universe@20: /*
universe@34:  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS HEADER. 
universe@57:  * Copyright 2018 Mike Becker. All rights reserved.
universe@34:  * 
universe@34:  * Redistribution and use in source and binary forms, with or without
universe@34:  * modification, are permitted provided that the following conditions are met:
universe@34:  * 
universe@34:  * 1. Redistributions of source code must retain the above copyright
universe@34:  * notice, this list of conditions and the following disclaimer.
universe@34:  * 
universe@34:  * 2. Redistributions in binary form must reproduce the above copyright
universe@34:  * notice, this list of conditions and the following disclaimer in the
universe@34:  * documentation and/or other materials provided with the distribution.
universe@34:  * 
universe@34:  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
universe@34:  * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
universe@34:  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
universe@34:  * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
universe@34:  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
universe@34:  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
universe@34:  * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
universe@34:  * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
universe@34:  * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
universe@57:  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
universe@20:  */
universe@20: 
universe@20: #include "bfile_heuristics.h"
universe@22: #include <ctype.h>
universe@20: 
universe@21: bfile_heuristics_t *new_bfile_heuristics_t() {
universe@21:   bfile_heuristics_t *ret = malloc(sizeof(bfile_heuristics_t));
universe@21:   ret->level = BFILE_MEDIUM_ACCURACY;
universe@22:   bfile_reset(ret);
universe@21:   return ret;
universe@20: }
universe@20: 
universe@21: void destroy_bfile_heuristics_t(bfile_heuristics_t *def) {
universe@21:   free(def);
universe@20: }
universe@21: 
universe@22: void bfile_reset(bfile_heuristics_t *def) {
universe@22:   def->bcount = 0;
universe@22:   def->tcount = 0;
universe@22: }
universe@22: 
universe@21: bool bfile_check(bfile_heuristics_t *def, int next_char) {
universe@21:   bool ret = false;
universe@22:   if (def->level != BFILE_IGNORE) {
universe@22:     def->tcount++;
universe@22:     if (!isprint(next_char) && !isspace(next_char)) {
universe@22:       def->bcount++;
universe@22:     }
universe@22: 
universe@23:     if (def->tcount > 1) { /* empty files are text files */
universe@23:       switch (def->level) {
universe@23:       case BFILE_LOW_ACCURACY:
universe@23:         if (def->tcount > 15 || next_char == EOF) {
universe@23:           ret = (1.0*def->bcount)/def->tcount > 0.32;
universe@23:         }
universe@23:         break;
universe@23:       case BFILE_HIGH_ACCURACY:
universe@23:         if (def->tcount > 500 || next_char == EOF) {
universe@23:           ret = (1.0*def->bcount)/def->tcount > 0.1;
universe@23:         }
universe@23:         break;
universe@23:       default: /* BFILE_MEDIUM_ACCURACY */
universe@23:         if (def->tcount > 100 || next_char == EOF) {
universe@23:           ret = (1.0*def->bcount)/def->tcount > 0.1;
universe@23:         }
universe@23:         break;
universe@22:       }
universe@22:     }
universe@22:   }
universe@21: 
universe@21:   return ret;
universe@21: }