X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Flanguage%2Fdata-io%2Fdata-parser.c;h=359f2df64c83ec7b05507f2fc498bc6ea624d34a;hb=fb5ee794ece0f21f0b52ee41e66a8b79c159bcf2;hp=960505a5ce25f322931f1e8050161f72f0698ede;hpb=1b5c2d8fe129a5f8db76c42250a9c199deccb773;p=pspp diff --git a/src/language/data-io/data-parser.c b/src/language/data-io/data-parser.c index 960505a5ce..359f2df64c 100644 --- a/src/language/data-io/data-parser.c +++ b/src/language/data-io/data-parser.c @@ -29,29 +29,32 @@ #include "data/file-handle-def.h" #include "data/settings.h" #include "language/data-io/data-reader.h" +#include "libpspp/intern.h" #include "libpspp/message.h" #include "libpspp/str.h" -#include "output/tab.h" +#include "output/pivot-table.h" #include "gl/xalloc.h" #include "gettext.h" +#define N_(msgid) msgid #define _(msgid) gettext (msgid) /* Data parser for textual data like that read by DATA LIST. */ struct data_parser { - const struct dictionary *dict; /*Dictionary of destination */ + struct dictionary *dict; /* Dictionary of destination */ enum data_parser_type type; /* Type of data to parse. */ int skip_records; /* Records to skip before first real data. */ struct field *fields; /* Fields to parse. */ - size_t field_cnt; /* Number of fields. */ + size_t n_fields; /* Number of fields. */ size_t field_allocated; /* Number of fields spaced allocated for. */ /* DP_DELIMITED parsers only. */ bool span; /* May cases span multiple records? */ bool empty_line_has_field; /* Does an empty line have an (empty) field? */ + bool warn_missing_fields; /* Should missing fields be considered errors? */ struct substring quotes; /* Characters that can quote separators. */ bool quote_escape; /* Doubled quote acts as escape? */ struct substring soft_seps; /* Two soft separators act like just one. */ @@ -78,7 +81,7 @@ static void set_any_sep (struct data_parser *parser); /* Creates and returns a new data parser. */ struct data_parser * -data_parser_create (const struct dictionary *dict) +data_parser_create (struct dictionary *dict) { struct data_parser *parser = xmalloc (sizeof *parser); @@ -86,12 +89,13 @@ data_parser_create (const struct dictionary *dict) parser->skip_records = 0; parser->fields = NULL; - parser->field_cnt = 0; + parser->n_fields = 0; parser->field_allocated = 0; - parser->dict = dict; + parser->dict = dict_ref (dict); parser->span = true; parser->empty_line_has_field = false; + parser->warn_missing_fields = true; ss_alloc_substring (&parser->quotes, ss_cstr ("\"'")); parser->quote_escape = false; ss_alloc_substring (&parser->soft_seps, ss_cstr (CC_SPACES)); @@ -112,7 +116,8 @@ data_parser_destroy (struct data_parser *parser) { size_t i; - for (i = 0; i < parser->field_cnt; i++) + dict_unref (parser->dict); + for (i = 0; i < parser->n_fields; i++) free (parser->fields[i].name); free (parser->fields); ss_dealloc (&parser->quotes); @@ -135,7 +140,7 @@ data_parser_get_type (const struct data_parser *parser) void data_parser_set_type (struct data_parser *parser, enum data_parser_type type) { - assert (parser->field_cnt == 0); + assert (parser->n_fields == 0); assert (type == DP_FIXED || type == DP_DELIMITED); parser->type = type; } @@ -186,6 +191,21 @@ data_parser_set_empty_line_has_field (struct data_parser *parser, parser->empty_line_has_field = empty_line_has_field; } + +/* If WARN_MISSING_FIELDS is true, configures PARSER to emit a warning + and cause an error condition when a missing field is encountered. + If WARN_MISSING_FIELDS is false, PARSER will silently fill such + fields with the system missing value. + + This setting affects parsing of DP_DELIMITED files only. */ +void +data_parser_set_warn_missing_fields (struct data_parser *parser, + bool warn_missing_fields) +{ + parser->warn_missing_fields = warn_missing_fields; +} + + /* Sets the characters that may be used for quoting field contents to QUOTES. If QUOTES is empty, quoting will be disabled. @@ -271,9 +291,9 @@ add_field (struct data_parser *p, const struct fmt_spec *format, int case_idx, { struct field *field; - if (p->field_cnt == p->field_allocated) + if (p->n_fields == p->field_allocated) p->fields = x2nrealloc (p->fields, &p->field_allocated, sizeof *p->fields); - field = &p->fields[p->field_cnt++]; + field = &p->fields[p->n_fields++]; field->format = *format; field->case_idx = case_idx; field->name = xstrdup (name); @@ -315,8 +335,8 @@ data_parser_add_fixed_field (struct data_parser *parser, int record, int first_column) { assert (parser->type == DP_FIXED); - assert (parser->field_cnt == 0 - || record >= parser->fields[parser->field_cnt - 1].record); + assert (parser->n_fields == 0 + || record >= parser->fields[parser->n_fields - 1].record); if (record > parser->records_per_case) parser->records_per_case = record; add_field (parser, format, case_idx, name, record, first_column); @@ -327,7 +347,7 @@ data_parser_add_fixed_field (struct data_parser *parser, bool data_parser_any_fields (const struct data_parser *parser) { - return parser->field_cnt > 0; + return parser->n_fields > 0; } static void @@ -386,10 +406,11 @@ data_parser_parse (struct data_parser *parser, struct dfm_reader *reader, *FIELD is set to the field content. The caller must not or destroy this constant string. - After parsing the field, sets the current position in the - record to just past the field and any trailing delimiter. - Returns 0 on failure or a 1-based column number indicating the - beginning of the field on success. */ + Sets *FIRST_COLUMN to the 1-based column number of the start of + the extracted field, and *LAST_COLUMN to the end of the extracted + field. + + Returns true on success, false on failure. */ static bool cut_field (const struct data_parser *parser, struct dfm_reader *reader, int *first_column, int *last_column, struct string *tmp, @@ -475,18 +496,22 @@ static void parse_error (const struct dfm_reader *reader, const struct field *field, int first_column, int last_column, char *error) { - struct msg m; - - m.category = MSG_C_DATA; - m.severity = MSG_S_WARNING; - m.file_name = CONST_CAST (char *, dfm_get_file_name (reader)); - m.first_line = dfm_get_line_number (reader); - m.last_line = m.first_line + 1; - m.first_column = first_column; - m.last_column = last_column; - m.text = xasprintf (_("Data for variable %s is not valid as format %s: %s"), - field->name, fmt_name (field->format.type), error); - msg_emit (&m); + int line_number = dfm_get_line_number (reader); + struct msg_location *location = xmalloc (sizeof *location); + *location = (struct msg_location) { + .file_name = intern_new (dfm_get_file_name (reader)), + .start = { .line = line_number, .column = first_column }, + .end = { .line = line_number, .column = last_column - 1 }, + }; + struct msg *m = xmalloc (sizeof *m); + *m = (struct msg) { + .category = MSG_C_DATA, + .severity = MSG_S_WARNING, + .location = location, + .text = xasprintf (_("Data for variable %s is not valid as format %s: %s"), + field->name, fmt_name (field->format.type), error), + }; + msg_emit (m); free (error); } @@ -520,18 +545,20 @@ parse_fixed (const struct data_parser *parser, struct dfm_reader *reader, dfm_expand_tabs (reader); line = dfm_get_record (reader); - for (; f < &parser->fields[parser->field_cnt] && f->record == row; f++) + for (; f < &parser->fields[parser->n_fields] && f->record == row; f++) { struct substring s = ss_substr (line, f->first_column - 1, f->format.w); union value *value = case_data_rw_idx (c, f->case_idx); char *error = data_in (s, input_encoding, f->format.type, + settings_get_fmt_settings (), value, fmt_var_width (&f->format), output_encoding); if (error == NULL) data_in_imply_decimals (s, input_encoding, f->format.type, - f->format.d, value); + f->format.d, settings_get_fmt_settings (), + value); else parse_error (reader, f, f->first_column, f->first_column + f->format.w, error); @@ -554,7 +581,7 @@ parse_delimited_span (const struct data_parser *parser, struct string tmp = DS_EMPTY_INITIALIZER; struct field *f; - for (f = parser->fields; f < &parser->fields[parser->field_cnt]; f++) + for (f = parser->fields; f < &parser->fields[parser->n_fields]; f++) { struct substring s; int first_column, last_column; @@ -578,6 +605,7 @@ parse_delimited_span (const struct data_parser *parser, const char *input_encoding = dfm_reader_get_encoding (reader); error = data_in (s, input_encoding, f->format.type, + settings_get_fmt_settings (), case_data_rw_idx (c, f->case_idx), fmt_var_width (&f->format), output_encoding); if (error != NULL) @@ -602,7 +630,7 @@ parse_delimited_no_span (const struct data_parser *parser, if (dfm_eof (reader)) return false; - end = &parser->fields[parser->field_cnt]; + end = &parser->fields[parser->n_fields]; for (f = parser->fields; f < end; f++) { int first_column, last_column; @@ -610,7 +638,7 @@ parse_delimited_no_span (const struct data_parser *parser, if (!cut_field (parser, reader, &first_column, &last_column, &tmp, &s)) { - if (f < end - 1 && settings_get_undefined ()) + if (f < end - 1 && settings_get_undefined () && parser->warn_missing_fields) msg (DW, _("Missing value(s) for all variables from %s onward. " "These will be filled with the system-missing value " "or blanks, as appropriate."), @@ -623,6 +651,7 @@ parse_delimited_no_span (const struct data_parser *parser, const char *input_encoding = dfm_reader_get_encoding (reader); error = data_in (s, input_encoding, f->format.type, + settings_get_fmt_settings (), case_data_rw_idx (c, f->case_idx), fmt_var_width (&f->format), output_encoding); if (error != NULL) @@ -646,37 +675,48 @@ static void dump_fixed_table (const struct data_parser *parser, const struct file_handle *fh) { - struct tab_table *t; - size_t i; - - t = tab_create (4, parser->field_cnt + 1); - tab_headers (t, 0, 0, 1, 0); - tab_text (t, 0, 0, TAB_CENTER | TAT_TITLE, _("Variable")); - tab_text (t, 1, 0, TAB_CENTER | TAT_TITLE, _("Record")); - tab_text (t, 2, 0, TAB_CENTER | TAT_TITLE, _("Columns")); - tab_text (t, 3, 0, TAB_CENTER | TAT_TITLE, _("Format")); - tab_box (t, TAL_1, TAL_1, TAL_0, TAL_1, 0, 0, 3, parser->field_cnt); - tab_hline (t, TAL_2, 0, 3, 1); - - for (i = 0; i < parser->field_cnt; i++) + /* XXX This should not be preformatted. */ + char *title = xasprintf (ngettext ("Reading %d record from %s.", + "Reading %d records from %s.", + parser->records_per_case), + parser->records_per_case, fh_get_name (fh)); + struct pivot_table *table = pivot_table_create__ ( + pivot_value_new_user_text (title, -1), "Fixed Data Records"); + free (title); + + pivot_dimension_create ( + table, PIVOT_AXIS_COLUMN, N_("Attributes"), + N_("Record"), N_("Columns"), N_("Format")); + + struct pivot_dimension *variables = pivot_dimension_create ( + table, PIVOT_AXIS_ROW, N_("Variable")); + variables->root->show_label = true; + for (size_t i = 0; i < parser->n_fields; i++) { struct field *f = &parser->fields[i]; - char fmt_string[FMT_STRING_LEN_MAX + 1]; - int row = i + 1; - - tab_text (t, 0, row, TAB_LEFT, f->name); - tab_text_format (t, 1, row, 0, "%d", f->record); - tab_text_format (t, 2, row, 0, "%3d-%3d", - f->first_column, f->first_column + f->format.w - 1); - tab_text (t, 3, row, TAB_LEFT | TAB_FIX, - fmt_to_string (&f->format, fmt_string)); + + /* XXX It would be better to have the actual variable here. */ + int variable_idx = pivot_category_create_leaf ( + variables->root, pivot_value_new_user_text (f->name, -1)); + + pivot_table_put2 (table, 0, variable_idx, + pivot_value_new_integer (f->record)); + + int first_column = f->first_column; + int last_column = f->first_column + f->format.w - 1; + char *columns = xasprintf ("%d-%d", first_column, last_column); + pivot_table_put2 (table, 1, variable_idx, + pivot_value_new_user_text (columns, -1)); + free (columns); + + char str[FMT_STRING_LEN_MAX + 1]; + pivot_table_put2 (table, 2, variable_idx, + pivot_value_new_user_text ( + fmt_to_string (&f->format, str), -1)); + } - tab_title (t, ngettext ("Reading %d record from %s.", - "Reading %d records from %s.", - parser->records_per_case), - parser->records_per_case, fh_get_name (fh)); - tab_submit (t); + pivot_table_submit (table); } /* Displays a table giving information on free-format variable parsing @@ -685,30 +725,32 @@ static void dump_delimited_table (const struct data_parser *parser, const struct file_handle *fh) { - struct tab_table *t; - size_t i; - - t = tab_create (2, parser->field_cnt + 1); - tab_headers (t, 0, 0, 1, 0); - tab_text (t, 0, 0, TAB_CENTER | TAT_TITLE, _("Variable")); - tab_text (t, 1, 0, TAB_CENTER | TAT_TITLE, _("Format")); - tab_box (t, TAL_1, TAL_1, TAL_0, TAL_1, 0, 0, 1, parser->field_cnt); - tab_hline (t, TAL_2, 0, 1, 1); - - for (i = 0; i < parser->field_cnt; i++) + struct pivot_table *table = pivot_table_create__ ( + pivot_value_new_text_format (N_("Reading free-form data from %s."), + fh_get_name (fh)), + "Free-Form Data Records"); + + pivot_dimension_create ( + table, PIVOT_AXIS_COLUMN, N_("Attributes"), N_("Format")); + + struct pivot_dimension *variables = pivot_dimension_create ( + table, PIVOT_AXIS_ROW, N_("Variable")); + variables->root->show_label = true; + for (size_t i = 0; i < parser->n_fields; i++) { struct field *f = &parser->fields[i]; - char str[FMT_STRING_LEN_MAX + 1]; - int row = i + 1; - tab_text (t, 0, row, TAB_LEFT, f->name); - tab_text (t, 1, row, TAB_LEFT | TAB_FIX, - fmt_to_string (&f->format, str)); - } + /* XXX It would be better to have the actual variable here. */ + int variable_idx = pivot_category_create_leaf ( + variables->root, pivot_value_new_user_text (f->name, -1)); - tab_title (t, _("Reading free-form data from %s."), fh_get_name (fh)); + char str[FMT_STRING_LEN_MAX + 1]; + pivot_table_put2 (table, 0, variable_idx, + pivot_value_new_user_text ( + fmt_to_string (&f->format, str), -1)); + } - tab_submit (t); + pivot_table_submit (table); } /* Displays a table giving information on how PARSER will read @@ -740,23 +782,35 @@ static const struct casereader_class data_parser_casereader_class; transferred to the dataset. */ void data_parser_make_active_file (struct data_parser *parser, struct dataset *ds, - struct dfm_reader *reader, - struct dictionary *dict) + struct dfm_reader *reader, + struct dictionary *dict, + struct casereader* (*func)(struct casereader *, + const struct dictionary *, + void *), + void *ud) { struct data_parser_casereader *r; - struct casereader *casereader; + struct casereader *casereader0; + struct casereader *casereader1; r = xmalloc (sizeof *r); r->parser = parser; r->reader = reader; r->proto = caseproto_ref (dict_get_proto (dict)); - casereader = casereader_create_sequential (NULL, r->proto, + casereader0 = casereader_create_sequential (NULL, r->proto, CASENUMBER_MAX, &data_parser_casereader_class, r); + + if (func) + casereader1 = func (casereader0, dict, ud); + else + casereader1 = casereader0; + dataset_set_dict (ds, dict); - dataset_set_source (ds, casereader); + dataset_set_source (ds, casereader1); } + static struct ccase * data_parser_casereader_read (struct casereader *reader UNUSED, void *r_) { @@ -772,14 +826,14 @@ data_parser_casereader_read (struct casereader *reader UNUSED, void *r_) } static void -data_parser_casereader_destroy (struct casereader *reader UNUSED, void *r_) +data_parser_casereader_destroy (struct casereader *reader, void *r_) { struct data_parser_casereader *r = r_; if (dfm_reader_error (r->reader)) casereader_force_error (reader); - data_parser_destroy (r->parser); dfm_close_reader (r->reader); caseproto_unref (r->proto); + data_parser_destroy (r->parser); free (r); }