Initial parser for the langv2 language file format

git-svn-id: svn://svn.rockbox.org/rockbox/trunk@7537 a1c6a512-1295-4272-9138-f99709370657

tools/genlang2

@@ -0,0 +1,252 @@
+#!/usr/bin/perl -s
+
+if(!$ARGV[0]) {
+    print <<MOO
+Usage: genlang2 [-p=<prefix>][-t=<target>][-v] <language file>
+
+<prefix>.h and <prefix>.c will be created in the current directory. <prefix>
+is "lang" by default.
+
+Use -v for verbose (debug) output.
+
+MOO
+;
+    exit;
+}
+
+my $prefix = $p;
+if(!$prefix) {
+    $prefix="lang";
+}
+my $target = $t;
+if(!$target) {
+    print "Please specify a target!\n";
+    exit;
+}
+my $verbose=$v;
+
+my %id; # string to num hash
+my @idnum; # num to string array
+
+my %source; # id string to source phrase hash
+my %dest; # id string to dest phrase hash
+my %voice; # id string to voice phrase hash
+
+
+my $input = $ARGV[0];
+
+open(HFILE, ">$prefix.h");
+open(CFILE, ">$prefix.c");
+
+print HFILE <<MOO
+/* This file was automatically generated using genlang2 */
+/*
+ * The str() macro/functions is how to access strings that might be
+ * translated. Use it like str(MACRO) and expect a string to be
+ * returned!
+ */
+#define str(x) language_strings[x]
+
+/* this is the array for holding the string pointers.
+   It will be initialized at runtime. */
+extern unsigned char *language_strings[];
+/* this contains the concatenation of all strings, separated by \\0 chars */
+extern const unsigned char language_builtin[];
+
+/* The enum below contains all available strings */
+enum {
+MOO
+    ;
+
+print CFILE <<MOO
+/* This file was automaticly generated using genlang2, the strings come
+   from "$input" */
+   
+#include "$prefix.h"
+
+unsigned char *language_strings[LANG_LAST_INDEX_IN_ARRAY];
+const unsigned char language_builtin[] =
+MOO
+    ;
+
+my @m;
+my $m="blank";
+
+sub match {
+    my ($string, $pattern)=@_;
+
+    $pattern =~ s/\*/.?*/g;
+    $pattern =~ s/\?/./g;
+
+    return ($string =~ $pattern);
+}
+
+sub blank {
+    # nothing to do
+}
+
+my %head;
+sub header {
+    my ($full, $n, $v)=@_;
+    $head{$n}=$v;
+}
+
+my %phrase;
+sub phrase {
+    my ($full, $n, $v)=@_;
+    $phrase{$n}=$v;
+}
+
+sub parsetarget {
+    my ($debug, $strref, $full, $n, $v)=@_;
+    my $string;
+    my @all= split(" *, *", $n);
+    my $test;
+    for $test (@all) {
+#        print "TEST ($debug) $target for $test\n";
+        if(match($target, $test)) {
+            $string = $v;
+#            print "MATCH: $test => $v\n";
+        }
+    }
+    if($string) {
+        $$strref = $string;
+    }
+    return $string;
+}
+
+my $src;
+sub source {
+    parsetarget("src", \$src, @_);
+}
+
+my $dest;
+sub dest {
+    parsetarget("dest", \$dest, @_);
+}
+
+my $voice;
+sub voice {
+    parsetarget("voice", \$voice, @_);
+}
+
+my $idcount; # counter for ID numbers
+
+open(LANG, "<$input");
+while(<LANG>) {
+    $line++;
+    if($_ =~ / *\#/) {
+        # comment
+        next;
+    }
+    # get rid of DOS newlines
+    $_ =~ s/\r//g;
+
+   # print "M: $m\n";
+
+    if(/ *<([^>]*)>/) {
+        my $part = $1;
+        #print "P: $part\n";
+        if($part =~ /^\//) {
+            if($part eq "/phrase") {
+                my $idstr = $phrase{'id'};
+
+                $id{$idstr} = $idcount;
+                $idnum[$idcount]=$idstr;
+
+                $source{$idstr}=$src;
+                $dest{$idstr}=$dest;
+                $voice{$idstr}=$voice;
+
+                if($verbose) {
+                    print "id: $phrase{id}\n";
+                    print "source: $src\n";
+                    print "dest: $dest\n";
+                    print "voice: $voice\n";
+                }
+
+                $idcount++;
+
+                undef $src;
+                undef $dest;
+                undef $voice;
+                undef %phrase;
+            }
+            # starts with a slash, this _ends_ this section
+            $m = pop @m; # get back old value
+            next;
+        }
+        push @m, $m; # store old value
+        $m = $1;
+        next;
+    }
+
+    if(/^ *([^:]+): *(.*)/) {
+        my ($name, $val)=($1, $2);
+        &$m($_, $name, $val);
+    }
+
+}
+close(LANG);
+
+# Output the ID names for the enum in the header file
+my $i;
+for $i (1 .. $idcount) {
+    my $name=$idnum[$i - 1]; # get the ID name
+
+    $name =~ s/\"//g; # cut off the quotes
+
+    printf HFILE ("    %s,\n", $name);
+}
+
+# Output separation marker for last string ID and the upcoming voice IDs
+
+print HFILE <<MOO
+    LANG_LAST_INDEX_IN_ARRAY, /* this is not a string, this is a marker */
+    /* --- below this follows voice-only strings --- */
+    VOICEONLY_DELIMITER = 0x8000,
+MOO
+    ;
+
+# TODO: add voice-only phrase IDs here
+
+# Output end of enum
+print HFILE <<MOO
+};
+/* end of generated enum list */
+MOO
+    ;
+
+# Output the target phrases for the source file
+for $i (1 .. $idcount) {
+    my $name=$idnum[$i - 1]; # get the ID
+    my $dest = $dest{$name}; # get the destination phrase
+
+    $dest =~ s:\"$:\\0\":; # insert a \0 before the second quote
+
+    printf CFILE ("    %s\n", $dest);
+}
+
+# Output end of string chunk
+print CFILE <<MOO
+;
+/* end of generated string list */
+MOO
+    ;
+
+close(HFILE);
+close(CFILE);
+
+if($verbose) {
+    printf("%d ID strings scanned\n", $idcount);
+
+    print "* head *\n";
+    for(keys %head) {
+        printf "$_: %s\n", $head{$_};
+    }
+}
+
+#print "* phrase *\n";
+#for(keys %phrase) {
+#    print "$_\n";
+#}