static char *
category_file(void)
{
if (dragonfly)
(void) snprintf(locpath, sizeof (locpath), "%s.%s",
locname, category_name());
else
(void) snprintf(locpath, sizeof (locpath), "%s/%s",
locname, category_name());
return (locpath);
}
int average()
{
struct Entry *temp;
if (head == NULL)
{
printf("No records found.\n");
return -1;
}
else
{
float possible_points = 0.0;
float earned_points = 0.0;
temp = head;
while (temp != NULL)
{
printf("\n\nName: %s\n", temp->name);
printf("Category: %s\n", category_name(temp->type));
printf("Possible Points: %f\n", temp->possible_points);
printf("Points Earned: %f\n", temp->points);
possible_points += temp->possible_points;
earned_points += temp->points;
temp = temp->next;
}
if (possible_points != 0)
printf("\n\nAverage Grade: %0.2f%%\n\n", (earned_points / possible_points) * 100);
return 0;
}
}
FILE *
open_category(void)
{
FILE *file;
if (verbose) {
(void) printf("Writing category %s: ", category_name());
(void) fflush(stdout);
}
/* make the parent directory */
if (!dragonfly)
(void) mkdir(dirname(category_file()), 0755);
/*
* note that we have to regenerate the file name, as dirname
* clobbered it.
*/
file = fopen(category_file(), "w");
if (file == NULL) {
errf(strerror(errno));
return (NULL);
}
return (file);
}
static char *
category_file(void)
{
(void) snprintf(locpath, sizeof (locpath), "%s/%s/LCL_DATA",
locname, category_name());
return (locpath);
}
/* Report versbose identifier info */
void input_verbose(cuuid_t u) {
char sflag;
int version = 0;
char structure[32];
char rev[32];
switch (u.version) {
case QUID_REV4:
version = 4;
strcpy(structure, "memgrep");
strcpy(rev, "REV2012");
break;
case QUID_REV7:
version = 7;
strcpy(structure, "ChaCha/4");
strcpy(rev, "REV2017");
break;
}
printf("---------------------------------------------\n");
printf("Health : %s\n", quid_validate(&u) ? "OK" : "INVALID");
printf("Timestamp (UTC) : %s", asctime(quid_timestamp(&u)));
printf("Microtime : %fms\n", quid_microtime(&u)/1000.0);
printf("Structure : %s\n", structure);
printf("Cuuid version : %d (%s)\n", version, rev);
printf("Tag : %s\n", quid_tag(&u));
printf("Category : %s\n", category_name(quid_category(&u)));
/* Remove NULL */
sflag = (quid_flag(&u) ^ IDF_NULL);
printf("Flags :\n");
if (sflag & FLAG_PUBLIC) {
printf(" * Public\n");
}
if (sflag & FLAG_IDSAFE) {
printf(" * Safe\n");
}
if (sflag & FLAG_MASTER) {
printf(" * master\n");
}
if (sflag & FLAG_SIGNED) {
printf(" * Signed\n");
}
if (sflag & FLAG_TAGGED) {
printf(" * Tagged\n");
}
if (sflag & FLAG_STRICT) {
printf(" * Strict\n");
}
printf("---------------------------------------------\n");
}
/*
* This function is used when copying the category from another
* locale. Note that the copy is actually performed using a hard
* link for efficiency.
*/
void
copy_category(char *src)
{
char srcpath[PATH_MAX];
int rv;
(void) snprintf(srcpath, sizeof (srcpath), "%s/%s",
src, category_name());
rv = access(srcpath, R_OK);
if ((rv != 0) && (strchr(srcpath, '/') == NULL)) {
/* Maybe we should try the system locale */
(void) snprintf(srcpath, sizeof (srcpath),
"/usr/lib/locale/%s/%s", src, category_name());
rv = access(srcpath, R_OK);
}
if (rv != 0) {
fprintf(stderr,"source locale data unavailable: %s", src);
return;
}
if (verbose > 1) {
(void) printf("Copying category %s from %s: ",
category_name(), src);
(void) fflush(stdout);
}
/* make the parent directory */
if (!dragonfly)
(void) mkdir(dirname(category_file()), 0755);
if (link(srcpath, category_file()) != 0) {
fprintf(stderr,"unable to copy locale data: %s",
strerror(errno));
return;
}
if (verbose > 1) {
(void) printf("done.\n");
}
}
wstring
TransferRule::gen_apertium_transfer_head(bool debug) {
locale loc(setlocale(LC_ALL,""));
wstring head=L"";
head+=L"<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n";
head+=L"<!-- -*- nxml -*- -->\n";
head+=L"<transfer>\n";
head+=L"<section-def-cats>\n";
set<pair<wstring,wstring> >::iterator it;
for(it=categories.begin(); it!=categories.end(); it++) {
head+=L" <def-cat n=\"CAT__"+category_name(it->first,it->second)+L"\">\n";
if (it->first.length()>0) //There is a lemma
{
head+=L" <cat-item lemma=\""+StringUtils::substitute(it->first,L"_",L" ")+L"\" tags=\""+escape_attr(it->second)+L"\"/>\n";
//Lemma in upper case too
head+=L" <cat-item lemma=\"";
head+=toupper<wchar_t>(StringUtils::substitute(it->first,L"_",L" ")[0],loc);
head+=StringUtils::substitute(it->first,L"_",L" ").substr(1)+L"\" tags=\""+escape_attr(it->second)+L"\"/>\n";
}
else
{
head+=L" <cat-item tags=\""+escape_attr(it->second)+L"\"/>\n";
if(sm_generalise || it->second[it->second.size()-1]==L'*' )
head+=L" <cat-item tags=\""+escape_attr((it->second).substr(0,it->second.size()-2))+L"\"/>\n";
}
head+=L" </def-cat>\n";
}
head+=L"<def-cat n=\"CAT__HASGENDER_NUMBER\"><cat-item tags=\"*.m.*\"/><cat-item tags=\"*.f.*\"/><cat-item tags=\"*.sg.*\"/><cat-item tags=\"*.pl.*\"/></def-cat>";
head+=L" <def-cat n=\"CAT__ND_GD\">\n";
head+=L" <cat-item tags=\"*.mf.*\"/>\n";
head+=L" <cat-item tags=\"*.sp.*\"/>\n";
head+=L" <cat-item tags=\"*.mf.sp.*\"/>\n";
head+=L" <cat-item tags=\"*.sp.mf.*\"/>\n";
head+=L" <cat-item tags=\"*.mf.*.sp.*\"/>\n";
head+=L" <cat-item tags=\"*.sp.*.mf.*\"/>\n";
head+=L" </def-cat>\n";
head+=L" <def-cat n=\"sent\">\n";
head+=L" <cat-item tags=\"sent\"/>\n";
head+=L" <cat-item tags=\"RESsent.sent\"/>\n";
head+=L" </def-cat>\n";
head+=L" <def-cat n=\"any\">\n";
head+=L" <cat-item tags=\"*\"/>\n";
head+=L" </def-cat>\n";
head+=L"</section-def-cats>\n";
head+=L"<section-def-attrs>\n";
//set<string>::iterator it2;
//for(it2=attributes.begin(); it2!=attributes.end(); it2++) {
// head+=" <def-attr n=\"ATTR__"+category_name("",(*it2))+"\">\n";
// head+=" <attr-item tags=\""+(*it2)+"\"/>\n";
// head+=" </def-attr>\n";
//}
map<wstring,pair< set<wstring>, set<wstring> > >::iterator it2;
for(it2=attributes.begin(); it2!=attributes.end(); it2++) {
head+=L" <def-attr n=\"learned_"+(*it2).first+L"\">\n";
set<wstring> tagsfromattr=(*it2).second.first;
for(set<wstring>::iterator it3= tagsfromattr.begin(); it3!=tagsfromattr.end(); ++it3)
head+=L" <attr-item tags=\""+StringUtils::substitute((*it3),L"+",L"\\+")+L"\"/>\n";
head+=L" </def-attr>\n";
}
head+=L" <def-attr n=\"learned_gen\">\n";
head+=L" <attr-item tags=\"m\"/>\n";
head+=L" <attr-item tags=\"f\"/>\n";
head+=L" <attr-item tags=\"mf\"/>\n";
head+=L" <attr-item tags=\"GD\"/>\n";
head+=L" </def-attr>\n";
head+=L" <def-attr n=\"learned_num\">\n";
head+=L" <attr-item tags=\"sg\"/>\n";
head+=L" <attr-item tags=\"pl\"/>\n";
head+=L" <attr-item tags=\"sp\"/>\n";
head+=L" <attr-item tags=\"ND\"/>\n";
head+=L" </def-attr>\n";
//head+=" <def-attr n=\"ATTR__notused\">\n";
//head+=" <attr-item tags=\"this.attr.will.not.be.used\"/>\n";
//head+=" </def-attr>\n";
head+=L"</section-def-attrs>\n";
head+=L"<section-def-vars>\n";
head+=L" <def-var n=\"genre\"/>\n";
head+=L" <def-var n=\"number\"/>\n";
head+=L"</section-def-vars>\n";
head+=L"<section-def-macros>\n";
head+=L"<def-macro n=\"f_bcond\" npar=\"2\">\n";
head+=L"<!--To test whether a blank contains format information.\n";
head+=L"If no format information is present it is removed. -->\n";
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <not>\n";
head+=L" <equal>\n";
head+=L" <b pos=\"1\"/>\n";
head+=L" <lit v=\" \"/>\n";
head+=L" </equal>\n";
head+=L" </not>\n";
head+=L" </test>\n";
head+=L" <out>\n";
head+=L" <b pos=\"1\"/>\n";
head+=L" </out>\n";
head+=L" </when>\n";
head+=L" </choose>\n";
head+=L"</def-macro>\n";
head+=L"<def-macro n=\"f_genre_num\" npar=\"1\">\n";
head+=L"<!--To set the global value storing the TL genre of the last seen word. -->\n";
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <clip pos=\"1\" side=\"tl\" part=\"learned_gen\"/>\n";
head+=L" <lit-tag v=\"m\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
head+=L" <let><var n=\"genre\"/><lit-tag v=\"m\"/></let>\n";
head+=L" </when>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <clip pos=\"1\" side=\"tl\" part=\"learned_gen\"/>\n";
head+=L" <lit-tag v=\"f\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
head+=L" <let><var n=\"genre\"/><lit-tag v=\"f\"/></let>\n";
head+=L" </when>\n";
head+=L" </choose>\n";
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <clip pos=\"1\" side=\"tl\" part=\"learned_num\"/>\n";
head+=L" <lit-tag v=\"sg\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
head+=L" <let><var n=\"number\"/><lit-tag v=\"sg\"/></let>\n";
head+=L" </when>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <clip pos=\"1\" side=\"tl\" part=\"learned_num\"/>\n";
head+=L" <lit-tag v=\"pl\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
head+=L" <let><var n=\"number\"/><lit-tag v=\"pl\"/></let>\n";
head+=L" </when>\n";
head+=L" </choose>\n";
head+=L"</def-macro>\n";
head+=L"<def-macro n=\"f_set_genre_num\" npar=\"1\">\n";
head+=L"<!--To set the genre of those words with GD, and the number of those words with ND. -->\n";
head+=L"<!--This is only used in no alignment template at all is applied. -->\n";
if(no_vars_determined)
{
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <clip pos=\"1\" side=\"tl\" part=\"learned_gen\"/>\n";
head+=L" <lit-tag v=\"GD\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
//head+=L" <let><clip pos=\"1\" side=\"tl\" part=\"gen\"/><lit-tag v=\"m\"/></let>\n";
head+=L" </when>\n";
head+=L" </choose>\n";
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <clip pos=\"1\" side=\"tl\" part=\"learned_num\"/>\n";
head+=L" <lit-tag v=\"ND\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
//head+=L" <let><clip pos=\"1\" side=\"tl\" part=\"num\"/><lit-tag v=\"sg\"/></let>\n";
head+=L" </when>\n";
head+=L" </choose>\n";
}
else
{
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <clip pos=\"1\" side=\"tl\" part=\"learned_gen\"/>\n";
head+=L" <lit-tag v=\"GD\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <var n=\"genre\"/>\n";
head+=L" <lit-tag v=\"f\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
head+=L" <let><clip pos=\"1\" side=\"tl\" part=\"learned_gen\"/><lit-tag v=\"f\"/></let>\n";
head+=L" </when>\n";
head+=L" <otherwise>\n";
head+=L" <let><clip pos=\"1\" side=\"tl\" part=\"learned_gen\"/><lit-tag v=\"m\"/></let>\n";
head+=L" </otherwise>\n";
head+=L" </choose>\n";
head+=L" </when>\n";
head+=L" </choose>\n";
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <clip pos=\"1\" side=\"tl\" part=\"learned_num\"/>\n";
head+=L" <lit-tag v=\"ND\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
head+=L" <choose>\n";
head+=L" <when>\n";
head+=L" <test>\n";
head+=L" <equal>\n";
head+=L" <var n=\"number\"/>\n";
head+=L" <lit-tag v=\"pl\"/>\n";
head+=L" </equal>\n";
head+=L" </test>\n";
head+=L" <let><clip pos=\"1\" side=\"tl\" part=\"learned_num\"/><lit-tag v=\"pl\"/></let>\n";
head+=L" </when>\n";
head+=L" <otherwise>\n";
head+=L" <let><clip pos=\"1\" side=\"tl\" part=\"learned_num\"/><lit-tag v=\"sg\"/></let>\n";
head+=L" </otherwise>\n";
head+=L" </choose>\n";
head+=L" </when>\n";
head+=L" </choose>\n";
}
head+=L"</def-macro>\n";
head+=L"</section-def-macros>\n";
head+=L"<section-rules>\n";
return head;
}
wstring
TransferRule::gen_apertium_transfer_rule(bool debug) {
locale loc(setlocale(LC_ALL,""));
wstring rule=L"";
bool include_otherwise=true;
if (ats.size()==0) {
cerr<<"Error in TransferRule::gen_apertium_transfer_rule: No alignment templates available\n";
exit(EXIT_FAILURE);
}
//Sort the AT so as to apply always the most frequent AT that
//satisfies the restrictions
AlignmentTemplateGreaterThanByCount atcomparer;
sort(ats.begin(), ats.end(), atcomparer);
//debug
//cerr<<"ats.size(): "<<ats.size()<<endl;
rule+=L"<rule>\n";
//The pattern to detect is the same for all AT within this transfer rule
rule+=L" <pattern>\n";
wstring chunkName=L"";
vector<wstring> generalpattern=StringUtils::split_wstring(source,L" ");
//for(unsigned i=0; i<ats[0].source.size(); i++) {
for(unsigned i=0; i<generalpattern.size(); i++) {
wstring lemma=Utils::get_lemma(generalpattern[i]);
wstring tags=Utils::tags2transferformat(Utils::get_tags(generalpattern[i]));
wstring tagsnogen=tags;
remove_generalised_tags(tagsnogen);
chunkName+=L"__";
if(sm_generalise)
{
wstring tmpstr=Utils::get_first_tag(Utils::get_tags(generalpattern[i])).substr(1);
rule+=L" <pattern-item n=\"CAT__"+category_name(L"",tmpstr.substr(0,tmpstr.size()-1)+L".*")+L"\"/>\n";
chunkName+=category_name(L"",tmpstr.substr(0,tmpstr.size()-1)+L".*");
}
else
{
rule+=L" <pattern-item n=\"CAT__"+category_name(lemma,tagsnogen)+L"\"/>\n";
chunkName+=category_name(lemma,tagsnogen);
}
}
rule+=L" </pattern>\n";
rule+=L" <action>\n";
rule+=L" <choose>\n";
//There is a set of different actions depending on the TL side of
//each AT. Consequently, there's one <when> statement per AT
for(unsigned i=0; i<ats.size(); i++) {
rule+=L" <when>";
rule+=L"<!--"+ats[i].to_wstring()+L"-->\n";
rule+=L" <test>\n";
int nconditions=0;
wstring teststr=L"";
//This AT can be applied if all restrictions are met
for(unsigned j=0; j<ats[i].restrictions.size(); j++){
if (ats[i].restrictions[j]!=L"__CLOSEWORD__") {
nconditions++;
//teststr+=L" <or>\n";
if (empty_restrictions_match_everything)
{
wstring source_tags_transfer_format=Utils::tags2transferformat(Utils::get_tags(ats[i].source[j]));
remove_generalised_tags(source_tags_transfer_format);
remove_final_asterisk(source_tags_transfer_format);
vector<wstring> nongenTags=StringUtils::split_wstring(source_tags_transfer_format,L".");
teststr+=L" <and>\n";
teststr+=L" <begins-with>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\"tags\" />\n";
teststr+=L" <lit-tag v=\""+escape_attr(StringUtils::split_wstring(Utils::tags2transferformat(ats[i].restrictions[j]),L".")[0])+L"\"/>\n";
teststr+=L" </begins-with>\n";
vector<wstring> resTags=StringUtils::split_wstring(Utils::tags2transferformat(ats[i].restrictions[j]),L".");
wstring partofspeech=resTags[0];
for (unsigned k=1 ; k<resTags.size(); k++)
{
wstring tag=resTags[k];
if(tag[0]==L'['){
wstring attr=tag.substr(1);
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\"learned_"+attr+L"\"/>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"sl\" part=\"learned_"+attr+L"\"/>\n";
teststr+=L" </equal>\n";
}
else if(tag[0]==L']'){
wstring attr=tag.substr(1);
teststr+=L" <not><equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\"learned_"+attr+L"\"/>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"sl\" part=\"learned_"+attr+L"\"/>\n";
teststr+=L" </equal></not>\n";
}
else{
wstring attr=get_attribute_for_tag(tag,partofspeech);
//teststr+=L" <or>\n";
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\""+attr+L"\"/>\n";
if(tag.size() >= 10 && tag.substr(0,10)==L"empty_tag_" )
teststr+=L" <lit v=\"\"/>\n";
else
teststr+=L" <lit-tag v=\""+escape_attr(tag)+L"\"/>\n";
teststr+=L" </equal>\n";
//Not necessary anymore
//A restriction with the same value that the sl tag may also mean that the tag dissappears (e.g., genders in es-en)
//if(std::find(nongenTags.begin(), nongenTags.end(), tag) != nongenTags.end()) {
//teststr+=L" <equal>\n";
//teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\""+attr+L"\"/>\n";
//teststr+=L" <lit v=\"\"/>\n";
//teststr+=L" </equal>\n";
//}
//teststr+=L" </or>\n";
}
}
teststr+=L" </and>\n";
}
else
{
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\"tags\" queue=\"no\"/>\n";
teststr+=L" <lit-tag v=\""+escape_attr(Utils::tags2transferformat(ats[i].restrictions[j]))+L"\"/>\n";
teststr+=L" </equal>\n";
}
int targetWordPos=ats[i].get_open_target_word_pos(j,false);
if(targetWordPos!=-1)
{
wstring target_transfer_format=Utils::tags2transferformat(Utils::get_tags(ats[i].target[targetWordPos]));
bool isGeneralised=false;
for(int myi=0; myi<target_transfer_format.size(); myi++)
{
if(target_transfer_format[myi] == L'*')
{
target_transfer_format=target_transfer_format.substr(0,myi-1);
isGeneralised=true;
break;
}
}
if(!isGeneralised){
if(!no_double_check_restrictions)
{
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\"tags\" queue=\"yes\"/>\n";
teststr+=L" <lit-tag v=\""+escape_attr(target_transfer_format)+L"\"/>\n";
teststr+=L" </equal>\n";
}
}
else{
vector<wstring> tltags=StringUtils::split_wstring(target_transfer_format,L".");
wstring partofspeech=tltags[0];
tltags.erase(tltags.begin());
vector<wstring> tlattrs;
for(vector<wstring>::iterator it=tltags.begin(); it!=tltags.end(); ++it){
wstring attribute=get_attribute_for_tag((*it),*(tltags.begin()));
tlattrs.push_back(attribute);
}
if(!no_double_check_restrictions)
{
if(tltags.size() > 0)
teststr+=L" <and>\n";
if(attributes.find(attribute_pos_group_name(partofspeech))==attributes.end())
{
pair<wstring,pair<set<wstring>, set<wstring> > > newelement;
newelement.first=attribute_pos_group_name(partofspeech);
pair<set<wstring>, wstring> newelementvalue;
newelementvalue.first.insert(partofspeech);
//newelementvalue.second=L"";
attributes.insert(newelement);
}
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\""+attribute_pos_group_name(partofspeech)+L"\"/>\n";
teststr+=L" <lit-tag v=\""+partofspeech+L"\"/>\n";
teststr+=L" </equal>\n";
for(int vcounter=0; vcounter < tltags.size(); vcounter++){
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\""+tlattrs[vcounter]+L"\"/>\n";
teststr+=L" <lit-tag v=\""+escape_attr(tltags[vcounter])+L"\"/>\n";
teststr+=L" </equal>\n";
}
if(tltags.size() > 0)
teststr+=L" </and>\n";
}
}
}
//teststr+=L" </or>\n";
}
//If we are working with generalised ATs, check also SL side, as
//ATs with different left side may be grouped in the same rule
if(sm_generalise || provided_patterns || using_explicit_empty_tags)
{
wstring source_tags_transfer_format=Utils::tags2transferformat(Utils::get_tags(ats[i].source[j]));
wstring source_lemma=Utils::get_lemma(ats[i].source[j]);
if(source_lemma.size() > 0 && (sm_generalise || provided_patterns))
{
nconditions++;
teststr+=L" <or>";
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"sl\" part=\"lem\" />\n";
teststr+=L" <lit v=\""+StringUtils::substitute(source_lemma,L"_",L" ")+L"\"/>\n";
teststr+=L" </equal>\n";
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"sl\" part=\"lem\" />\n";
teststr+=L" <lit v=\"";
teststr+=toupper<wchar_t>(StringUtils::substitute(source_lemma,L"_",L" ")[0],loc);
teststr+=StringUtils::substitute(source_lemma,L"_",L" ").substr(1)+L"\"/>\n";
teststr+=L" </equal>\n";
teststr+=L" </or>\n";
}
vector<wstring> alltags=StringUtils::split_wstring(source_tags_transfer_format,L".");
remove_generalised_tags(source_tags_transfer_format);
remove_final_asterisk(source_tags_transfer_format);
vector<wstring> nongenTags=StringUtils::split_wstring(source_tags_transfer_format,L".");
for(int tagi=1; tagi<nongenTags.size(); tagi++)
{
if( !(nongenTags[tagi].size()>=3 && nongenTags[tagi].substr(0,3)==L"RES") )
{
nconditions++;
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"sl\" part=\""+get_attribute_for_tag(nongenTags[tagi],nongenTags[0])+L"\" />\n";
if(nongenTags[tagi].size() >= 10 && nongenTags[tagi].substr(0,10)==L"empty_tag_" )
teststr+=L" <lit v=\"\"/>\n";
else
teststr+=L" <lit-tag v=\""+escape_attr(nongenTags[tagi])+L"\"/>\n";
teststr+=L" </equal>\n";
}
}
//I think this chunk of code should be removed anyway but...
if (!empty_restrictions_match_everything)
for(int tagi=1; tagi<alltags.size(); tagi++)
{
wstring tag=alltags[tagi];
if(tag.size()>0 && tag[0]==L'*')
{
nconditions++;
teststr+=L" <equal>\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"sl\" part=\"learned_"+tag.substr(1)+L"\" />\n";
teststr+=L" <clip pos=\""+Utils::itoa(j+1)+L"\" side=\"tl\" part=\"learned_"+tag.substr(1)+L"\" />\n";
teststr+=L" </equal>\n";
}
}
}
}
if (nconditions==0) { //All words were close words. We introduce a
//condition that is always true
teststr+=L" <equal>\n";
teststr+=L" <lit v=\"TRUE\"/>\n";
teststr+=L" <lit v=\"TRUE\"/>\n";
teststr+=L" </equal>\n";
//LO COMENTO PARA QUE MIS SCRIPTS DE DEPURACION FUNCIONEN BIEN
//include_otherwise=false;
}
if (nconditions>1) // There are more than one restriction to test
rule+=L" <and>\n";
rule+=teststr;
if (nconditions>1)
rule+=L" </and>\n";
rule+=L" </test>\n";
if (debug) {
wstring s=StringUtils::substitute(ats[i].to_wstring(), L"><", L".");
s=StringUtils::substitute(s,L"<",L"-");
s=StringUtils::substitute(s,L">",L"");
rule+=L" <out>\n";
rule+=L" <lu><lit v=\"(rid:"+Utils::itoa(rule_id)+L" at:"+s+L")\"/></lu>\n";
rule+=L" </out>\n";
}
wstring letvarStrs=L"";
rule+=L" <out>\n";
if (generate_chunks)
{
rule+=L" <chunk name=\""+chunkName+L"\" case=\"caseFirstWord\">";
rule+=L" <tags>";
rule+=L" <tag><lit-tag v=\"LRN\"/></tag>";
rule+=L" </tags>";
}
int blank_pos=0;
for(unsigned j=0; j<ats[i].target.size(); j++) {
if (ats[i].target[j][0]!='<') { //It's a lexicalized word, we copy it as is
wstring target_tags=Utils::tags2transferformat(Utils::get_tags(ats[i].target[j]));
wstring tagstoprint=Utils::tags2transferformat(Utils::get_tags(ats[i].target[j]));
vector<wstring> attributeNames=extract_attribute_names(tagstoprint);
vector<wstring> tagvector=StringUtils::split_wstring(tagstoprint,L".");
//remove_generalised_tags(tagstoprint);
//remove_final_asterisk(tagstoprint);
int pos=-1;
//Some tags come from bilingual dictionary. Get correct gender and number in case of GD/ND
//if(attributeNames.size() > 0)
// {
// pos=ats[i].get_source_word_pos(j);
// rule+=L" <call-macro n=\"f_set_genre_num\">\n";
// rule+=L" <with-param pos=\""+Utils::itoa(pos+1)+L"\"/>\n";
// rule+=L" </call-macro>\n";
// }
// rule+=L" <out>\n";
rule+=L" <lu>\n";
rule+=L" <lit v=\""+StringUtils::substitute(Utils::get_lemma_without_queue(ats[i].target[j]),L"_",L" ")+L"\"/>\n";
//rule+=L" <lit-tag v=\""+tagstoprint+L"\"/>\n";
//Some tags come from bilingual dictionary.
for(vector<wstring>::iterator it=tagvector.begin(); it!=tagvector.end(); ++it){
if( (*it).substr(0,1)==L"*" )
rule+=L" <clip pos=\""+Utils::itoa(pos+1)+L"\" side=\"sl\" part=\"learned_"+(*it).substr(1)+L"\"/>\n";
else if ( (*it).substr(0,1)==L")" )
{
long locpos=Utils::wtol((*it).substr(1,4));
rule+=L" <clip pos=\""+Utils::itoa((int) (locpos+1))+L"\" side=\"tl\" part=\"learned_"+(*it).substr(4)+L"\"/>\n";
}
else if ( (*it).substr(0,1)==L"(" )
{
long locpos=Utils::wtol((*it).substr(1,4));
rule+=L" <clip pos=\""+Utils::itoa((int) (locpos+1))+L"\" side=\"sl\" part=\"learned_"+(*it).substr(4)+L"\"/>\n";
}
else
rule+=L" <lit-tag v=\""+escape_attr((*it))+L"\"/>\n";
}
//if(attributeNames.size() > 0)
//{
// for(vector<wstring>::iterator it=attributeNames.begin(); it!=attributeNames.end(); ++it){
// rule+=L" <clip pos=\""+Utils::itoa(pos+1)+L"\" side=\"sl\" part=\""+(*it)+L"\"/>\n";
// }
//}
//rule+=L" <lit-tag v=\""+target_tags+L"\"/>\n";
rule+=L" <lit v=\""+StringUtils::substitute(Utils::get_queue(ats[i].target[j]),L"_",L" ")+L"\"/>\n";
rule+=L" </lu>\n";
// rule+=L" </out>\n";
//Some tags come from bilingual dictionary. Copy gender/number to global variables
// if(attributeNames.size() > 0)
// {
// rule+=L" <call-macro n=\"f_genre_num\">\n";
// rule+=L" <with-param pos=\""+Utils::itoa(pos+1)+L"\"/>\n";
// rule+=L" </call-macro>\n";
// }
//Copy gender/number to global variables in case that they come directly from AT
wstring genre=Utils::get_tag_value(tagstoprint,L"m|f");
if(genre.length()>0)
letvarStrs+=L" <let><var n=\"genre\"/><lit-tag v=\""+genre+L"\"/></let>\n";
wstring number=Utils::get_tag_value(tagstoprint,L"sg|pl");
if(number.length()>0)
letvarStrs+=L" <let><var n=\"number\"/><lit-tag v=\""+number+L"\"/></let>\n";
} else {
wstring tagstoprint=Utils::tags2transferformat(Utils::get_tags(ats[i].target[j]));
vector<wstring> attributeNames=extract_attribute_names(tagstoprint);
vector<wstring> tagvector=StringUtils::split_wstring(tagstoprint,L".");
//remove_generalised_tags(tagstoprint);
//remove_final_asterisk(tagstoprint);
int pos=ats[i].get_first_open_source_word_pos(j);
//Some tags come from bilingual dictionary. Get correct gender and number in case of GD/ND
// if(attributeNames.size() > 0)
// {
// int pos=ats[i].get_source_word_pos(j);
// rule+=L" <call-macro n=\"f_set_genre_num\">\n";
// rule+=L" <with-param pos=\""+Utils::itoa(pos+1)+L"\"/>\n";
// rule+=L" </call-macro>\n";
// }
// rule+=L" <out>\n";
rule+=L" <lu>\n";
if(do_rbpe){
rule+=L" <clip pos=\""+Utils::itoa(pos+1)+L"\" side=\"sl\" part=\"lemh\"/>\n";
}
else{
rule+=L" <clip pos=\""+Utils::itoa(pos+1)+L"\" side=\"tl\" part=\"lemh\"/>\n";
}
for(vector<wstring>::iterator it=tagvector.begin(); it!=tagvector.end(); ++it){
if( (*it).substr(0,1)==L"*" )
rule+=L" <clip pos=\""+Utils::itoa(pos+1)+L"\" side=\"sl\" part=\"learned_"+(*it).substr(1)+L"\"/>\n";
else if ( (*it).substr(0,1)==L")" )
{
long locpos=Utils::wtol((*it).substr(1,4));
rule+=L" <clip pos=\""+Utils::itoa((int) (locpos+1))+L"\" side=\"tl\" part=\"learned_"+(*it).substr(4)+L"\"/>\n";
}
else if ( (*it).substr(0,1)==L"(" )
{
long locpos=Utils::wtol((*it).substr(1,4));
rule+=L" <clip pos=\""+Utils::itoa((int) (locpos+1))+L"\" side=\"sl\" part=\"learned_"+(*it).substr(4)+L"\"/>\n";
}
else
rule+=L" <lit-tag v=\""+escape_attr((*it))+L"\"/>\n";
}
//rule+=L" <lit-tag v=\""+tagstoprint+L"\"/>\n";
//for(vector<wstring>::iterator it=attributeNames.begin(); it!=attributeNames.end(); ++it){
// rule+=L" <clip pos=\""+Utils::itoa(pos+1)+L"\" side=\"tl\" part=\""+(*it)+L"\"/>\n";
//}
if(do_rbpe){
rule+=L" <clip pos=\""+Utils::itoa(pos+1)+L"\" side=\"sl\" part=\"lemq\"/>\n";
}
else{
rule+=L" <clip pos=\""+Utils::itoa(pos+1)+L"\" side=\"tl\" part=\"lemq\"/>\n";
}
rule+=L" </lu>\n";
// rule+=L" </out>\n";
//Some tags come from bilingual dictionary. Copy gender/number to global variables
// if(attributeNames.size() > 0)
// {
// rule+=L" <call-macro n=\"f_genre_num\">\n";
// rule+=L" <with-param pos=\""+Utils::itoa(pos+1)+L"\"/>\n";
// rule+=L" </call-macro>\n";
// }
//Copy gender/number to global variables in case that they come directly from AT
wstring genre=Utils::get_tag_value(tagstoprint,L"m|f");
if(genre.length()>0)
letvarStrs+=L" <let><var n=\"genre\"/><lit-tag v=\""+genre+L"\"/></let>\n";
wstring number=Utils::get_tag_value(tagstoprint,L"sg|pl");
if(number.length()>0)
letvarStrs+=L" <let><var n=\"number\"/><lit-tag v=\""+number+L"\"/></let>\n";
}
if (blank_pos<(int)(ats[i].source.size()-1)) {
// rule+=L" <out>\n";
rule+=L" <b pos=\""+Utils::itoa(blank_pos+1)+L"\"/>\n";
// rule+=L" </out>\n";
blank_pos++;
} else if (j<(ats[i].target.size()-1)) {
//TL output string has more words than the SL pattern detected
// rule+=L" <out>\n";
rule+=L" <b/>\n";
// rule+=L" </out>\n";
}
}
if (generate_chunks)
{
rule+=L" </chunk>\n";
}
rule+=L" </out>\n";
if (debug) {
rule+=L" <out>\n";
rule+=L" <lu><lit v=\"(END)\"/></lu>\n";
rule+=L" </out>\n";
}
//If there are remaining blanks we print them out if they have
//format information inside. This is caused by a SL input string
//longer than the TL output one
for (unsigned j=ats[i].target.size(); j<ats[i].source.size(); j++) {
rule+=L" <call-macro n=\"f_bcond\">\n";
rule+=L" <with-param pos=\""+Utils::itoa(j)+L"\"/>\n";
rule+=L" <with-param pos=\""+Utils::itoa(j+1)+L"\"/>\n";
rule+=L" </call-macro>\n";
}
rule+=letvarStrs;
rule+=L" </when>\n";
if(!include_otherwise) {
//As the condition will always be met it has no sense to include
//further ATs
break;
}
}
//Actions to perform when none of the ATs can be applied
//word-for-word translation
if(include_otherwise) {
rule+=L" <otherwise><!--Word-for-word translation-->\n";
if (debug) {
rule+=L" <out>\n";
rule+=L" <lu><lit v=\"(rid:"+Utils::itoa(rule_id)+L" at:word-for-word)\"/></lu>\n";
rule+=L" </out>\n";
}
if(use_discard_rule)
{
rule+=L" <reject-current-rule shifting=\"no\" />\n";
rule+=L" <out><chunk name=\"any\" case=\"caseFirstWord\"><tags> <tag><lit-tag v=\"LRN\"/></tag> </tags> <lu>";
if(do_rbpe){
rule+=L"<clip pos=\"1\" side=\"sl\" part=\"whole\"/>";
}
else{
rule+=L"<clip pos=\"1\" side=\"tl\" part=\"whole\"/>";
}
rule+=L"</lu></chunk></out> </action><!--isolated word-->\n";
}
else
{
for(unsigned i=0; i<ats[0].source.size(); i++) {
rule+=L" <call-macro n=\"f_genre_num\">\n";
rule+=L" <with-param pos=\""+Utils::itoa(i+1)+L"\"/>\n";
rule+=L" </call-macro>\n";
rule+=L" <call-macro n=\"f_set_genre_num\">\n";
rule+=L" <with-param pos=\""+Utils::itoa(i+1)+L"\"/>\n";
rule+=L" </call-macro>\n";
rule+=L" <out>\n";
rule+=L" <lu>\n";
rule+=L" <clip pos=\""+Utils::itoa(i+1)+L"\" side=\"tl\" part=\"whole\"/>\n";
//rule+=L" <clip pos=\""+Utils::itoa(i+1)+L"\" side=\"tl\" part=\"lemh\"/>\n";
//rule+=L" <clip pos=\""+Utils::itoa(i+1)+L"\" side=\"tl\" part=\"tags\"/>\n";
//rule+=L" <clip pos=\""+Utils::itoa(i+1)+L"\" side=\"tl\" part=\"lemq\"/>\n";
rule+=L" </lu>\n";
if (i<(ats[0].source.size()-1))
rule+=L" <b pos=\""+Utils::itoa(i+1)+L"\"/>\n";
rule+=L" </out>\n";
}
if (debug) {
rule+=L" <out>\n";
rule+=L" <lu><lit v=\"(END)\"/></lu>\n";
rule+=L" </out>\n";
}
}
rule+=L" </otherwise>\n";
}
rule+=L" </choose>\n";
rule+=L" </action>\n";
rule+=L"</rule>\n";
return rule;
}
/* Program main */
int main(int argc, char *argv[]) {
cuuid_t cuuid;
int c, rtn;
unsigned int n = 1;
char *fname = NULL;
FILE *fp = NULL;
int fout = 0, nout = 0, fmat = PRINT_FORMAT_HEX_BACKET, vbose = 0, gen = 1;
int option_index;
char tag[3] = {0x0, 0x0, 0x0};
static struct option long_options[] = {
{"category", required_argument, 0, 0},
{"set-safe", no_argument, 0, 0},
{"set-master", no_argument, 0, 0},
{"set-public", no_argument, 0, 0},
{"set-sign", no_argument, 0, 0},
{"set-tag", no_argument, 0, 0},
{"set-strict", no_argument, 0, 0},
{"list-categories",no_argument, 0, 0},
{"rev", required_argument, 0, 0},
{"tag", required_argument, 0, 0},
{"rand-seed", required_argument, 0, 0},
{"memory-seed", required_argument, 0, 0},
{"output-hex", no_argument, 0, 'x'},
{"output-number", no_argument, 0, 'i'},
{"verbose", no_argument, 0, 'V'},
{"version", no_argument, 0, 'v'},
{"help", no_argument, 0, 'h'},
{0, 0, 0, 0 }
};
/* Register interrupt handler */
signal(SIGINT, set_signint);
while (1) {
option_index = 0;
c = getopt_long(argc, argv, "c:d:o:qxivVh", long_options, &option_index);
if (c == -1)
break;
switch (c) {
case 0:
if (!strcmp("rand-seed", long_options[option_index].name)) {
quid_set_rnd_seed(atoi(optarg));
} else if (!strcmp("memory-seed", long_options[option_index].name)) {
quid_set_mem_seed(atoi(optarg));
} else if (!strcmp("rev", long_options[option_index].name)) {
switch (atoi(optarg)) {
case 4:
cuuid.version = QUID_REV4;
break;
case 7:
default:
cuuid.version = QUID_REV7;
break;
}
} else if (!strcmp("tag", long_options[option_index].name)) {
if (strlen(optarg) != 3) {
printf("tag must be 3 characters\n");
printf("see --help for more information\n");
return 1;
}
tag[0] = optarg[0];
tag[1] = optarg[1];
tag[2] = optarg[2];
} else if (!strcmp("list-categories", long_options[option_index].name)) {
printf("%d) %s\n", CLS_CMON, category_name(CLS_CMON));
printf("%d) %s\n", CLS_INFO, category_name(CLS_INFO));
printf("%d) %s\n", CLS_WARN, category_name(CLS_WARN));
printf("%d) %s\n", CLS_ERROR, category_name(CLS_ERROR));
gen = 0;
} else if (!strcmp("set-safe", long_options[option_index].name)) {
flg |= IDF_IDSAFE;
} else if (!strcmp("set-public", long_options[option_index].name)) {
flg |= IDF_PUBLIC;
} else if (!strcmp("set-master", long_options[option_index].name)) {
flg |= IDF_MASTER;
} else if (!strcmp("set-tag", long_options[option_index].name)) {
flg |= IDF_TAGGED;
} else if (!strcmp("set-strict", long_options[option_index].name)) {
flg |= IDF_STRICT;
} else if (!strcmp("set-sign", long_options[option_index].name)) {
flg |= IDF_SIGNED;
} else if (!strcmp("category", long_options[option_index].name)) {
cat = (char)atoi(optarg);
switch (cat) {
case CLS_CMON:
case CLS_INFO:
case CLS_WARN:
case CLS_ERROR:
break;
default:
printf("unknown category %d\n", cat);
printf("see --help for more information\n");
return 1;
}
}
break;
case 'c':
n = atoi(optarg);
break;
case 'd':
delay = atoi(optarg);
break;
case 'o':
fname = optarg;
fout = 1;
break;
case 'x':
fmat = PRINT_FORMAT_HEX;
break;
case 'i':
fmat = PRINT_FORMAT_DEC;
break;
case 'q':
nout = 1;
break;
case 'V':
vbose = 1;
break;
case 'v':
print_version();
return 0;
case 'h':
case '?':
default:
usage();
return 0;
}
}
/* Identifier as positional argument */
if (optind < argc) {
cuuid_t uuid;
while (optind < argc) {
printf("%s\t", argv[optind]);
if (quid_parse(argv[optind], &uuid)) {
printf("VALID\n");
if (vbose)
input_verbose(uuid);
} else
printf("INVALID\n");
optind++;
}
return 0;
}
/* Output new identifiers */
if (gen) {
gettimeofday(&t1, NULL);
/* File output */
if (fout) {
rtn = check_fname(fname);
if (!rtn) {
fp = fopen(fname, "a");
} else if (rtn == 1) {
printf("%s is a directory\n", fname);
return 1;
} else if (rtn == 2){
printf("%s already exists\n", fname);
return 1;
}
}
for (i=0; i<n; ++i) {
assert(cat != 0);
quid_create(&cuuid, flg, cat, tag);
if (intflag) {
break;
}
if (!fout){
if (!nout) {
quid_print(cuuid, fmat);
}
} else {
quid_print_file(fp, cuuid, fmat);
}
if (delay) {
qusleep(delay * 1000);
}
ticks = clock();
}
if (fp) {
fclose(fp);
}
gettimeofday(&t2, NULL);
}
/* Show counters */
if (vbose && gen) {
generate_verbose();
}
return 0;
}
