#include #include #include #include #include #include #include "mystdlib.h" static char xlat_table[] = { 0x00, 0x09, 0x02, 0x03, 0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27, 0x28, 0x29, 0x2a, 0x2b, 0x2c, 0x2d, 0x2e, 0x2f, 0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x3a, 0x3b, 0x3c, 0x3d, 0x3e, 0x3f, 0x40, 0x41, 0x42, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48, 0x49, 0x4a, 0x4b, 0x4c, 0x4d, 0x4e, 0x4f, 0x50, 0x51, 0x52, 0x53, 0x54, 0x55, 0x56, 0x57, 0x58, 0x59, 0x5a, 0x5f, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f, 0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0xc7, 0xfc, 0xe9, 0xe2, 0xe4, 0xe0, 0xe5, 0xe7, 0xea, 0xeb, 0xe8, 0xef, 0xee, 0xec, 0xc4, 0xc5, 0xc9, 0xe6, 0xc6, 0xf4, 0xf6, 0xf2, 0xfb, 0xf9, 0xff, 0xd6, 0xdc, 0xe1, 0xed, 0xf3, 0xfa, 0xf1, 0xd1, 0xdf, 0x5b, 0x5c, 0x5d, 0x60 }; static char cp437_to_iso8859_1_table[] = { 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f, 0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27, 0x28, 0x29, 0x2a, 0x2b, 0x2c, 0x2d, 0x2e, 0x2f, 0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x3a, 0x3b, 0x3c, 0x3d, 0x3e, 0x3f, 0x40, 0x41, 0x42, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48, 0x49, 0x4a, 0x4b, 0x4c, 0x4d, 0x4e, 0x4f, 0x50, 0x51, 0x52, 0x53, 0x54, 0x55, 0x56, 0x57, 0x58, 0x59, 0x5a, 0x5b, 0x5c, 0x5d, 0x5e, 0x5f, 0x60, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f, 0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0x7b, 0x7c, 0x7d, 0x7e, 0x7f, 0xc7, 0xfc, 0xe9, 0xe2, 0xe4, 0xe0, 0xe5, 0xe7, 0xea, 0xeb, 0xe8, 0xef, 0xee, 0xec, 0xc4, 0xc5, 0xc9, 0xe6, 0xc6, 0xf4, 0xf6, 0xf2, 0xfb, 0xf9, 0xff, 0xd6, 0xdc, 0xa2, 0xa3, 0xa5, 0x2e, 0x2e, 0xe1, 0xed, 0xf3, 0xfa, 0xf1, 0xd1, 0xaa, 0xba, 0xbf, 0x2e, 0xac, 0xbd, 0xbc, 0xa1, 0xab, 0xbb, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0xdf, 0x2e, 0x2e, 0x2e, 0x2e, 0xb5, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0x2e, 0xb1, 0x2e, 0x2e, 0x2e, 0x2e, 0xf7, 0x2e, 0xb0, 0x2e, 0xb7, 0x2e, 0x2e, 0xb2, 0x2e, 0xa0 }; static struct { int outfiles[15]; char * vorwahl; char * ort; char * zip; } g_state; void decode_7bit_string( uint8_t const *source, uint8_t *dest ) { uint16_t acc = 0, acc_bits = 0; while( 1 ) { acc = acc*256+*(source++); ++acc_bits; again: *(dest++) = xlat_table[ 0x7f & ( acc >> acc_bits ) ]; if( !dest[-1] ) return; if( acc_bits == 7 ) { acc_bits = 0; goto again; } } } void split_to_files( uint8_t *entries, int num_entries ) { int entry = 0, column; char *e = (char*)entries; for( entry = 0; entry < num_entries; ++entry ) { /* mimic flags from telefonbuch v3 */ write( g_state.outfiles[0], entry ? "02\n" : ( num_entries > 1 ? "01\n" : "00\n" ), 3 ); dprintf( g_state.outfiles[12], "%s\n", g_state.ort ); /* Only part of zip code, Zustellpostamt or eastern German code in column 5 (06_unknown) */ dprintf( g_state.outfiles[13], "%s\n", g_state.zip ); for( column = 0; column < 11; ++column ) { char * end = strchr( e, '\t' ); if( end ) { *end = '\n'; if( column == 9 && *e >= '0' && *e <= '9' ) write( g_state.outfiles[column+1], "0", 1 ); /* Augment Vorwahl */ write( g_state.outfiles[column+1], e, end - e + 1); e = end + 1; } else dprintf( g_state.outfiles[column+1], "%s\n", e); } } } void act_on_record( uint8_t *file, int flag, uint8_t *page, uint16_t record_off ) { uint8_t outbuf[8192], *out_dest = outbuf, *record; int num_entries; if( record_off > 0x1fff ) { /* Here comes the reference to a blob page */ int page_nr = *(uint16_t*)(page + ( record_off ^ 0xffff ) + 1); page = file + 0x800 + 0x2000 * page_nr; num_entries = *(uint16_t*)page; record = page + 0x6; } else { record = page + record_off; num_entries = *(uint16_t*)record; if( !flag ) { /* Flags says whether page has a prefix row */ record += 2; if( *(uint16_t*)record ) { decode_7bit_string( page + *(uint16_t*)record, outbuf ); out_dest += strlen( (char*)outbuf ); } } record += 2; } decode_7bit_string( record, out_dest ); split_to_files( outbuf, num_entries ); } /* Page is always 0x2000 */ void act_on_page( uint8_t *file, uint8_t *page, int page_nr ) { int record, flag = page[0]; int blob_len = *(uint16_t*)(page + 0x2); int num_records = *(uint16_t*)(page + 0x4); uint16_t * recs = (uint16_t*)(page + 0xe); /* printf( " Acting on page %04d with flag %d and %04d records.\n", page_nr, flag, num_records ); */ (void)page_nr; /* silence warning about unused variable */ if( !blob_len ) /* Handle blob, pages when they're referenced */ for( record = 0; record < num_records; ++record ) act_on_record( file, flag, page + 0xe, recs[record] ); } void act_on_file( uint8_t *file ) { int page, num_pages = *(uint16_t*)(file+0x40); int num_records = *(uint32_t*)(file+0x42); char *gasse = (char *)file + 0x8e; char *ort = gasse + 1 + strlen(gasse); char *zip = ort + 1 + strlen(ort); char *vorwahl = zip + 1 + strlen(zip); char ort_conv[1024]; int i; for(i=0; ort[i]; ++i ) ort_conv[i] = cp437_to_iso8859_1_table[((uint8_t*)ort)[i]]; ort_conv[i] = 0; /* printf( "Working on a %04d page and %06d records file, city: %4s %-32s with prefix %s\n", num_pages, num_records, zip, ort_conv, vorwahl ); */ g_state.ort = ort_conv; g_state.zip = zip; g_state.vorwahl = vorwahl; for( page = 0; page < num_pages; ++page ) act_on_page( file, file + 0x800 + 0x2000 * page, page ); } int main( ) { char filename[1024]; MAP f; int i; for( i=0; i<14; ++i ) { sprintf( filename, "%02d_unknown", i+1 ); g_state.outfiles[i] = open( filename, O_WRONLY | O_APPEND | O_CREAT, 0644 ); } while( fgets( filename, sizeof(filename), stdin ) ) { filename[strlen(filename)-1] = 0; /* fgets sucks */ f = map_file( filename, 1 ); act_on_file( f->addr ); unmap_file( &f ); } for( i=0; i<14; ++i ) close( g_state.outfiles[i] ); return 0; }