| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # gsprintf.pm --
|
|---|
| 4 | # A component of the Greenstone digital library software
|
|---|
| 5 | # from the New Zealand Digital Library Project at the
|
|---|
| 6 | # University of Waikato, New Zealand.
|
|---|
| 7 | #
|
|---|
| 8 | # Copyright (C) 1999 New Zealand Digital Library Project
|
|---|
| 9 | #
|
|---|
| 10 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 11 | # it under the terms of the GNU General Public License as published by
|
|---|
| 12 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 13 | # (at your option) any later version.
|
|---|
| 14 | #
|
|---|
| 15 | # This program is distributed in the hope that it will be useful,
|
|---|
| 16 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 17 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 18 | # GNU General Public License for more details.
|
|---|
| 19 | #
|
|---|
| 20 | # You should have received a copy of the GNU General Public License
|
|---|
| 21 | # along with this program; if not, write to the Free Software
|
|---|
| 22 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 23 | #
|
|---|
| 24 | ###########################################################################
|
|---|
| 25 | use strict;
|
|---|
| 26 | no strict 'refs';
|
|---|
| 27 |
|
|---|
| 28 | package gsprintf;
|
|---|
| 29 | require Exporter;
|
|---|
| 30 | @gsprintf::ISA=qw(Exporter);
|
|---|
| 31 |
|
|---|
| 32 | use unicode;
|
|---|
| 33 | use util;
|
|---|
| 34 |
|
|---|
| 35 | @gsprintf::EXPORT_OK = qw(gsprintf); # functions we can export into namespace
|
|---|
| 36 |
|
|---|
| 37 |
|
|---|
| 38 | # Language-specific resource bundle
|
|---|
| 39 | my %specialresourcebundle = ();
|
|---|
| 40 | our $specialoutputencoding; # our, so that it can be changed outside.
|
|---|
| 41 |
|
|---|
| 42 | # Default resource bundle
|
|---|
| 43 | my %defaultresourcebundle;
|
|---|
| 44 | my $defaultoutputencoding;
|
|---|
| 45 |
|
|---|
| 46 | # English resource bundle
|
|---|
| 47 | my %englishresourcebundle;
|
|---|
| 48 | my $englishoutputencoding;
|
|---|
| 49 |
|
|---|
| 50 | # Ignore the OutputEncoding strings in the resource bundles and output all text in UTF-8
|
|---|
| 51 | my $outputstringsinUTF8 = 0;
|
|---|
| 52 |
|
|---|
| 53 |
|
|---|
| 54 | sub gsprintf
|
|---|
| 55 | {
|
|---|
| 56 | my ($handle, $text_string, @text_arguments) = @_;
|
|---|
| 57 |
|
|---|
| 58 | # Return unless the required arguments were supplied
|
|---|
| 59 | return unless (defined($handle) && defined($text_string));
|
|---|
| 60 |
|
|---|
| 61 | # Look up all the strings in the dictionary
|
|---|
| 62 | $text_string =~ s/(\{[^\}]+\})/&lookup_string($1)/eg;
|
|---|
| 63 |
|
|---|
| 64 | # Resolve the string arguments using sprintf, then write out to the handle
|
|---|
| 65 | print $handle sprintf($text_string, @text_arguments);
|
|---|
| 66 | }
|
|---|
| 67 |
|
|---|
| 68 |
|
|---|
| 69 | sub lookup_string
|
|---|
| 70 | {
|
|---|
| 71 | my ($stringkey) = @_;
|
|---|
| 72 | return "" unless defined $stringkey;
|
|---|
| 73 | # Try the language-specific resource bundle first
|
|---|
| 74 | my $utf8string = $specialresourcebundle{$stringkey};
|
|---|
| 75 | my $outputencoding = $specialoutputencoding;
|
|---|
| 76 |
|
|---|
| 77 | # Try the default resource bundle next
|
|---|
| 78 | if (!defined($utf8string)) {
|
|---|
| 79 | # Load the default resource bundle if it is not already loaded
|
|---|
| 80 | &load_default_resource_bundle() if (!%defaultresourcebundle);
|
|---|
| 81 |
|
|---|
| 82 | $utf8string = $defaultresourcebundle{$stringkey};
|
|---|
| 83 | $outputencoding = $defaultoutputencoding;
|
|---|
| 84 | }
|
|---|
| 85 |
|
|---|
| 86 | # Try the English resource bundle last
|
|---|
| 87 | if (!defined($utf8string)) {
|
|---|
| 88 | # Load the English resource bundle if it is not already loaded
|
|---|
| 89 | &load_english_resource_bundle() if (!%englishresourcebundle);
|
|---|
| 90 |
|
|---|
| 91 | $utf8string = $englishresourcebundle{$stringkey};
|
|---|
| 92 | $outputencoding = $englishoutputencoding;
|
|---|
| 93 | }
|
|---|
| 94 |
|
|---|
| 95 | # No matching string was found, so just return the key
|
|---|
| 96 | if (!defined($utf8string)) {
|
|---|
| 97 | return $stringkey;
|
|---|
| 98 | }
|
|---|
| 99 |
|
|---|
| 100 | # Return the string matching the key
|
|---|
| 101 | return $utf8string if (!defined($outputencoding) || $outputstringsinUTF8
|
|---|
| 102 | || $outputencoding eq "utf8");
|
|---|
| 103 |
|
|---|
| 104 | # If an 8-bit output encoding has been defined, encode the string appropriately
|
|---|
| 105 | my $encoded=unicode::unicode2singlebyte(&unicode::utf82unicode($utf8string), $outputencoding);
|
|---|
| 106 |
|
|---|
| 107 | # If we successfully encoded it, return it
|
|---|
| 108 | if ($encoded) { return $encoded }
|
|---|
| 109 |
|
|---|
| 110 | # Otherwise, we can't convert to the requested encoding. return the utf8?
|
|---|
| 111 | $specialoutputencoding='utf8';
|
|---|
| 112 | return $utf8string;
|
|---|
| 113 | }
|
|---|
| 114 |
|
|---|
| 115 |
|
|---|
| 116 | sub load_language_specific_resource_bundle
|
|---|
| 117 | {
|
|---|
| 118 | my $language = shift(@_);
|
|---|
| 119 |
|
|---|
| 120 | # Read the specified resource bundle
|
|---|
| 121 | my $resourcebundlename = "strings_" . $language . ".properties";
|
|---|
| 122 |
|
|---|
| 123 | %specialresourcebundle
|
|---|
| 124 | = &read_resource_bundle_and_extensions($ENV{'GSDLHOME'},"perllib",$resourcebundlename);
|
|---|
| 125 | return if (!%specialresourcebundle);
|
|---|
| 126 |
|
|---|
| 127 | # Read the output encoding to use from the resource bundle
|
|---|
| 128 | if ($ENV{'GSDLOS'} =~ /windows/) {
|
|---|
| 129 | $specialoutputencoding = $specialresourcebundle{"{OutputEncoding.windows}"};
|
|---|
| 130 | }
|
|---|
| 131 | else {
|
|---|
| 132 | # see if there is an encoding set in the appropriate locale env var
|
|---|
| 133 |
|
|---|
| 134 | foreach my $envvar ('LC_ALL', 'LANG') {
|
|---|
| 135 | if (!exists $ENV{$envvar}) { next }
|
|---|
| 136 | my $locale=$ENV{$envvar};
|
|---|
| 137 | if ($locale !~ /^\w+\.(.+)$/) { next }
|
|---|
| 138 | my $enc=lc($1);
|
|---|
| 139 | $enc =~ s/-/_/g;
|
|---|
| 140 | if ($enc eq 'utf_8') { $enc='utf8' } # normalise to this name
|
|---|
| 141 | $specialoutputencoding = $enc;
|
|---|
| 142 | return;
|
|---|
| 143 | }
|
|---|
| 144 | $specialoutputencoding = $specialresourcebundle{"{OutputEncoding.unix}"};
|
|---|
| 145 | }
|
|---|
| 146 | }
|
|---|
| 147 |
|
|---|
| 148 |
|
|---|
| 149 | sub load_default_resource_bundle
|
|---|
| 150 | {
|
|---|
| 151 | # Read the default resource bundle
|
|---|
| 152 | my $resourcebundlename = "strings.properties";
|
|---|
| 153 |
|
|---|
| 154 | %defaultresourcebundle
|
|---|
| 155 | = &read_resource_bundle_and_extensions($ENV{'GSDLHOME'},"perllib",$resourcebundlename);
|
|---|
| 156 | if (!%defaultresourcebundle) {
|
|---|
| 157 | # $! will still have the error value for the last failed syscall
|
|---|
| 158 | print STDERR "$! $resourcebundlename\n";
|
|---|
| 159 | # set something so we don't bother trying to load it again
|
|---|
| 160 | $defaultresourcebundle{0}=undef;
|
|---|
| 161 | return;
|
|---|
| 162 | }
|
|---|
| 163 |
|
|---|
| 164 | # Read the output encoding to use from the resource bundle
|
|---|
| 165 | if ($ENV{'GSDLOS'} =~ /windows/) {
|
|---|
| 166 | $defaultoutputencoding = $defaultresourcebundle{"{OutputEncoding.windows}"};
|
|---|
| 167 | }
|
|---|
| 168 | else {
|
|---|
| 169 | $defaultoutputencoding = $defaultresourcebundle{"{OutputEncoding.unix}"};
|
|---|
| 170 | }
|
|---|
| 171 | }
|
|---|
| 172 |
|
|---|
| 173 |
|
|---|
| 174 | sub load_english_resource_bundle
|
|---|
| 175 | {
|
|---|
| 176 | # Ensure the English resource bundle hasn't already been loaded
|
|---|
| 177 | if (%specialresourcebundle && $specialresourcebundle{"{Language.code}"} eq "en") {
|
|---|
| 178 | %englishresourcebundle = %specialresourcebundle;
|
|---|
| 179 | $englishoutputencoding = $specialoutputencoding;
|
|---|
| 180 | }
|
|---|
| 181 |
|
|---|
| 182 | if ($defaultresourcebundle{"{Language.code}"} &&
|
|---|
| 183 | $defaultresourcebundle{"{Language.code}"} eq "en") {
|
|---|
| 184 | %englishresourcebundle = %defaultresourcebundle;
|
|---|
| 185 | $englishoutputencoding = $defaultoutputencoding;
|
|---|
| 186 | }
|
|---|
| 187 |
|
|---|
| 188 | # Read the English resource bundle
|
|---|
| 189 | my $resourcebundlename = "strings_en.properties";
|
|---|
| 190 |
|
|---|
| 191 | %englishresourcebundle
|
|---|
| 192 | = &read_resource_bundle_and_extensions($ENV{'GSDLHOME'},"perllib",$resourcebundlename);
|
|---|
| 193 | return if (!%englishresourcebundle);
|
|---|
| 194 |
|
|---|
| 195 | # Read the output encoding to use from the resource bundle
|
|---|
| 196 | if ($ENV{'GSDLOS'} =~ /windows/) {
|
|---|
| 197 | $englishoutputencoding = $englishresourcebundle{"{OutputEncoding.windows}"};
|
|---|
| 198 | }
|
|---|
| 199 | else {
|
|---|
| 200 | $englishoutputencoding = $englishresourcebundle{"{OutputEncoding.unix}"};
|
|---|
| 201 | }
|
|---|
| 202 | }
|
|---|
| 203 |
|
|---|
| 204 |
|
|---|
| 205 | sub read_resource_bundle_and_extensions
|
|---|
| 206 | {
|
|---|
| 207 | my ($bundle_base,$primary_dir,$resourcename) = @_;
|
|---|
| 208 |
|
|---|
| 209 | my $primary_resourcebundlefile
|
|---|
| 210 | = &util::filename_cat($bundle_base,$primary_dir,$resourcename);
|
|---|
| 211 |
|
|---|
| 212 | my $resourcebundle = read_resource_bundle($primary_resourcebundlefile);
|
|---|
| 213 | return if (!defined $resourcebundle);
|
|---|
| 214 |
|
|---|
| 215 | if (defined $ENV{'GSDLEXTS'}) {
|
|---|
| 216 | my @extensions = split(/:/,$ENV{'GSDLEXTS'});
|
|---|
| 217 | foreach my $e (@extensions) {
|
|---|
| 218 | my $ext_base
|
|---|
| 219 | = &util::filename_cat($bundle_base,"ext",$e);
|
|---|
| 220 |
|
|---|
| 221 | my $ext_resourcebundlefile
|
|---|
| 222 | = &util::filename_cat($ext_base,$primary_dir,$resourcename);
|
|---|
| 223 |
|
|---|
| 224 | # can ignore return value (will be same reference to $resourcebundle)
|
|---|
| 225 | read_resource_bundle($ext_resourcebundlefile,$resourcebundle);
|
|---|
| 226 | }
|
|---|
| 227 | }
|
|---|
| 228 | if (defined $ENV{'GSDL3EXTS'}) {
|
|---|
| 229 | my @extensions = split(/:/,$ENV{'GSDL3EXTS'});
|
|---|
| 230 | foreach my $e (@extensions) {
|
|---|
| 231 | my $ext_base
|
|---|
| 232 | = &util::filename_cat($ENV{'GSDL3SRCHOME'},"ext",$e);
|
|---|
| 233 |
|
|---|
| 234 | my $ext_resourcebundlefile
|
|---|
| 235 | = &util::filename_cat($ext_base,$primary_dir,$resourcename);
|
|---|
| 236 |
|
|---|
| 237 | # can ignore return value (will be same reference to $resourcebundle)
|
|---|
| 238 | read_resource_bundle($ext_resourcebundlefile,$resourcebundle);
|
|---|
| 239 | }
|
|---|
| 240 | }
|
|---|
| 241 |
|
|---|
| 242 | return %$resourcebundle;
|
|---|
| 243 | }
|
|---|
| 244 |
|
|---|
| 245 |
|
|---|
| 246 | sub read_resource_bundle
|
|---|
| 247 | {
|
|---|
| 248 | my ($resourcebundlefilepath,$resourcebundle) = @_;
|
|---|
| 249 |
|
|---|
| 250 | if (!open(RESOURCE_BUNDLE, "<$resourcebundlefilepath")) {
|
|---|
| 251 | # When called for the first time (primary resource), $resourcebundle
|
|---|
| 252 | # is not defined (=undef). If the file does not exist, then we return
|
|---|
| 253 | # this 'undef' to signal it was not found
|
|---|
| 254 | # For an extension resource bundle, if it does not exist this
|
|---|
| 255 | # is not so serious (in fact quite likely) => return what we
|
|---|
| 256 | # have built up so far
|
|---|
| 257 |
|
|---|
| 258 | return $resourcebundle;
|
|---|
| 259 | }
|
|---|
| 260 |
|
|---|
| 261 | if (!defined $resourcebundle) {
|
|---|
| 262 | # resource files exists, so exect some content to be stored
|
|---|
| 263 | $resourcebundle = {};
|
|---|
| 264 | }
|
|---|
| 265 |
|
|---|
| 266 | # Load this resource bundle
|
|---|
| 267 | my @resourcebundlelines = <RESOURCE_BUNDLE>;
|
|---|
| 268 | close(RESOURCE_BUNDLE);
|
|---|
| 269 |
|
|---|
| 270 | # Parse the resource bundle
|
|---|
| 271 |
|
|---|
| 272 | foreach my $line (@resourcebundlelines) {
|
|---|
| 273 | # Remove any trailing whitespace
|
|---|
| 274 | $line =~ s/(\s*)$//;
|
|---|
| 275 |
|
|---|
| 276 | # Ignore comments and empty lines
|
|---|
| 277 | if ($line !~ /^\#/ && $line ne "") {
|
|---|
| 278 | # Parse key (everything up to the first colon)
|
|---|
| 279 | if ($line =~ m/^([^:]+):(.+)$/) {
|
|---|
| 280 | my $linekey = "{" . $1 . "}";
|
|---|
| 281 | my $linetext = $2;
|
|---|
| 282 | $linetext =~ s/(\s*)\#\s+Updated\s+(\d?\d-\D\D\D-\d\d\d\d).*$//i;
|
|---|
| 283 |
|
|---|
| 284 | # Map key to text
|
|---|
| 285 | $resourcebundle->{$linekey} = $linetext;
|
|---|
| 286 | }
|
|---|
| 287 | }
|
|---|
| 288 | }
|
|---|
| 289 |
|
|---|
| 290 | return $resourcebundle;
|
|---|
| 291 | }
|
|---|
| 292 |
|
|---|
| 293 |
|
|---|
| 294 | sub output_strings_in_UTF8
|
|---|
| 295 | {
|
|---|
| 296 | $outputstringsinUTF8 = 1;
|
|---|
| 297 | }
|
|---|
| 298 |
|
|---|
| 299 |
|
|---|
| 300 | sub debug_unicode_string
|
|---|
| 301 | {
|
|---|
| 302 | join("",
|
|---|
| 303 | map { $_ > 255 ? # if wide character...
|
|---|
| 304 | sprintf("\\x{%04X}", $_) : # \x{...}
|
|---|
| 305 | chr($_)
|
|---|
| 306 | } unpack("U*", $_[0])); # unpack Unicode characters
|
|---|
| 307 | }
|
|---|
| 308 |
|
|---|
| 309 |
|
|---|
| 310 | 1;
|
|---|
