--- loncom/localize/lonlocal.pm	2008/09/13 02:37:29	1.48
+++ loncom/localize/lonlocal.pm	2020/10/29 23:04:39	1.68
@@ -1,7 +1,7 @@
 # The LearningOnline Network with CAPA
 # Localization routines
 #
-# $Id: lonlocal.pm,v 1.48 2008/09/13 02:37:29 raeburn Exp $
+# $Id: lonlocal.pm,v 1.68 2020/10/29 23:04:39 raeburn Exp $
 #
 # Copyright Michigan State University Board of Trustees
 #
@@ -38,7 +38,7 @@ Apache::lonlocal - provides localization
 
 lonlocal provides localization services for LON-CAPA programmers based
 on Locale::Maketext. See
-C<http://search.cpan.org/dist/Locale-Maketext/lib/Locale/Maketext.pod>
+C<https://metacpan.org/pod/Locale::Maketext>
 for more information on Maketext.
 
 =head1 OVERVIEWX<internationalization>
@@ -81,7 +81,7 @@ Inside might be something like this
      if ($status eq 'WON') {
         $message='You have won.';
      } elsif ($status eq 'LOST') {
-        $message='You are a total looser.';
+        $message='You are a total loser.';
      }
      return $message;
  }
@@ -103,7 +103,7 @@ The first two examples are easy:
      if ($status eq 'WON') {
         $message='You have won.';
      } elsif ($status eq 'LOST') {
-        $message='You are a total looser.';
+        $message='You are a total loser.';
      }
      return &mt($message);
  }
@@ -135,7 +135,7 @@ like this:
     'You have won.'
  => 'Sie haben gewonnen.',
 
-    'You are a total looser.'
+    'You are a total loser.'
  => 'Sie sind der totale Verlierer.',
 
     'Rules'
@@ -153,7 +153,7 @@ ignore the rest of the line.
 This is a relatively easy task, and any help is appreciated.
 
 Maketext can do a whole lot more, see
-C<http://search.cpan.org/dist/Locale-Maketext/lib/Locale/Maketext.pod>
+C<https://metacpan.org/pod/Locale::Maketext>
 but for most purposes, we do not have to mess with that.
 
 =cut
@@ -166,15 +166,18 @@ use locale;
 use POSIX qw(locale_h strftime);
 use DateTime();
 use DateTime::TimeZone;
+use DateTime::Locale;
 
 require Exporter;
 
 our @ISA = qw (Exporter);
-our @EXPORT = qw(mt mtn ns mt_user);
+our @EXPORT = qw(mt mtn ns mt_user js_escape html_escape);
+
+my %mtcache=();
 
 # ========================================================= The language handle
 
-use vars qw($lh);
+use vars qw($lh $current_language);
 
 # ===================================================== The "MakeText" function
 
@@ -190,7 +193,13 @@ sub mt (@) {
                 return $_[0];
             }
         } else {
-            return $lh->maketext(@_);
+            if ($#_>0) { return $lh->maketext(@_); }
+            if ($mtcache{$current_language.':'.$_[0]}) {
+               return $mtcache{$current_language.':'.$_[0]};
+            }
+            my $translation=$lh->maketext(@_);
+            $mtcache{$current_language.':'.$_[0]}=$translation;
+            return $translation; 
         }
     } else {
 	if (wantarray) {
@@ -232,13 +241,65 @@ sub current_language {
     return 'en';
 }
 
+sub preferred_languages {
+    my @languages=();
+    if (($Apache::lonnet::env{'request.role.adv'}) && ($Apache::lonnet::env{'form.languages'})) {
+        @languages=(@languages,split(/\s*(\,|\;|\:)\s*/,$Apache::lonnet::env{'form.languages'}));
+    }
+    if ($Apache::lonnet::env{'course.'.$Apache::lonnet::env{'request.course.id'}.'.languages'}) {
+        @languages=(@languages,split(/\s*(\,|\;|\:)\s*/,
+                 $Apache::lonnet::env{'course.'.$Apache::lonnet::env{'request.course.id'}.'.languages'}));
+    }
+
+    if ($Apache::lonnet::env{'environment.languages'}) {
+        @languages=(@languages,
+                    split(/\s*(\,|\;|\:)\s*/,$Apache::lonnet::env{'environment.languages'}));
+    }
+    my $browser=$ENV{'HTTP_ACCEPT_LANGUAGE'};
+    if ($browser) {
+        my @browser =
+            map { (split(/\s*;\s*/,$_))[0] } (split(/\s*,\s*/,$browser));
+        push(@languages,@browser);
+    }
+
+    my $defdom = &Apache::lonnet::default_login_domain();
+    foreach my $domtype ($Apache::lonnet::env{'user.domain'},$Apache::lonnet::env{'request.role.domain'},$defdom) {
+        if (($domtype ne '') && ($domtype ne 'public')) {
+            my %domdefs = &Apache::lonnet::get_domain_defaults($domtype);
+            if ($domdefs{'lang_def'} ne '') {
+                push(@languages,$domdefs{'lang_def'});
+            }
+        }
+    }
+    return &get_genlanguages(@languages);
+}
+
+sub get_genlanguages {
+    my (@languages) = @_;
+# turn "en-ca" into "en-ca,en"
+    my @genlanguages;
+    foreach my $lang (@languages) {
+        unless ($lang=~/\w/) { next; }
+        push(@genlanguages,$lang);
+        if ($lang=~/(\-|\_)/) {
+            push(@genlanguages,(split(/(\-|\_)/,$lang))[0]);
+        }
+    }
+    #uniqueify the languages list
+    my %count;
+    @genlanguages = map { $count{$_}++ == 0 ? $_ : () } @genlanguages;
+    return @genlanguages;
+}
+
 # ============================================================== What encoding?
 
 sub current_encoding {
     my $default='UTF-8';
-    if ($Apache::lonnet::env{'browser.os'} eq 'win' && 
-	$Apache::lonnet::env{'browser.type'} eq 'explorer') {
-        $default='ISO-8859-1';
+    unless ($Apache::lonnet::env{'browser.unicode'}) {
+        if ($Apache::lonnet::env{'browser.os'} eq 'win' && 
+	    $Apache::lonnet::env{'browser.type'} eq 'explorer') {
+            $default='ISO-8859-1';
+        }
     }
     if ($lh) {
 	my $enc=$lh->maketext('char_encoding');
@@ -264,7 +325,7 @@ sub current_locale {
 
 sub texthash {
     my %hash=@_;
-    foreach (keys %hash) {
+    foreach (keys(%hash)) {
 	$hash{$_}=&mt($hash{$_});
     }
     return %hash;
@@ -273,14 +334,20 @@ sub texthash {
 # ========= Get a handle (do not invoke in vain, leave this to access handlers)
 
 sub get_language_handle {
-    my $r=shift;
+    my ($r,$chosen) = @_;
     if ($r) {
 	my $headers=$r->headers_in;
 	$ENV{'HTTP_ACCEPT_LANGUAGE'}=$headers->{'Accept-language'};
     }
-    my @languages=&Apache::loncommon::preferred_languages;
+    my @languages;
+    if ($chosen ne '') {
+        @languages=($chosen);
+    } else {
+        @languages=&preferred_languages();
+    }
     $ENV{'HTTP_ACCEPT_LANGUAGE'}='';
     $lh=Apache::localize->get_handle(@languages);
+    $current_language=&current_language();
     if ($r) {
 	$r->content_languages([&current_language()]);
     }
@@ -289,17 +356,39 @@ sub get_language_handle {
 
 # ========================================================== Localize localtime
 sub gettimezone {
-    my $timezone; 
-    if ($Apache::lonnet::env{'course.'.$Apache::lonnet::env{'request.course.id'}.'.timezone'}) {
-        $timezone = $Apache::lonnet::env{'course.'.$Apache::lonnet::env{'request.course.id'}.'.timezone'};    
-    } elsif ($Apache::lonnet::env{'request.course.id'} ne '') {
-        my $cdom = $Apache::lonnet::env{'course.'.$Apache::lonnet::env{'request.course.id'}.'.domain'};
-        if ($cdom ne '') {
-            my %domdefaults = &Apache::lonnet::get_domain_defaults($cdom);
-            if ($domdefaults{'timezone_def'} ne '') {
-                $timezone = $domdefaults{'timezone_def'};
+    my ($timezone) = @_;
+    if ($timezone ne '') {
+        if (!DateTime::TimeZone->is_valid_name($timezone)) {
+            $timezone = 'local';
+        }
+        return $timezone;
+    }
+    my $cid = $Apache::lonnet::env{'request.course.id'};  
+    if ($cid ne '') {
+        if ($Apache::lonnet::env{'course.'.$cid.'.timezone'}) {
+            $timezone = $Apache::lonnet::env{'course.'.$cid.'.timezone'};    
+        } else {
+            my $cdom = $Apache::lonnet::env{'course.'.$cid.'.domain'};
+            if ($cdom ne '') {
+                my %domdefaults = &Apache::lonnet::get_domain_defaults($cdom);
+                if ($domdefaults{'timezone_def'} ne '') {
+                    $timezone = $domdefaults{'timezone_def'};
+                }
             }
         }
+    } elsif ($Apache::lonnet::env{'request.role.domain'} ne '') {
+        my %uroledomdefs = 
+            &Apache::lonnet::get_domain_defaults($Apache::lonnet::env{'request.role.domain'});
+        if ($uroledomdefs{'timezone_def'} ne '') {
+            $timezone = $uroledomdefs{'timezone_def'};
+        }
+    } elsif (($Apache::lonnet::env{'user.domain'} ne '') && 
+             ($Apache::lonnet::env{'user.domain'} ne 'public')) {
+        my %udomdefaults = 
+            &Apache::lonnet::get_domain_defaults($Apache::lonnet::env{'user.domain'});
+        if ($udomdefaults{'timezone_def'} ne '') {
+            $timezone = $udomdefaults{'timezone_def'};
+        }
     }
     if ($timezone ne '') {
         if (DateTime::TimeZone->is_valid_name($timezone)) {
@@ -310,7 +399,8 @@ sub gettimezone {
 }
 
 sub locallocaltime {
-    my $thistime=shift;
+    my ($thistime,$timezone) = @_;
+
     if (!defined($thistime) || $thistime eq '') {
 	return &mt('Never');
     }
@@ -324,7 +414,16 @@ sub locallocaltime {
     }
 
     my $dt = DateTime->from_epoch(epoch => $thistime)
-                     ->set_time_zone(&gettimezone());
+                     ->set_time_zone(gettimezone($timezone));
+
+    # TimeZone tries to determine the 'local' timezone from $ENV{TZ} if this
+    # fails it searches through various system files. Under certain
+    # circumstances this is an extremly expensive operation.
+    # So after the first run we store the timezone in $ENV{TZ} to significantly
+    # speed up future lookups. 
+    $ENV{TZ} = $dt->time_zone()->name() 
+        if (! $ENV{TZ} && gettimezone($timezone) eq 'local');
+
     if ((&current_language=~/^en/) || (!$lh)) {
 
 	return $dt->strftime("%a %b %e %I:%M:%S %P %Y (%Z)");
@@ -366,12 +465,52 @@ sub locallocaltime {
     }
 }
 
-# ==================== Normalize string (reduce fragility in the lexicon files)
+sub getdatelocale {
+    my ($datelocale,$locale_obj);
+    if ($Apache::lonnet::env{'course.'.$Apache::lonnet::env{'request.course.id'}.'.datelocale'}) {
+        $datelocale = $Apache::lonnet::env{'course.'.$Apache::lonnet::env{'request.course.id'}.'.datelocale'};
+    } elsif ($Apache::lonnet::env{'request.course.id'} ne '') {
+        my $cdom = $Apache::lonnet::env{'course.'.$Apache::lonnet::env{'request.course.id'}.'.domain'};
+        if ($cdom ne '') {
+            my %domdefaults = &Apache::lonnet::get_domain_defaults($cdom);
+            if ($domdefaults{'datelocale_def'} ne '') {
+                $datelocale = $domdefaults{'datelocale_def'};
+            }
+        }
+    } elsif ($Apache::lonnet::env{'user.domain'} ne '') {
+        my %udomdefaults = &Apache::lonnet::get_domain_defaults($Apache::lonnet::env{'user.domain'});
+        if ($udomdefaults{'datelocale_def'} ne '') {
+            $datelocale = $udomdefaults{'datelocale_def'};
+        }
+    }
+    if ($datelocale ne '') {
+        eval {
+            $locale_obj = DateTime::Locale->load($datelocale);
+        };
+        if (!$@) {
+            if ($locale_obj->id() eq $datelocale) {
+                return $locale_obj;
+            }
+        }
+    }
+    return $locale_obj;
+}
+
+=pod
+
+=over 
+
+=item * normalize_string()
+
+Normalize string (reduce fragility in the lexicon files)
+
+This normalizes a string to reduce fragility in the lexicon files of
+huge messages (such as are used by the helper), and allow useful
+formatting: reduce all consecutive whitespace to a single space,
+and remove all HTML
+
+=cut
 
-# This normalizes a string to reduce fragility in the lexicon files of
-# huge messages (such as are used by the helper), and allow useful
-# formatting: reduce all consecutive whitespace to a single space,
-# and remove all HTML
 sub normalize_string {
     my $s = shift;
     $s =~ s/\s+/ /g;
@@ -382,13 +521,27 @@ sub normalize_string {
     return $s;
 }
 
-# alias for normalize_string; recommend using it only in the lexicon
+=pod 
+
+=item * ns()
+
+alias for normalize_string; recommend using it only in the lexicon
+
+=cut
+
 sub ns {
     return normalize_string(@_);
 }
 
-# mtn: call the mt function and the normalization function easily.
-# Returns original non-normalized string if there was no translation
+=pod
+
+=item * mtn()
+
+mtn: call the mt function and the normalization function easily.
+Returns original non-normalized string if there was no translation
+
+=cut
+
 sub mtn (@) {
     my @args = @_; # don't want to modify caller's string; if we
 		   # didn't care about that we could set $_[0]
@@ -411,7 +564,7 @@ sub transstatic {
 
 =pod 
 
-=item * mt_escape
+=item * mt_escape()
 
 mt_escape takes a string reference and escape the [] in there so mt
 will leave them as is and not try to expand them
@@ -424,6 +577,119 @@ sub mt_escape {
     $$str_ref =~s/([\[\]])/~$1/g;
 }
 
+=pod 
+
+=item * js_escape()
+
+js_escape takes a string, string reference or hash reference,
+and escapes the values so that they can be used within a <script> element.
+It replaces all instances of \ by \\, ' by \', " by \" and \n by \\n.
+It is typically used with localized strings, which might contain quotes.
+
+=cut
+
+sub js_escape {
+    my ($v) = @_;
+    my $ref = ref($v);
+    if ($ref eq 'SCALAR') {
+        $$v =~ s/\\/\\\\/g;
+        $$v =~ s/'/\\'/g;
+        $$v =~ s/"/\\"/g;
+        $$v =~ s/\n/\\n/g;
+    } elsif ($ref eq 'HASH') {
+        foreach my $key (keys %$v) {
+            $v->{$key} =~ s/\\/\\\\/g;
+            $v->{$key} =~ s/'/\\'/g;
+            $v->{$key} =~ s/"/\\"/g;
+            $v->{$key} =~ s/\n/\\n/g;
+        }
+    } else {
+        $v =~ s/\\/\\\\/g;
+        $v =~ s/'/\\'/g;
+        $v =~ s/"/\\"/g;
+        $v =~ s/\n/\\n/g;
+        return $v;
+    }
+}
+
+=pod 
+
+=item * html_escape()
+
+html_escape takes a string, string reference or hash reference,
+and escapes the values so that they can be used as HTML.
+It encodes <, >, &, ' and ".
+
+=cut
+
+sub html_escape {
+    my ($v) = @_;
+    my $ref = ref($v);
+    if ($ref eq 'SCALAR') {
+        $$v =~ s/&/&amp;/g;
+        $$v =~ s/</&lt;/g;
+        $$v =~ s/>/&gt;/g;
+        $$v =~ s/'/&apos;/g;
+        $$v =~ s/"/&quot;/g;
+    } elsif ($ref eq 'HASH') {
+        foreach my $key (keys %$v) {
+            $v->{$key} =~ s/&/&amp;/g;
+            $v->{$key} =~ s/</&lt;/g;
+            $v->{$key} =~ s/>/&gt;/g;
+            $v->{$key} =~ s/'/&apos;/g;
+            $v->{$key} =~ s/"/&quot;/g;
+        }
+    } else {
+        $v =~ s/&/&amp;/g;
+        $v =~ s/</&lt;/g;
+        $v =~ s/>/&gt;/g;
+        $v =~ s/'/&apos;/g;
+        $v =~ s/"/&quot;/g;
+        return $v;
+    }
+    # NOTE: we could also turn \n into <br> if needed
+}
+
+=pod
+
+=item * choose_language()
+
+choose_language prompts a user to enter a two letter language code via
+keyboard when running a script from the command line. Default is en.
+
+=back
+
+=cut
+
+sub choose_language {
+    my %languages = (
+                      ar => 'Arabic',
+                      de => 'German',
+                      en => 'English',
+                      es => 'Spanish',
+                      fa => 'Persian',
+                      fr => 'French',
+                      he => 'Hebrew',
+                      ja => 'Japanese',
+                      pt => 'Portuguese',
+                      ru => 'Russian',
+                      tr => 'Turkish',
+                      zh => 'Chinese (Simplified)'
+                   );
+    my @posslangs = sort(keys(%languages));
+    my $langlist = join('|',@posslangs);
+    my $lang = 'en';
+    print 'Language: English (en). Change? ['.$langlist.']? ';
+    my $langchoice = <STDIN>;
+    chomp($langchoice);
+    $langchoice =~ s/(^\s+|\s+$)//g;
+    $langchoice = lc($langchoice);
+    if (defined($languages{$langchoice})) {
+        $lang = $langchoice;
+    }
+    return $lang;
+}
+
 1;
 
 __END__