// ----------------------------------------------------------------
static void mapper_step_free(void* pvstate) {
mapper_step_state_t* pstate = pvstate;
slls_free(pstate->pstepper_names);
slls_free(pstate->pvalue_field_names);
slls_free(pstate->pgroup_by_field_names);
// xxx free the level-2's 1st
lhmslv_free(pstate->groups);
}
static lrec_t* lrec_reader_csvex_process(void* pvhandle, void* pvstate, context_t* pctx) {
lrec_reader_csvex_state_t* pstate = pvstate;
// xxx byte-reader open ...
// if (pstate->pfr == NULL) {
// pstate->pfr = pfr_alloc((FILE*)pvhandle, pstate->peek_buf_len);
// }
if (pstate->expect_header_line_next) {
slls_t* pheader_fields = lrec_reader_csvex_get_fields(pstate);
if (pheader_fields == NULL)
return NULL;
pstate->ilno++;
pstate->expect_header_line_next = FALSE;
pstate->pheader_keeper = lhmslv_get(pstate->pheader_keepers, pheader_fields);
if (pstate->pheader_keeper == NULL) {
pstate->pheader_keeper = header_keeper_alloc(NULL, pheader_fields);
lhmslv_put(pstate->pheader_keepers, pheader_fields, pstate->pheader_keeper);
} else { // Re-use the header-keeper in the header cache
slls_free(pheader_fields);
}
}
pstate->ilno++;
slls_t* pdata_fields = lrec_reader_csvex_get_fields(pstate);
return paste_header_and_data(pstate, pdata_fields);
}
static lrec_t* lrec_reader_mmap_csvlite_process_multi_seps(void* pvstate, void* pvhandle, context_t* pctx) {
file_reader_mmap_state_t* phandle = pvhandle;
lrec_reader_mmap_csvlite_state_t* pstate = pvstate;
while (TRUE) {
if (pstate->expect_header_line_next) {
slls_t* pheader_fields = lrec_reader_mmap_csvlite_get_header_multi_seps(phandle, pstate);
if (pheader_fields == NULL) // EOF
return NULL;
pstate->expect_header_line_next = FALSE;
pstate->pheader_keeper = lhmslv_get(pstate->pheader_keepers, pheader_fields);
if (pstate->pheader_keeper == NULL) {
pstate->pheader_keeper = header_keeper_alloc(NULL, pheader_fields);
lhmslv_put(pstate->pheader_keepers, pheader_fields, pstate->pheader_keeper);
} else { // Re-use the header-keeper in the header cache
slls_free(pheader_fields);
}
}
int end_of_stanza = FALSE;
lrec_t* prec = lrec_reader_mmap_csvlite_get_record_multi_seps(phandle, pstate, pstate->pheader_keeper,
&end_of_stanza);
if (end_of_stanza) {
pstate->expect_header_line_next = TRUE;
} else if (prec == NULL) { // EOF
return NULL;
} else {
return prec;
}
}
}
// ----------------------------------------------------------------
static void lrec_writer_pprint_process(FILE* output_stream, lrec_t* prec, void* pvstate) {
lrec_writer_pprint_state_t* pstate = pvstate;
int drain = FALSE;
if (prec == NULL) {
drain = TRUE;
} else {
if (pstate->pprev_keys != NULL && !lrec_keys_equal_list(prec, pstate->pprev_keys)) {
drain = TRUE;
}
}
if (drain) {
if (pstate->num_blocks_written > 0LL) // separate blocks with empty line
fputs(pstate->ors, output_stream);
print_and_free_record_list(pstate->precords, output_stream, pstate->ors, pstate->ofs, pstate->left_align);
if (pstate->pprev_keys != NULL) {
slls_free(pstate->pprev_keys);
pstate->pprev_keys = NULL;
}
pstate->precords = sllv_alloc();
pstate->num_blocks_written++;
}
if (prec != NULL) {
sllv_append(pstate->precords, prec);
if (pstate->pprev_keys == NULL)
pstate->pprev_keys = mlr_copy_keys_from_record(prec);
}
}
static void lrec_writer_csv_free(void* pvstate) {
lrec_writer_csv_state_t* pstate = pvstate;
if (pstate->plast_header_output != NULL) {
slls_free(pstate->plast_header_output);
pstate->plast_header_output = NULL;
}
}
// ----------------------------------------------------------------
static void mapper_tail_free(void* pvstate) {
mapper_tail_state_t* pstate = pvstate;
if (pstate->pgroup_by_field_names != NULL)
slls_free(pstate->pgroup_by_field_names);
if (pstate->precord_lists_by_group != NULL)
// xxx free the void-star payloads 1st
lhmslv_free(pstate->precord_lists_by_group);
}
// ----------------------------------------------------------------
static void mapper_head_free(void* pvstate) {
mapper_head_state_t* pstate = (mapper_head_state_t*)pvstate;
if (pstate->pgroup_by_field_names != NULL)
slls_free(pstate->pgroup_by_field_names);
// xxx recursively free void-stars ... here & elsewhere.
lhmslv_free(pstate->precord_lists_by_group);
}
static lrec_t* lrec_reader_stdio_csv_process(void* pvhandle, void* pvstate, context_t* pctx) {
FILE* input_stream = pvhandle;
lrec_reader_stdio_csv_state_t* pstate = pvstate;
while (TRUE) {
if (pstate->expect_header_line_next) {
// xxx cmt
while (TRUE) {
char* hline = mlr_get_line(input_stream, pstate->irs);
if (hline == NULL) // EOF
return NULL;
pstate->ilno++;
slls_t* pheader_fields = split_csv_header_line(hline, pstate->ifs, pstate->allow_repeat_ifs);
if (pheader_fields->length == 0) {
pstate->expect_header_line_next = TRUE;
if (pstate->pheader_keeper != NULL) {
pstate->pheader_keeper = NULL;
}
} else {
pstate->expect_header_line_next = FALSE;
pstate->pheader_keeper = lhmslv_get(pstate->pheader_keepers, pheader_fields);
if (pstate->pheader_keeper == NULL) {
pstate->pheader_keeper = header_keeper_alloc(hline, pheader_fields);
lhmslv_put(pstate->pheader_keepers, pheader_fields, pstate->pheader_keeper);
} else { // Re-use the header-keeper in the header cache
slls_free(pheader_fields);
free(hline);
}
break;
}
}
}
char* line = mlr_get_line(input_stream, pstate->irs);
if (line == NULL) // EOF
return NULL;
// xxx empty-line check ... make a lib func is_empty_modulo_whitespace().
if (!*line) {
if (pstate->pheader_keeper != NULL) {
pstate->pheader_keeper = NULL;
pstate->expect_header_line_next = TRUE;
free(line);
continue;
}
} else {
pstate->ifnr++;
return lrec_parse_stdio_csv_data_line(pstate->pheader_keeper, line, pstate->ifs, pstate->allow_repeat_ifs);
}
}
}
static void lrec_writer_pprint_free(lrec_writer_t* pwriter) {
lrec_writer_pprint_state_t* pstate = pwriter->pvstate;
if (pstate->precords != NULL) {
sllv_free(pstate->precords);
pstate->precords = NULL;
}
if (pstate->pprev_keys != NULL) {
slls_free(pstate->pprev_keys);
pstate->pprev_keys = NULL;
}
free(pstate);
free(pwriter);
}
// ----------------------------------------------------------------
static lrec_t* lrec_reader_mmap_csv_process(void* pvstate, void* pvhandle, context_t* pctx) {
lrec_reader_mmap_csv_state_t* pstate = pvstate;
file_reader_mmap_state_t* phandle = pvhandle;
if (pstate->expect_header_line_next) {
if (!lrec_reader_mmap_csv_get_fields(pstate, pstate->pfields, phandle))
return NULL;
pstate->ilno++;
slls_t* pheader_fields = slls_alloc();
int i = 0;
for (rsllse_t* pe = pstate->pfields->phead; i < pstate->pfields->length && pe != NULL; pe = pe->pnext, i++) {
if (*pe->value == 0) {
fprintf(stderr, "%s: unacceptable empty CSV key at file \"%s\" line %lld.\n",
MLR_GLOBALS.argv0, pctx->filename, pstate->ilno);
exit(1);
}
// Transfer pointer-free responsibility from the rslls to the
// header fields in the header keeper
slls_append(pheader_fields, pe->value, pe->free_flag);
pe->free_flag = 0;
}
rslls_reset(pstate->pfields);
pstate->pheader_keeper = lhmslv_get(pstate->pheader_keepers, pheader_fields);
if (pstate->pheader_keeper == NULL) {
pstate->pheader_keeper = header_keeper_alloc(NULL, pheader_fields);
lhmslv_put(pstate->pheader_keepers, pheader_fields, pstate->pheader_keeper,
NO_FREE); // freed by header-keeper
} else { // Re-use the header-keeper in the header cache
slls_free(pheader_fields);
}
pstate->expect_header_line_next = FALSE;
}
int rc = lrec_reader_mmap_csv_get_fields(pstate, pstate->pfields, phandle);
pstate->ilno++;
if (rc == FALSE) // EOF
return NULL;
else {
lrec_t* prec = pstate->use_implicit_header
? paste_indices_and_data(pstate, pstate->pfields, pctx)
: paste_header_and_data(pstate, pstate->pfields, pctx);
rslls_reset(pstate->pfields);
return prec;
}
}
// ----------------------------------------------------------------
static void lrec_writer_csv_process(FILE* output_stream, lrec_t* prec, void* pvstate) {
if (prec == NULL)
return;
lrec_writer_csv_state_t* pstate = pvstate;
char *ors = pstate->ors;
char *ofs = pstate->ofs;
if (pstate->plast_header_output != NULL) {
if (!lrec_keys_equal_list(prec, pstate->plast_header_output)) {
slls_free(pstate->plast_header_output);
pstate->plast_header_output = NULL;
if (pstate->num_header_lines_output > 0LL)
fputs(ors, output_stream);
}
}
if (pstate->plast_header_output == NULL) {
int nf = 0;
if (!pstate->headerless_csv_output) {
for (lrece_t* pe = prec->phead; pe != NULL; pe = pe->pnext) {
if (nf > 0)
fputs(ofs, output_stream);
pstate->pquoted_output_func(output_stream, pe->key, pstate->ors, pstate->ofs,
pstate->orslen, pstate->ofslen);
nf++;
}
fputs(ors, output_stream);
}
pstate->plast_header_output = mlr_copy_keys_from_record(prec);
pstate->num_header_lines_output++;
}
int nf = 0;
for (lrece_t* pe = prec->phead; pe != NULL; pe = pe->pnext) {
if (nf > 0)
fputs(ofs, output_stream);
pstate->pquoted_output_func(output_stream, pe->value, pstate->ors, pstate->ofs,
pstate->orslen, pstate->ofslen);
nf++;
}
fputs(ors, output_stream);
pstate->onr++;
// See ../README.md for memory-management conventions
lrec_free(prec);
}
void ap_free(ap_state_t* pstate) {
if (pstate == NULL)
return;
for (sllve_t* pe = pstate->pflag_defs->phead; pe != NULL; pe = pe->pnext) {
ap_flag_def_t* pdef = pe->pvdata;
if (pdef->type == AP_STRING_LIST_FLAG && pdef->pval != NULL) {
slls_t** pplist = pdef->pval;
if (*pplist != NULL)
slls_free(*pplist);
}
free(pdef);
}
sllv_free(pstate->pflag_defs);
free(pstate);
}
static void lrec_writer_csv_process(FILE* output_stream, lrec_t* prec, void* pvstate) {
if (prec == NULL)
return;
lrec_writer_csv_state_t* pstate = pvstate;
char ors = pstate->ors;
char ofs = pstate->ofs;
if (pstate->plast_header_output != NULL) {
// xxx make a fcn to compare these w/o copy: put it in mixutil.
if (!lrec_keys_equal_list(prec, pstate->plast_header_output)) {
slls_free(pstate->plast_header_output);
pstate->plast_header_output = NULL;
if (pstate->num_header_lines_output > 0LL)
fputc(ors, output_stream);
}
}
if (pstate->plast_header_output == NULL) {
int nf = 0;
for (lrece_t* pe = prec->phead; pe != NULL; pe = pe->pnext) {
if (nf > 0)
fputc(ofs, output_stream);
fputs(pe->key, output_stream);
nf++;
}
fputc(ors, output_stream);
pstate->plast_header_output = mlr_copy_keys_from_record(prec);
pstate->num_header_lines_output++;
}
int nf = 0;
for (lrece_t* pe = prec->phead; pe != NULL; pe = pe->pnext) {
if (nf > 0)
fputc(ofs, output_stream);
fputs(pe->value, output_stream);
nf++;
}
fputc(ors, output_stream);
pstate->onr++;
lrec_free(prec); // xxx cmt mem-mgmt
}
// ----------------------------------------------------------------
static lrec_t* lrec_reader_stdio_xtab_process(void* pvstate, void* pvhandle, context_t* pctx) {
FILE* input_stream = pvhandle;
lrec_reader_stdio_xtab_state_t* pstate = pvstate;
if (pstate->at_eof)
return NULL;
slls_t* pxtab_lines = slls_alloc();
while (TRUE) {
char* line = (pstate->ifslen == 1)
? mlr_get_cline(input_stream, pstate->ifs[0])
: mlr_get_sline(input_stream, pstate->ifs, pstate->ifslen);
if (line == NULL) { // EOF
// EOF or blank line terminates the stanza.
pstate->at_eof = TRUE;
if (pxtab_lines->length == 0) {
slls_free(pxtab_lines);
return NULL;
} else {
return (pstate->ipslen == 1)
? lrec_parse_stdio_xtab_single_ips(pxtab_lines, pstate->ips[0], pstate->allow_repeat_ips)
: lrec_parse_stdio_xtab_multi_ips(pxtab_lines, pstate->ips, pstate->ipslen, pstate->allow_repeat_ips);
}
} else if (*line == '\0') {
free(line);
if (pxtab_lines->length > 0) {
return (pstate->ipslen == 1)
? lrec_parse_stdio_xtab_single_ips(pxtab_lines, pstate->ips[0], pstate->allow_repeat_ips)
: lrec_parse_stdio_xtab_multi_ips(pxtab_lines, pstate->ips, pstate->ipslen, pstate->allow_repeat_ips);
}
} else {
slls_add_with_free(pxtab_lines, line);
}
}
}
static void lrec_free_multiline_backing(lrec_t* prec) {
slls_free(prec->pxtab_lines);
}
static void lrec_writer_csv_free(lrec_writer_t* pwriter) {
lrec_writer_csv_state_t* pstate = pwriter->pvstate;
slls_free(pstate->plast_header_output);
free(pstate);
free(pwriter);
}
// ----------------------------------------------------------------
static lrec_t* lrec_reader_stdio_csvlite_process(void* pvstate, void* pvhandle, context_t* pctx) {
FILE* input_stream = pvhandle;
lrec_reader_stdio_csvlite_state_t* pstate = pvstate;
while (TRUE) {
if (pstate->expect_header_line_next) {
while (TRUE) {
char* hline = (pstate->irslen == 1)
? mlr_get_cline(input_stream, pstate->irs[0])
: mlr_get_sline(input_stream, pstate->irs, pstate->irslen);
if (hline == NULL) // EOF
return NULL;
pstate->ilno++;
slls_t* pheader_fields = (pstate->ifslen == 1)
? split_csvlite_header_line_single_ifs(hline, pstate->ifs[0], pstate->allow_repeat_ifs)
: split_csvlite_header_line_multi_ifs(hline, pstate->ifs, pstate->ifslen, pstate->allow_repeat_ifs);
if (pheader_fields->length == 0) {
pstate->expect_header_line_next = TRUE;
if (pstate->pheader_keeper != NULL) {
pstate->pheader_keeper = NULL;
}
} else {
for (sllse_t* pe = pheader_fields->phead; pe != NULL; pe = pe->pnext) {
if (*pe->value == 0) {
fprintf(stderr, "%s: unacceptable empty CSV key at file \"%s\" line %lld.\n",
MLR_GLOBALS.argv0, pctx->filename, pstate->ilno);
exit(1);
}
}
pstate->expect_header_line_next = FALSE;
pstate->pheader_keeper = lhmslv_get(pstate->pheader_keepers, pheader_fields);
if (pstate->pheader_keeper == NULL) {
pstate->pheader_keeper = header_keeper_alloc(hline, pheader_fields);
lhmslv_put(pstate->pheader_keepers, pheader_fields, pstate->pheader_keeper,
NO_FREE); // freed by header-keeper
} else { // Re-use the header-keeper in the header cache
slls_free(pheader_fields);
free(hline);
}
break;
}
}
}
char* line = (pstate->irslen == 1)
? mlr_get_cline(input_stream, pstate->irs[0])
: mlr_get_sline(input_stream, pstate->irs, pstate->irslen);
if (line == NULL) // EOF
return NULL;
pstate->ilno++;
if (!*line) {
if (pstate->pheader_keeper != NULL) {
pstate->pheader_keeper = NULL;
pstate->expect_header_line_next = TRUE;
free(line);
continue;
}
} else {
pstate->ifnr++;
if (pstate->ifslen == 1) {
return pstate->use_implicit_header
? lrec_parse_stdio_csvlite_data_line_single_ifs_implicit_header(
pstate->pheader_keeper, pctx->filename, pstate->ilno, line,
pstate->ifs[0], pstate->allow_repeat_ifs)
: lrec_parse_stdio_csvlite_data_line_single_ifs(pstate->pheader_keeper, pctx->filename, pstate->ilno, line,
pstate->ifs[0], pstate->allow_repeat_ifs);
} else {
return pstate->use_implicit_header
? lrec_parse_stdio_csvlite_data_line_multi_ifs_implicit_header(
pstate->pheader_keeper, pctx->filename, pstate->ilno, line,
pstate->ifs, pstate->ifslen, pstate->allow_repeat_ifs)
: lrec_parse_stdio_csvlite_data_line_multi_ifs(pstate->pheader_keeper, pctx->filename, pstate->ilno, line,
pstate->ifs, pstate->ifslen, pstate->allow_repeat_ifs);
}
}
}
}
// ----------------------------------------------------------------
static lrec_t* lrec_reader_stdio_csv_process(void* pvstate, void* pvhandle, context_t* pctx) {
lrec_reader_stdio_csv_state_t* pstate = pvstate;
// Ingest the next header line, if expected
if (pstate->expect_header_line_next) {
while (TRUE) {
if (!lrec_reader_stdio_csv_get_fields(pstate, pstate->pfields, pctx, TRUE))
return NULL;
pstate->ilno++;
// We check for comments here rather than within the parser since it's important
// for users to be able to comment out lines containing double-quoted newlines.
if (pstate->comment_string != NULL && pstate->pfields->phead != NULL) {
if (streqn(pstate->pfields->phead->value, pstate->comment_string, pstate->comment_string_length)) {
if (pstate->comment_handling == PASS_COMMENTS) {
int i = 0;
for (
rsllse_t* pe = pstate->pfields->phead;
i < pstate->pfields->length && pe != NULL;
pe = pe->pnext, i++)
{
if (i > 0)
fputs(pstate->ifs, stdout);
fputs(pe->value, stdout);
}
if (pstate->do_auto_line_term) {
fputs(pctx->auto_line_term, stdout);
} else {
fputs(pstate->irs, stdout);
}
}
rslls_reset(pstate->pfields);
continue;
}
}
slls_t* pheader_fields = slls_alloc();
int i = 0;
for (rsllse_t* pe = pstate->pfields->phead; i < pstate->pfields->length && pe != NULL; pe = pe->pnext) {
if (*pe->value == 0) {
fprintf(stderr, "%s: unacceptable empty CSV key at file \"%s\" line %lld.\n",
MLR_GLOBALS.bargv0, pctx->filename, pstate->ilno);
exit(1);
}
// Transfer pointer-free responsibility from the rslls to the
// header fields in the header keeper
slls_append(pheader_fields, pe->value, pe->free_flag);
pe->free_flag = 0;
}
rslls_reset(pstate->pfields);
pstate->pheader_keeper = lhmslv_get(pstate->pheader_keepers, pheader_fields);
if (pstate->pheader_keeper == NULL) {
pstate->pheader_keeper = header_keeper_alloc(NULL, pheader_fields);
lhmslv_put(pstate->pheader_keepers, pheader_fields, pstate->pheader_keeper,
NO_FREE); // freed by header-keeper
} else { // Re-use the header-keeper in the header cache
slls_free(pheader_fields);
}
pstate->expect_header_line_next = FALSE;
break;
}
}
// Ingest the next data line, if expected
while (TRUE) {
int rc = lrec_reader_stdio_csv_get_fields(pstate, pstate->pfields, pctx, FALSE);
pstate->ilno++;
if (rc == FALSE) // EOF
return NULL;
// We check for comments here rather than within the parser since it's important
// for users to be able to comment out lines containing double-quoted newlines.
if (pstate->comment_string != NULL && pstate->pfields->phead != NULL) {
if (streqn(pstate->pfields->phead->value, pstate->comment_string, pstate->comment_string_length)) {
if (pstate->comment_handling == PASS_COMMENTS) {
int i = 0;
for (
rsllse_t* pe = pstate->pfields->phead;
i < pstate->pfields->length && pe != NULL;
pe = pe->pnext, i++)
{
if (i > 0)
fputs(pstate->ifs, stdout);
fputs(pe->value, stdout);
}
if (pstate->do_auto_line_term) {
fputs(pctx->auto_line_term, stdout);
} else {
fputs(pstate->irs, stdout);
}
}
rslls_reset(pstate->pfields);
continue;
}
}
lrec_t* prec = pstate->use_implicit_header
? paste_indices_and_data(pstate, pstate->pfields, pctx)
: paste_header_and_data(pstate, pstate->pfields, pctx);
rslls_reset(pstate->pfields);
return prec;
}
}
// ----------------------------------------------------------------
static void mapper_reorder_free(void* pvstate) {
mapper_reorder_state_t* pstate = (mapper_reorder_state_t*)pvstate;
if (pstate->pfield_name_list != NULL)
slls_free(pstate->pfield_name_list);
}
// ----------------------------------------------------------------
int ap_parse(ap_state_t* pstate, char* verb, int* pargi, int argc, char** argv) {
int argi = *pargi;
int ok = TRUE;
while (argi < argc) {
if (argv[argi][0] != '-') {
break;
}
if (streq(argv[argi], "-h") || streq(argv[argi], "--help")) {
ok = FALSE;
break;
}
ap_flag_def_t* pdef = ap_find(pstate, argv[argi]);
if (pdef == NULL) {
ok = FALSE;
break;
}
if ((argc-argi) < pdef->count) {
fprintf(stderr, "%s %s: option %s requires an argument.\n",
argv[0], verb, argv[argi]);
fprintf(stderr, "\n");
ok = FALSE;
break;
}
if (pdef->type == AP_INT_VALUE_FLAG) {
*(int *)pdef->pval = pdef->intval;
} else if (pdef->type == AP_CHAR_FLAG) {
if (!try_sep_from_arg(argv[argi+1], (char *)pdef->pval)) {
fprintf(stderr, "%s %s: couldn't parse \"%s\" after \"%s\" as character.\n",
argv[0], verb, argv[argi+1], argv[argi]);
fprintf(stderr, "\n");
}
} else if (pdef->type == AP_INT_FLAG) {
if (sscanf(argv[argi+1], "%d", (int *)pdef->pval) != 1) {
fprintf(stderr, "%s %s: couldn't parse \"%s\" after \"%s\" as integer.\n",
argv[0], verb, argv[argi+1], argv[argi]);
fprintf(stderr, "\n");
}
} else if (pdef->type == AP_DOUBLE_FLAG) {
if (!mlr_try_double_from_string(argv[argi+1], (double *)pdef->pval)) {
fprintf(stderr, "%s %s: couldn't parse \"%s\" after \"%s\" as double.\n",
argv[0], verb, argv[argi+1], argv[argi]);
fprintf(stderr, "\n");
}
} else if (pdef->type == AP_STRING_FLAG) {
char** pstring = pdef->pval;
*pstring = argv[argi+1];
pdef->pval = pstring;
} else if (pdef->type == AP_STRING_LIST_FLAG) {
slls_t** pplist = pdef->pval;
if (*pplist != NULL)
slls_free(*pplist);
*pplist = slls_from_line(argv[argi+1], ',', FALSE);
pdef->pval = pplist;
} else {
ok = FALSE;
fprintf(stderr, "argparse.c: coding error: flag-def type %x not recognized.\n", pdef->type);
fprintf(stderr, "\n");
break;
}
argi += pdef->count;
}
*pargi = argi;
return ok;
}