/**
* Loads a match list. Match lists are supposed to have been
* generated by the Locate program.
*/
struct match_list* load_match_list(U_FILE* f,OutputPolicy *output_policy,unichar *header,Abstract_allocator prv_alloc) {
struct match_list* l=NULL;
struct match_list* end_of_list=NULL;
int start,end,start_char,end_char,start_letter,end_letter;
Ustring* line=new_Ustring();
char is_an_output;
/* We read the header */
unichar foo=0;
if (header==NULL) {
header=&foo;
}
u_fscanf(f,"#%C\n",header);
OutputPolicy policy;
switch(*header) {
case 'D': {
policy=DEBUG_OUTPUTS;
/* In debug mode, we have to skip the debug header */
int n_graphs;
u_fscanf(f,"%d\n",&n_graphs);
while ((n_graphs--)>-1) {
/* -1, because we also have to skip the #[IMR] line */
readline(line,f);
}
break;
}
case 'M': policy=MERGE_OUTPUTS; break;
case 'R':
case 'T':
case 'X': policy=REPLACE_OUTPUTS; break;
case 'I':
default: policy=IGNORE_OUTPUTS; break;
}
if (output_policy!=NULL) {
(*output_policy)=policy;
}
while (6==u_fscanf(f,"%d.%d.%d %d.%d.%d",&start,&start_char,&start_letter,&end,&end_char,&end_letter)) {
/* We look if there is an output or not, i.e. a space or a new line */
int c=u_fgetc(f);
if (c==' ') {
/* If we have an output to read */
readline(line,f);
/* In debug mode, we have to stop at the char #1 */
int i=-1;
while (line->str[++i]!=1 && line->str[i]!='\0') {
}
line->str[i]='\0';
}
is_an_output=(policy!=IGNORE_OUTPUTS);
if (l==NULL) {
l=new_match(start,end,start_char,end_char,start_letter,end_letter,is_an_output?line->str:NULL,-1,NULL,prv_alloc);
end_of_list=l;
} else {
end_of_list->next=new_match(start,end,start_char,end_char,start_letter,end_letter,is_an_output?line->str:NULL,-1,NULL,prv_alloc);
end_of_list=end_of_list->next;
}
}
free_Ustring(line);
return l;
}
static int read_number(lua_State *L, UFILE *ufile) {
lua_Number d;
if (u_fscanf(ufile, LUA_NUMBER_SCAN, &d) == 1) {
lua_pushnumber(L, d);
return 1;
}
else return 0; /* read fails */
}
/**
* Loads the given offset file. Returns NULL in case of error.
*/
vector_offset* load_offsets(const VersatileEncodingConfig* vec,const char* name) {
U_FILE* f=u_fopen(vec,name,U_READ);
if (f==NULL) return NULL;
int a,b,c,d,n;
vector_offset* res=new_vector_offset();
while ((n=u_fscanf(f,"%d%d%d%d",&a,&b,&c,&d))!=EOF) {
if (n!=4) {
fatal_error("Corrupted offset file %s\n",name);
}
vector_offset_add(res,a,b,c,d);
}
u_fclose(f);
return res;
}
int main_SpellCheck(int argc,char* const argv[]) {
if (argc==1) {
usage();
return SUCCESS_RETURN_CODE;
}
VersatileEncodingConfig vec=VEC_DEFAULT;
int val,index=-1;
char mode=0;
char snt[FILENAME_MAX]="";
char txt[FILENAME_MAX]="";
char output[FILENAME_MAX]="";
char output_set=0;
char output_op='A';
SpellCheckConfig config;
config.max_errors=1;
config.max_SP_INSERT=1;
config.max_SP_SUPPR=1;
config.max_SP_SWAP=1;
config.max_SP_CHANGE=1;
for (int i=0;i<N_SPSubOp;i++) {
config.score[i]=default_scores[i];
}
config.min_length1=4;
config.min_length2=6;
config.min_length3=12;
config.input_op='D';
config.keyboard=NULL;
config.allow_uppercase_initial=0;
char foo;
bool only_verify_arguments = false;
UnitexGetOpt options;
while (EOF!=(val=options.parse_long(argc,argv,optstring_SpellCheck,lopts_SpellCheck,&index))) {
switch(val) {
case 's': {
strcpy(snt,options.vars()->optarg);
mode='s';
break;
}
case 'f': {
strcpy(txt,options.vars()->optarg);
mode='f';
break;
}
case 'o': {
if (options.vars()->optarg!=NULL) {
strcpy(output,options.vars()->optarg);
}
output_set=1;
break;
}
case 'I': {
if (!strcmp(options.vars()->optarg,"D") || !strcmp(options.vars()->optarg,"M") || !strcmp(options.vars()->optarg,"U")) {
config.input_op=options.vars()->optarg[0];
} else {
error("Invalid argument %s for option --input-op: should in [DMU]\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 'O': {
if (!strcmp(options.vars()->optarg,"O") || !strcmp(options.vars()->optarg,"A")) {
output_op=options.vars()->optarg[0];
} else {
error("Invalid argument %s for option --output-op: should in [OA]\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 1: {
config.keyboard=get_Keyboard(options.vars()->optarg);
if (config.keyboard==NULL) {
error("Invalid argument %s for option --keyboard:\nUse --show-keyboards to see possible values\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 2: {
print_available_keyboards(U_STDOUT);
return SUCCESS_RETURN_CODE;
}
case 10: {
if (1!=sscanf(options.vars()->optarg,"%u%c",&config.max_errors,&foo)) {
error("Invalid argument %s for --max-errors: should be an integer >=0\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 11: {
if (1!=sscanf(options.vars()->optarg,"%u%c",&config.max_SP_INSERT,&foo)) {
error("Invalid argument %s for --max-insert: should be an integer >=0\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 12: {
if (1!=sscanf(options.vars()->optarg,"%u%c",&config.max_SP_SUPPR,&foo)) {
error("Invalid argument %s for --max-suppr: should be an integer >=0\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 13: {
if (1!=sscanf(options.vars()->optarg,"%u%c",&config.max_SP_CHANGE,&foo)) {
error("Invalid argument %s for --max-change: should be an integer >=0\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 14: {
if (1!=sscanf(options.vars()->optarg,"%u%c",&config.max_SP_SWAP,&foo)) {
error("Invalid argument %s for --max-swap: should be an integer >=0\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 20: {
int* scores=config.score;
if (N_SPSubOp!=sscanf(options.vars()->optarg,"%d,%d,%d,%d,%d,%d,%d,%d,%d%c",
scores,scores+1,scores+2,scores+3,scores+4,scores+5,
scores+6,scores+7,scores+8,&foo)) {
error("Invalid argument %s for option --scores. See --help-scores\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 21: {
usage_scores();
return SUCCESS_RETURN_CODE;
}
case 22: {
if (3!=sscanf(options.vars()->optarg,"%u,%u,%u%c",
&config.min_length1,&config.min_length2,&config.min_length3,&foo)) {
error("Invalid argument %s for option --min-lengths\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 23: {
if (!strcmp(options.vars()->optarg,"yes")) {
config.allow_uppercase_initial=1;
} else if (!strcmp(options.vars()->optarg,"no")) {
config.allow_uppercase_initial=0;
} else {
error("Invalid argument %s for option --upper-initial\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
break;
}
case 'k': if (options.vars()->optarg[0]=='\0') {
error("Empty input_encoding argument\n");
return USAGE_ERROR_CODE;
}
decode_reading_encoding_parameter(&(vec.mask_encoding_compatibility_input),options.vars()->optarg);
break;
case 'q': if (options.vars()->optarg[0]=='\0') {
error("Empty output_encoding argument\n");
return USAGE_ERROR_CODE;
}
decode_writing_encoding_parameter(&(vec.encoding_output),&(vec.bom_output),options.vars()->optarg);
break;
case 'V': only_verify_arguments = true;
break;
case 'h': usage();
return SUCCESS_RETURN_CODE;
case ':': index==-1 ? error("Missing argument for option -%c\n",options.vars()->optopt) :
error("Missing argument for option --%s\n",lopts_SpellCheck[index].name);
return USAGE_ERROR_CODE;
case '?': index==-1 ? error("Invalid option -%c\n",options.vars()->optopt) :
error("Invalid option --%s\n",options.vars()->optarg);
return USAGE_ERROR_CODE;
}
index=-1;
}
if (options.vars()->optind==argc) {
error("Invalid arguments: rerun with --help\n");
return USAGE_ERROR_CODE;
}
if (mode==0) {
error("You must use either --snt or --file\n");
return USAGE_ERROR_CODE;
}
if (only_verify_arguments) {
// freeing all allocated memory
return SUCCESS_RETURN_CODE;
}
config.n_dics=argc-options.vars()->optind;
config.dics=(Dictionary**)malloc(config.n_dics*sizeof(Dictionary*));
if (config.dics==NULL) {
alloc_error("main_SpellCheck");
return ALLOC_ERROR_CODE;
}
for (int i=0;i<config.n_dics;i++) {
config.dics[i]=new_Dictionary(&vec,argv[i+options.vars()->optind]);
if (config.dics[i]==NULL) {
error("Cannot load dictionary %s\n",argv[i+options.vars()->optind]);
}
}
config.out=U_STDOUT;
config.n_input_lines=0;
config.n_output_lines=0;
if (mode=='s') {
/* When working with a .snt, we actually want to work on its err file */
get_snt_path(snt,txt);
strcat(txt,"err");
/* the output must be dlf, and we note the number of lines in the existing
* dlf file, if any */
get_snt_path(snt,output);
strcat(output,"dlf.n");
U_FILE* f=u_fopen(&vec,output,U_READ);
if (f!=NULL) {
u_fscanf(f,"%d",&(config.n_output_lines));
u_fclose(f);
}
get_snt_path(snt,output);
strcat(output,"dlf");
output_set=1;
/* and we force the values for -I and -O */
config.input_op='U';
output_op='A';
} else {
/* If mode=='f', we don't have anything to do since we already
* defined the default output to stdout */
}
if (output_set) {
if (output_op=='O') {
config.out=u_fopen(&vec,output,U_WRITE);
} else {
config.out=u_fopen(&vec,output,U_APPEND);
}
if (config.out==NULL) {
error("Cannot open output file %s\n",output);
for (int i=0;i<config.n_dics;i++) {
free_Dictionary(config.dics[i]);
}
free(config.dics);
return DEFAULT_ERROR_CODE;
}
}
config.modified_input=NULL;
char modified_input[FILENAME_MAX]="";
if (config.input_op!='D') {
strcpy(modified_input,txt);
strcat(modified_input,".tmp");
config.modified_input=u_fopen(&vec,modified_input,U_WRITE);
if (config.modified_input==NULL) {
error("Cannot open tmp file %s\n",modified_input);
if (config.out!=U_STDOUT) {
u_fclose(config.out);
}
for (int i=0;i<config.n_dics;i++) {
free_Dictionary(config.dics[i]);
}
free(config.dics);
return DEFAULT_ERROR_CODE;
}
}
config.in=u_fopen(&vec,txt,U_READ);
if (config.in==NULL) {
error("Cannot open file %s\n",txt);
u_fclose(config.modified_input);
if (config.out!=U_STDOUT) {
u_fclose(config.out);
}
for (int i=0;i<config.n_dics;i++) {
free_Dictionary(config.dics[i]);
}
free(config.dics);
return DEFAULT_ERROR_CODE;
}
/* We perform spellchecking */
spellcheck(&config);
/* And we clean */
u_fclose(config.in);
if (config.modified_input!=NULL) {
/* If we used a tmp file because the input file has to be modified,
* it's now time to actually modify it */
u_fclose(config.modified_input);
af_remove(txt);
af_rename(modified_input,txt);
}
if (config.out!=U_STDOUT) {
u_fclose(config.out);
}
for (int i=0;i<config.n_dics;i++) {
free_Dictionary(config.dics[i]);
}
free(config.dics);
/* Finally, we update the dlf.n and err.n files if mode=='s' */
if (mode=='s') {
get_snt_path(snt,output);
strcat(output,"err.n");
U_FILE* f=u_fopen(&vec,output,U_WRITE);
if (f!=NULL) {
u_fprintf(f,"%d",config.n_input_lines);
u_fclose(f);
}
if (config.input_op!='D') {
get_snt_path(snt,output);
strcat(output,"dlf.n");
U_FILE* fw=u_fopen(&vec,output,U_WRITE);
if (fw!=NULL) {
u_fprintf(fw,"%d",config.n_output_lines);
u_fclose(fw);
}
}
}
return SUCCESS_RETURN_CODE;
}