void run_cg(std::istream& is, std::ostream& os, const hfst::HfstTransducer& t)
{
std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> utf16conv;
// Simple debug function
std::ostringstream ss;
for (std::string line; std::getline(is, line);) {
os << line << std::endl;
std::match_results<const char*> result;
std::regex_match(line.c_str(), result, CG_LINE);
if(!result.empty() && result[3].length() != 0) {
const auto& sugg = proc_line(t, line);
if(sugg.suggest) {
const auto& ana = sugg.ana;
const auto& formv = sugg.sforms;
if(formv.empty()) {
os << ana << "\t" << "?" << std::endl;
}
else {
os << ana << "\t" << u16join(formv) << std::endl;
}
}
else {
const auto& errtype = sugg.errtype;
if(!errtype.empty()) {
os << utf16conv.to_bytes(errtype) << std::endl;
}
}
}
}
}
int
main(int argc, char *argv[])
{
static char dflt_fmt[] = "%Y-%m-%d\n\0H:%M:%S %Z\n";
yuck_t argi[1U];
char *outfmt = dflt_fmt;
char **infmt;
size_t ninfmt;
char **input;
size_t ninput;
int quietp;
int res = 0;
if (yuck_parse(argi, argc, argv)) {
res = 1;
goto out;
}
if (argi->format_arg) {
outfmt = argi->format_arg;
/* unescape sequences, maybe */
if (argi->backslash_escapes_flag) {
dt_io_unescape(outfmt);
}
} else if (argi->time_flag) {
outfmt[8] = ' ';
outfmt[9] = '%';
}
if (!argi->input_format_nargs) {
infmt = argi->args;
ninfmt = argi->nargs;
input = NULL;
ninput = 0;
} else {
infmt = argi->input_format_args;
ninfmt = argi->input_format_nargs;
input = argi->args;
ninput = argi->nargs;
}
/* get quiet predicate */
quietp = argi->quiet_flag;
/* get lines one by one, apply format string and print date/time */
if (ninput == 0) {
/* read from stdin */
proc_lines((const char*const*)infmt, ninfmt, outfmt, quietp);
} else {
const char *const *cinfmt = (const char*const*)infmt;
for (size_t i = 0; i < ninput; i++) {
proc_line(input[i], cinfmt, ninfmt, outfmt, quietp);
}
}
out:
yuck_free(argi);
return res;
}
int main(int argc, const char * argv[])
{
FILE *file = fopen(argv[1], "r");
char line[1024];
while (fgets(line, 1024, file)) {
proc_line(line);
}
return 0;
}
static void
mass_add_dur(const struct mass_add_clo_s *clo)
{
/* read lines from stdin
* interpret as dates
* add to reference duration
* output */
size_t lno = 0;
for (char *line; prchunk_haslinep(clo->pctx); lno++) {
size_t llen = prchunk_getline(clo->pctx, &line);
proc_line(clo, line, llen);
}
return;
}
static void
proc_lines(const char *const *fmt, size_t nfmt, const char *ofmt, int quietp)
{
size_t lno = 0;
void *pctx;
/* using the prchunk reader now */
if ((pctx = init_prchunk(STDIN_FILENO)) == NULL) {
serror("Error: could not open stdin");
return;
}
while (prchunk_fill(pctx) >= 0) {
for (char *line; prchunk_haslinep(pctx); lno++) {
(void)prchunk_getline(pctx, &line);
/* check if line matches */
proc_line(line, fmt, nfmt, ofmt, quietp);
}
}
/* get rid of resources */
free_prchunk(pctx);
return;
}
int
main(int argc, char *argv[])
{
static char dflt_fmt[] = "%Y-%m-%d\n\0H:%M:%S %Z\n";
yuck_t argi[1U];
char *outfmt = dflt_fmt;
char **infmt;
size_t ninfmt;
char **input;
size_t ninput;
int quietp;
int res = 0;
if (yuck_parse(argi, argc, argv)) {
res = 1;
goto out;
}
if (argi->format_arg) {
outfmt = argi->format_arg;
/* unescape sequences, maybe */
if (argi->backslash_escapes_flag) {
dt_io_unescape(outfmt);
}
} else if (argi->time_flag) {
outfmt[8] = ' ';
outfmt[9] = '%';
}
if (!argi->input_format_nargs) {
infmt = argi->args;
ninfmt = argi->nargs;
input = NULL;
ninput = 0;
} else {
infmt = argi->input_format_args;
ninfmt = argi->input_format_nargs;
input = argi->args;
ninput = argi->nargs;
}
/* get quiet predicate */
quietp = argi->quiet_flag;
/* set locale specific/independent behaviour */
with (const char *loc) {
if (!argi->locale_flag) {
loc = "C";
/* we need to null out TZ for UTC */
setenv("TZ", "", 1);
} else {
loc = "";
}
/* actually set our findings in stone */
setlocale(LC_TIME, loc);
tzset();
}
/* get lines one by one, apply format string and print date/time */
if (ninput == 0) {
/* read from stdin */
proc_lines((const char*const*)infmt, ninfmt, outfmt, quietp);
} else {
const char *const *cinfmt = (const char*const*)infmt;
for (size_t i = 0; i < ninput; i++) {
proc_line(input[i], cinfmt, ninfmt, outfmt, quietp);
}
}
out:
yuck_free(argi);
return res;
}
int
main(int argc, char *argv[])
{
struct gengetopt_args_info argi[1];
const char *ofmt;
char **fmt;
size_t nfmt;
int res = 0;
zif_t fromz = NULL;
zif_t z = NULL;
if (cmdline_parser(argc, argv, argi)) {
res = 1;
goto out;
}
/* init and unescape sequences, maybe */
ofmt = argi->format_arg;
fmt = argi->input_format_arg;
nfmt = argi->input_format_given;
if (argi->backslash_escapes_given) {
dt_io_unescape(argi->format_arg);
for (size_t i = 0; i < nfmt; i++) {
dt_io_unescape(fmt[i]);
}
}
/* try and read the from and to time zones */
if (argi->from_zone_given) {
fromz = zif_open(argi->from_zone_arg);
}
if (argi->zone_given) {
z = zif_open(argi->zone_arg);
}
if (argi->default_given) {
struct dt_dt_s dflt = dt_strpdt(argi->default_arg, NULL, NULL);
dt_set_default(dflt);
}
if (argi->inputs_num) {
for (size_t i = 0; i < argi->inputs_num; i++) {
const char *inp = argi->inputs[i];
struct dt_dt_s d = dt_io_strpdt(inp, fmt, nfmt, fromz);
if (!dt_unk_p(d)) {
dt_io_write(d, ofmt, z, '\n');
} else if (!argi->quiet_given) {
dt_io_warn_strpdt(inp);
}
}
} else {
/* read from stdin */
size_t lno = 0;
struct grep_atom_s __nstk[16], *needle = __nstk;
size_t nneedle = countof(__nstk);
struct grep_atom_soa_s ndlsoa;
void *pctx;
struct prln_ctx_s prln = {
.ndl = &ndlsoa,
.ofmt = ofmt,
.fromz = fromz,
.outz = z,
.sed_mode_p = argi->sed_mode_given,
.quietp = argi->quiet_given,
};
/* no threads reading this stream */
__io_setlocking_bycaller(stdout);
/* lest we overflow the stack */
if (nfmt >= nneedle) {
/* round to the nearest 8-multiple */
nneedle = (nfmt | 7) + 1;
needle = calloc(nneedle, sizeof(*needle));
}
/* and now build the needles */
ndlsoa = build_needle(needle, nneedle, fmt, nfmt);
/* using the prchunk reader now */
if ((pctx = init_prchunk(STDIN_FILENO)) == NULL) {
error(errno, "Error: could not open stdin");
goto ndl_free;
}
while (prchunk_fill(pctx) >= 0) {
for (char *line; prchunk_haslinep(pctx); lno++) {
size_t llen = prchunk_getline(pctx, &line);
proc_line(prln, line, llen);
}
}
/* get rid of resources */
free_prchunk(pctx);
ndl_free:
if (needle != __nstk) {
free(needle);
}
}
if (argi->from_zone_given) {
zif_close(fromz);
}
if (argi->zone_given) {
zif_close(z);
}
out:
cmdline_parser_free(argi);
return res;
}
void run_json(std::istream& is, std::ostream& os, const hfst::HfstTransducer& t, const msgmap& msgs)
{
json::sanity_test();
int pos = 0;
std::u16string errtype = u"default";
bool first_err = true;
LineType prevtype = BlankL;
bool is_addcohort = true;
std::ostringstream text;
Cohort c;
// TODO: could use http://utfcpp.sourceforge.net, but it's not in macports;
// and ICU seems overkill just for iterating codepoints
std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> utf16conv;
os << "{"
<< json::key(u"errs")
<< "[";
for (std::string line; std::getline(is, line);) {
std::match_results<const char*> result;
std::regex_match(line.c_str(), result, CG_LINE);
bool appendsugg = is_addcohort && prevtype != WordformL && !c.err.empty();
if (!result.empty() && ((result[2].length() != 0 && !appendsugg)
|| result[6].length() != 0)) {
proc_cohort(pos,
first_err,
c,
text,
os,
t,
msgs,
utf16conv);
c = Cohort();
}
if (!result.empty() && result[2].length() != 0) {
if(appendsugg) {
c.err = sugg_append(utf16conv.from_bytes(result[2]),
c.err);
}
is_addcohort = true;
c.form = utf16conv.from_bytes(result[2]);
prevtype = WordformL;
}
else if(!result.empty() && result[3].length() != 0) {
// TODO: doesn't do anything with subreadings yet; needs to keep track of previous line(s) for that
const auto& sugg = proc_line(t, line);
if(!sugg.errtype.empty()) {
errtype = sugg.errtype;
}
if(sugg.suggest) {
c.err[errtype].insert(sugg.sforms.begin(),
sugg.sforms.end());
}
else {
is_addcohort = false; // Seen at least one non-suggestion reading
}
prevtype = ReadingL;
}
else if(!result.empty() && result[6].length() != 0) {
const auto blank = clean_blank(result[6]);
pos += utf16conv.from_bytes(blank).size();
text << blank;
prevtype = BlankL;
}
else {
// Blank lines without the prefix don't go into text output!
prevtype = BlankL;
}
}
proc_cohort(pos,
first_err,
c,
text,
os,
t,
msgs,
utf16conv);
os << "]"
<< "," << json::key(u"text") << json::str(utf16conv.from_bytes(text.str()))
<< "}";
}
