <= hex($srcEnd); $i++) {
+ $data{sprintf("%04X",$i)} = sprintf("%04X",
+ hex($destBegin) + $i - hex($srcBegin));
+ }
+
+ <INFILE> =~ /^End of Item $itemId\s*$/ or die;
+ }
+ elsif ($format == 1) { # Mapping
+ <INFILE> =~ /srcEnd = ([[:xdigit:]]+)/ or die;
+ $srcEnd = $1;
+ <INFILE> =~ /mappingOffset = ([[:xdigit:]]+)/ or die;
+ $mappingOffset = hex($1);
+ die unless $mappingOffset == $mappingPos;
+ <INFILE> =~ /Mapping =\s*$/ or die;
+ until ($_ = <INFILE>, /^End of Item/) {
+ chop;
+ for $i (split ' ') {
+ $key = sprintf("%04X", hex($srcBegin) - $mappingOffset + $mappingPos++);
+ next if $i eq "FFFD";
+ if (defined($data{$key})) {
+ print "Error: doubly defined. $key was $data{$key}, and now $i.\n";
+ } else {
+ $data{$key} = $i;
+ }
+ }
+ }
+ die unless $mappingPos - $mappingOffset == hex($srcEnd) - hex($srcBegin) + 1;
+ /^End of Item $itemId\s*$/ or die;
+ }
+ else { # Single ($format == 2)
+ <INFILE> =~ /destBegin = ([[:xdigit:]]+)/ or die;
+ $destBegin = $1;
+ $data{$srcBegin} = $destBegin;
+ <INFILE> =~ /^End of Item $itemId\s*$/ or die;
+ }
+ }
+}
+
+# Generate conversion table
+for $key (sort keys %data) {
+ if ($mode eq "ut") {
+ print "0x$key\t0x$data{$key}\n";
+ } elsif ($mode eq "uf") {
+ print "0x$data{$key}\t0x$key\n";
+ } else {
+ die;
+ }
+}
+
+close INFILE;