X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;ds=sidebyside;f=src%2Flanguage%2Fdata-io%2Fget-data.c;h=fa033f22e86c2de2b126d5d50a826f3bb4132c13;hb=4785da39b9cb392c0a6a2c38a8abac199a72d615;hp=7e75b413ea49cf82fcb87a81daea8ba570e4b905;hpb=81579d9e9f994fb2908f50af41c3eb033d216e58;p=pspp diff --git a/src/language/data-io/get-data.c b/src/language/data-io/get-data.c index 7e75b413ea..fa033f22e8 100644 --- a/src/language/data-io/get-data.c +++ b/src/language/data-io/get-data.c @@ -1,5 +1,5 @@ /* PSPP - a program for statistical analysis. - Copyright (C) 2007, 2008, 2009, 2010, 2011 Free Software Foundation, Inc. + Copyright (C) 2007, 2008, 2009, 2010, 2011, 2012, 2013 Free Software Foundation, Inc. This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by @@ -18,10 +18,14 @@ #include +#include + +#include "data/dataset.h" #include "data/dictionary.h" #include "data/format.h" #include "data/gnumeric-reader.h" -#include "data/procedure.h" +#include "data/ods-reader.h" +#include "data/spreadsheet-reader.h" #include "data/psql-reader.h" #include "data/settings.h" #include "language/command.h" @@ -31,6 +35,8 @@ #include "language/data-io/placement-parser.h" #include "language/lexer/format-parser.h" #include "language/lexer/lexer.h" +#include "libpspp/cast.h" +#include "libpspp/i18n.h" #include "libpspp/message.h" #include "gl/xalloc.h" @@ -39,13 +45,18 @@ #define _(msgid) gettext (msgid) #define N_(msgid) (msgid) -static int parse_get_gnm (struct lexer *lexer, struct dataset *); +static bool parse_spreadsheet (struct lexer *lexer, struct spreadsheet_read_info *sri, + struct spreadsheet_read_options *opts); + +static void destroy_spreadsheet_read_info (struct spreadsheet_read_info *, struct spreadsheet_read_options *); + static int parse_get_txt (struct lexer *lexer, struct dataset *); static int parse_get_psql (struct lexer *lexer, struct dataset *); int cmd_get_data (struct lexer *lexer, struct dataset *ds) { + char *tok = NULL; lex_force_match (lexer, T_SLASH); if (!lex_force_match_id (lexer, "TYPE")) @@ -53,14 +64,47 @@ cmd_get_data (struct lexer *lexer, struct dataset *ds) lex_force_match (lexer, T_EQUALS); - if (lex_match_id (lexer, "GNM")) - return parse_get_gnm (lexer, ds); - else if (lex_match_id (lexer, "TXT")) - return parse_get_txt (lexer, ds); + tok = strdup (lex_tokcstr (lexer)); + if (lex_match_id (lexer, "TXT")) + { + free (tok); + return parse_get_txt (lexer, ds); + } else if (lex_match_id (lexer, "PSQL")) - return parse_get_psql (lexer, ds); + { + free (tok); + return parse_get_psql (lexer, ds); + } + else if (lex_match_id (lexer, "GNM") || + lex_match_id (lexer, "ODS")) + { + struct casereader *reader = NULL; + struct dictionary *dict = NULL; + struct spreadsheet_read_info sri; + struct spreadsheet_read_options opts; + if (!parse_spreadsheet (lexer, &sri, &opts)) + goto error; + + if ( 0 == strncasecmp (tok, "GNM", 3)) + reader = gnumeric_open_reader (&sri, &opts, &dict); + else if (0 == strncasecmp (tok, "ODS", 3)) + reader = ods_open_reader (&sri, &opts, &dict); + + if (reader) + { + dataset_set_dict (ds, dict); + dataset_set_source (ds, reader); + destroy_spreadsheet_read_info (&sri, &opts); + free (tok); + return CMD_SUCCESS; + } + destroy_spreadsheet_read_info (&sri, &opts); + } + else + msg (SE, _("Unsupported TYPE %s."), tok); - msg (SE, _("Unsupported TYPE %s."), lex_tokcstr (lexer)); + error: + free (tok); return CMD_FAILURE; } @@ -121,7 +165,10 @@ parse_get_psql (struct lexer *lexer, struct dataset *ds) struct casereader *reader = psql_open_reader (&psql, &dict); if ( reader ) - proc_set_active_file (ds, reader, dict); + { + dataset_set_dict (ds, dict); + dataset_set_source (ds, reader); + } } ds_destroy (&psql.sql); @@ -137,10 +184,16 @@ parse_get_psql (struct lexer *lexer, struct dataset *ds) return CMD_FAILURE; } -static int -parse_get_gnm (struct lexer *lexer, struct dataset *ds) +static bool +parse_spreadsheet (struct lexer *lexer, struct spreadsheet_read_info *sri, + struct spreadsheet_read_options *opts) { - struct gnumeric_read_info gri = {NULL, NULL, NULL, 1, true, -1}; + opts->sheet_index = 1; + opts->sheet_name = NULL; + opts->cell_range = NULL; + sri->read_names = true; + sri->asw = -1; + sri->file_name = NULL; lex_force_match (lexer, T_SLASH); @@ -152,7 +205,7 @@ parse_get_gnm (struct lexer *lexer, struct dataset *ds) if (!lex_force_string (lexer)) goto error; - gri.file_name = ss_xstrdup (lex_tokss (lexer)); + sri->file_name = utf8_to_filename (lex_tokcstr (lexer)); lex_get (lexer); @@ -161,7 +214,7 @@ parse_get_gnm (struct lexer *lexer, struct dataset *ds) if ( lex_match_id (lexer, "ASSUMEDSTRWIDTH")) { lex_match (lexer, T_EQUALS); - gri.asw = lex_integer (lexer); + sri->asw = lex_integer (lexer); lex_get (lexer); } else if (lex_match_id (lexer, "SHEET")) @@ -172,18 +225,27 @@ parse_get_gnm (struct lexer *lexer, struct dataset *ds) if ( ! lex_force_string (lexer) ) goto error; - gri.sheet_name = ss_xstrdup (lex_tokss (lexer)); - gri.sheet_index = -1; + opts->sheet_name = ss_xstrdup (lex_tokss (lexer)); + opts->sheet_index = -1; lex_get (lexer); } else if (lex_match_id (lexer, "INDEX")) { - gri.sheet_index = lex_integer (lexer); + sri->sheet_index = lex_integer (lexer); + if (sri->sheet_index <= 0) + { + msg (SE, _("The sheet index must be greater than or equal to 1")); + goto error; + } lex_get (lexer); } else - goto error; + { + msg (SE, _("%s must be followed by either \"%s\" or \"%s\"."), + "/SHEET", "NAME", "INDEX"); + goto error; + } } else if (lex_match_id (lexer, "CELLRANGE")) { @@ -191,18 +253,22 @@ parse_get_gnm (struct lexer *lexer, struct dataset *ds) if (lex_match_id (lexer, "FULL")) { - gri.cell_range = NULL; + opts->cell_range = NULL; } else if (lex_match_id (lexer, "RANGE")) { if ( ! lex_force_string (lexer) ) goto error; - gri.cell_range = ss_xstrdup (lex_tokss (lexer)); + opts->cell_range = ss_xstrdup (lex_tokss (lexer)); lex_get (lexer); } else - goto error; + { + msg (SE, _("%s must be followed by either \"%s\" or \"%s\"."), + "/CELLRANGE", "FULL", "RANGE"); + goto error; + } } else if (lex_match_id (lexer, "READNAMES")) { @@ -210,14 +276,18 @@ parse_get_gnm (struct lexer *lexer, struct dataset *ds) if ( lex_match_id (lexer, "ON")) { - gri.read_names = true; + sri->read_names = true; } else if (lex_match_id (lexer, "OFF")) { - gri.read_names = false; + sri->read_names = false; } else - goto error; + { + msg (SE, _("%s must be followed by either \"%s\" or \"%s\"."), + "/READNAMES", "ON", "OFF"); + goto error; + } } else { @@ -226,27 +296,14 @@ parse_get_gnm (struct lexer *lexer, struct dataset *ds) } } - { - struct dictionary *dict = NULL; - struct casereader *reader = gnumeric_open_reader (&gri, &dict); - - if ( reader ) - proc_set_active_file (ds, reader, dict); - } - - free (gri.file_name); - free (gri.sheet_name); - free (gri.cell_range); - return CMD_SUCCESS; + return true; error: - - free (gri.file_name); - free (gri.sheet_name); - free (gri.cell_range); - return CMD_FAILURE; + destroy_spreadsheet_read_info (sri, opts); + return false; } + static bool set_type (struct data_parser *parser, const char *subcommand, enum data_parser_type type, bool *has_type) @@ -272,9 +329,10 @@ static int parse_get_txt (struct lexer *lexer, struct dataset *ds) { struct data_parser *parser = NULL; - struct dictionary *dict = dict_create (); + struct dictionary *dict = dict_create (get_default_encoding ()); struct file_handle *fh = NULL; struct dfm_reader *reader = NULL; + char *encoding = NULL; char *name = NULL; int record; @@ -286,7 +344,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) if (!lex_force_match_id (lexer, "FILE")) goto error; lex_force_match (lexer, T_EQUALS); - fh = fh_parse (lexer, FH_REF_FILE | FH_REF_INLINE); + fh = fh_parse (lexer, FH_REF_FILE | FH_REF_INLINE, NULL); if (fh == NULL) goto error; @@ -302,7 +360,18 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) if (!lex_force_match (lexer, T_SLASH)) goto error; - if (lex_match_id (lexer, "ARRANGEMENT")) + if (lex_match_id (lexer, "ENCODING")) + { + lex_match (lexer, T_EQUALS); + if (!lex_force_string (lexer)) + goto error; + + free (encoding); + encoding = ss_xstrdup (lex_tokss (lexer)); + + lex_get (lexer); + } + else if (lex_match_id (lexer, "ARRANGEMENT")) { bool ok; @@ -314,7 +383,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) DP_DELIMITED, &has_type); else { - lex_error (lexer, _("expecting %s or %s"), "FIXED", "DELIMITED"); + lex_error_expecting (lexer, "FIXED", "DELIMITED", NULL_SENTINEL); goto error; } if (!ok) @@ -352,7 +421,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) } else { - lex_error (lexer, _("expecting %s or %s"), "LINE", "VARIABLES"); + lex_error_expecting (lexer, "LINE", "VARIABLES", NULL_SENTINEL); goto error; } } @@ -418,6 +487,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) if (!lex_force_string (lexer)) goto error; + /* XXX should support multibyte UTF-8 characters */ s = lex_tokss (lexer); if (ss_match_string (&s, ss_cstr ("\\t"))) ds_put_cstr (&hard_seps, "\t"); @@ -443,6 +513,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) if (!lex_force_string (lexer)) goto error; + /* XXX should support multibyte UTF-8 characters */ if (settings_get_syntax () == COMPATIBLE && ss_length (lex_tokss (lexer)) != 1) { @@ -461,7 +532,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) break; else { - lex_error (lexer, _("expecting %s"), "VARIABLES"); + lex_error_expecting (lexer, "VARIABLES", NULL_SENTINEL); goto error; } } @@ -500,7 +571,8 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) lex_get (lexer); } - if (!lex_force_id (lexer)) + if (!lex_force_id (lexer) + || !dict_id_is_valid (dict, lex_tokcstr (lexer), true)) goto error; name = xstrdup (lex_tokcstr (lexer)); lex_get (lexer); @@ -510,19 +582,46 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) if (!parse_format_specifier (lexer, &input) || !fmt_check_input (&input)) goto error; + + output = fmt_for_output_from_input (&input); } else { + char fmt_type_name[FMT_TYPE_LEN_MAX + 1]; + enum fmt_type fmt_type; + int w, d; + if (!parse_column_range (lexer, 0, &fc, &lc, NULL)) goto error; - if (!parse_format_specifier_name (lexer, &input.type)) + + /* Accept a format (e.g. F8.2) or just a type name (e.g. DOLLAR). */ + if (!parse_abstract_format_specifier (lexer, fmt_type_name, &w, &d)) goto error; + if (!fmt_from_name (fmt_type_name, &fmt_type)) + { + msg (SE, _("Unknown format type `%s'."), fmt_type_name); + goto error; + } + + /* Compose input format. */ + input.type = fmt_type; input.w = lc - fc + 1; input.d = 0; if (!fmt_check_input (&input)) goto error; + + /* Compose output format. */ + if (w != 0) + { + output.type = fmt_type; + output.w = w; + output.d = d; + if (!fmt_check_output (&output)) + goto error; + } + else + output = fmt_for_output_from_input (&input); } - output = fmt_for_output_from_input (&input); v = dict_create_var (dict, name, fmt_var_width (&input)); if (v == NULL) @@ -544,12 +643,13 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) } while (lex_token (lexer) != T_ENDCMD); - reader = dfm_open_reader (fh, lexer); + reader = dfm_open_reader (fh, lexer, encoding); if (reader == NULL) goto error; data_parser_make_active_file (parser, ds, reader, dict); fh_unref (fh); + free (encoding); return CMD_SUCCESS; error: @@ -557,5 +657,16 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) dict_destroy (dict); fh_unref (fh); free (name); + free (encoding); return CMD_CASCADING_FAILURE; } + + +static void +destroy_spreadsheet_read_info (struct spreadsheet_read_info *sri, + struct spreadsheet_read_options *opts) +{ + free (opts->sheet_name); + free (opts->cell_range); + free (sri->file_name); +}