SphinxBase 0.6

src/sphinx_jsgf2fsg/main.c

00001 /* -*- c-basic-offset: 4; indent-tabs-mode: nil -*- */
00002 /* ====================================================================
00003  * Copyright (c) 2007 Carnegie Mellon University.  All rights
00004  * reserved.
00005  *
00006  * Redistribution and use in source and binary forms, with or without
00007  * modification, are permitted provided that the following conditions
00008  * are met:
00009  *
00010  * 1. Redistributions of source code must retain the above copyright
00011  *    notice, this list of conditions and the following disclaimer. 
00012  *
00013  * 2. Redistributions in binary form must reproduce the above copyright
00014  *    notice, this list of conditions and the following disclaimer in
00015  *    the documentation and/or other materials provided with the
00016  *    distribution.
00017  *
00018  * This work was supported in part by funding from the Defense Advanced 
00019  * Research Projects Agency and the National Science Foundation of the 
00020  * United States of America, and the CMU Sphinx Speech Consortium.
00021  *
00022  * THIS SOFTWARE IS PROVIDED BY CARNEGIE MELLON UNIVERSITY ``AS IS'' AND 
00023  * ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, 
00024  * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
00025  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL CARNEGIE MELLON UNIVERSITY
00026  * NOR ITS EMPLOYEES BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
00027  * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT 
00028  * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 
00029  * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 
00030  * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 
00031  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE 
00032  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
00033  *
00034  * ====================================================================
00035  *
00036  */
00037 
00038 #include <string.h>
00039 
00040 #include <sphinxbase/hash_table.h>
00041 #include <sphinxbase/fsg_model.h>
00042 #include <sphinxbase/jsgf.h>
00043 #include <sphinxbase/err.h>
00044 #include <sphinxbase/strfuncs.h>
00045 
00046 static const arg_t defn[] = {
00047   { "-help",
00048     ARG_BOOLEAN,
00049     "no",
00050     "Shows the usage of the tool"},
00051 
00052   { "-jsgf",
00053     REQARG_STRING,
00054     NULL,
00055     "Input grammar in jsgf format (required)"},
00056 
00057   { "-rule",
00058     ARG_STRING,
00059     NULL,
00060     "Root rule name (optional)"},
00061 
00062   { "-fsg",
00063     ARG_STRING,
00064     NULL,
00065     "Output grammar in fsg format"},
00066 
00067   { "-fsm",
00068     ARG_STRING,
00069     NULL,
00070     "Output grammar in FSM format"},
00071 
00072   { "-symtab",
00073     ARG_STRING,
00074     NULL,
00075     "Output symtab for grammar in FSM format"},
00076 
00077   { "-compile",
00078     ARG_BOOLEAN,
00079     "no",
00080     "Compute grammar closure to speedup loading"},
00081 
00082   { NULL, 0, NULL, NULL }
00083 };
00084 
00085 
00086 static void
00087 usagemsg(char *pgm)
00088 {
00089     E_INFO("Usage: %s -jsgf <input.jsgf> -rule <rule name>\\\n", pgm);
00090     E_INFOCONT("\t[-fsm yes/no] [-compile yes/no]\n");
00091     E_INFOCONT("\t-fsg <output.fsg>\n");
00092 
00093     exit(0);
00094 }
00095 
00096 static fsg_model_t *
00097 get_fsg(jsgf_t *grammar, const char *name)
00098 {
00099     jsgf_rule_iter_t *itor;
00100     logmath_t *lmath = logmath_init(1.0001, 0, 0);
00101     fsg_model_t *fsg = NULL;
00102 
00103     for (itor = jsgf_rule_iter(grammar); itor;
00104          itor = jsgf_rule_iter_next(itor)) {
00105         jsgf_rule_t *rule = jsgf_rule_iter_rule(itor);
00106         char const *rule_name = jsgf_rule_name(rule);
00107 
00108         if ((name == NULL && jsgf_rule_public(rule))
00109             || (name && strlen(rule_name)-2 == strlen(name) &&
00110                 0 == strncmp(rule_name + 1, name, strlen(rule_name) - 2))) {
00111             fsg = jsgf_build_fsg_raw(grammar, rule, logmath_retain(lmath), 1.0);
00112             jsgf_rule_iter_free(itor);
00113             break;
00114         }
00115     }
00116 
00117     logmath_free(lmath);
00118     return fsg;
00119 }
00120 
00121 int
00122 main(int argc, char *argv[])
00123 {
00124     jsgf_t *jsgf;
00125     fsg_model_t *fsg;
00126     cmd_ln_t *config;
00127         
00128     if ((config = cmd_ln_parse_r(NULL, defn, argc, argv, TRUE)) == NULL)
00129         return 1;
00130                 
00131     if (cmd_ln_boolean_r(config, "-help")) {
00132         usagemsg(argv[0]);
00133     }
00134 
00135     jsgf = jsgf_parse_file(cmd_ln_str_r(config, "-jsgf"), NULL);
00136     if (jsgf == NULL) {
00137         return 1;
00138     }
00139 
00140     fsg = get_fsg(jsgf, cmd_ln_str_r(config, "-rule") ? cmd_ln_str_r(config, "-rule") : NULL);
00141     
00142     if (cmd_ln_boolean_r(config, "-compile")) {
00143         fsg_model_null_trans_closure(fsg, NULL);
00144     }
00145 
00146     
00147     if (cmd_ln_str_r(config, "-fsm")) {
00148         const char* outfile = cmd_ln_str_r(config, "-fsm");
00149         const char* symfile = cmd_ln_str_r(config, "-symtab");
00150         if (outfile)
00151             fsg_model_writefile_fsm(fsg, outfile);
00152         else
00153             fsg_model_write_fsm(fsg, stdout);
00154         if (symfile)
00155             fsg_model_writefile_symtab(fsg, symfile);
00156     }
00157     else {
00158         const char *outfile = cmd_ln_str_r(config, "-fsg");
00159         if (outfile)
00160             fsg_model_writefile(fsg, outfile);
00161         else
00162             fsg_model_write(fsg, stdout);
00163     }
00164     fsg_model_free(fsg);
00165     jsgf_grammar_free(jsgf);
00166 
00167     return 0;
00168 }
00169 
00170 
00171 #if defined(_WIN32_WCE)
00172 #pragma comment(linker,"/entry:mainWCRTStartup")
00173 #include <windows.h>
00174 
00175 //Windows Mobile has the Unicode main only
00176 int wmain(int32 argc, wchar_t *wargv[]) {
00177     char** argv;
00178     size_t wlen;
00179     size_t len;
00180     int i;
00181 
00182     argv = malloc(argc*sizeof(char*));
00183     for (i=0; i<argc; i++){
00184         wlen = lstrlenW(wargv[i]);
00185         len = wcstombs(NULL, wargv[i], wlen);
00186         argv[i] = malloc(len+1);
00187         wcstombs(argv[i], wargv[i], wlen);
00188     }
00189 
00190     //assuming ASCII parameters
00191     return main(argc, argv);
00192 }
00193 #endif