X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Flanguage%2Fdata-io%2Fget-data.c;h=2bf419659d18e52ae99a91455c05dcd8a5b698df;hb=1b5c2d8fe129a5f8db76c42250a9c199deccb773;hp=9b878c553a92bf1a16e37f5985506e025db924c6;hpb=32ee0e0402d6d56674f53a47d879ec5c07dabe09;p=pspp diff --git a/src/language/data-io/get-data.c b/src/language/data-io/get-data.c index 9b878c553a..2bf419659d 100644 --- a/src/language/data-io/get-data.c +++ b/src/language/data-io/get-data.c @@ -1,5 +1,6 @@ /* 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, 2015, 2016 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 @@ -35,6 +36,7 @@ #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" @@ -44,8 +46,32 @@ #define _(msgid) gettext (msgid) #define N_(msgid) (msgid) -static struct spreadsheet_read_info *parse_spreadsheet (struct lexer *lexer); -static void destroy_spreadsheet_read_info (struct spreadsheet_read_info *); + +#ifdef ODF_READ_SUPPORT +static const bool odf_read_support = true; +#else +static const bool odf_read_support = false; +struct spreadsheet *ods_probe (const char *filename, bool report_errors){} +struct casereader * ods_make_reader (struct spreadsheet *spreadsheet, + const struct spreadsheet_read_options *opts){} +void ods_unref (struct spreadsheet *r){} +#endif + +#ifdef GNM_READ_SUPPORT +static const bool gnm_read_support = true; +#else +static const bool gnm_read_support = false; +struct spreadsheet *gnumeric_probe (const char *filename, bool report_errors){} +struct casereader * gnumeric_make_reader (struct spreadsheet *spreadsheet, + const struct spreadsheet_read_options *opts){} +void gnumeric_unref (struct spreadsheet *r){} + +#endif + +static bool parse_spreadsheet (struct lexer *lexer, char **filename, + struct spreadsheet_read_options *opts); + +static void destroy_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 *); @@ -54,50 +80,78 @@ int cmd_get_data (struct lexer *lexer, struct dataset *ds) { char *tok = NULL; + struct spreadsheet_read_options opts; + + opts.sheet_name = NULL; + opts.sheet_index = -1; + opts.cell_range = NULL; + opts.read_names = false; + opts.asw = -1; + lex_force_match (lexer, T_SLASH); if (!lex_force_match_id (lexer, "TYPE")) - return CMD_FAILURE; + goto error; lex_force_match (lexer, T_EQUALS); 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")) { + free (tok); return parse_get_psql (lexer, ds); } else if (lex_match_id (lexer, "GNM") || lex_match_id (lexer, "ODS")) { + char *filename = NULL; struct casereader *reader = NULL; struct dictionary *dict = NULL; - struct spreadsheet_read_info *sri = parse_spreadsheet (lexer); - if (NULL == sri) + + if (!parse_spreadsheet (lexer, &filename, &opts)) goto error; - if ( 0 == strncasecmp (tok, "GNM", 3)) - reader = gnumeric_open_reader (sri, &dict); - else if (0 == strncasecmp (tok, "ODS", 3)) - reader = ods_open_reader (sri, &dict); + if ( gnm_read_support && 0 == strncasecmp (tok, "GNM", 3)) + { + struct spreadsheet *spreadsheet = gnumeric_probe (filename, true); + if (spreadsheet == NULL) + goto error; + reader = gnumeric_make_reader (spreadsheet, &opts); + dict = spreadsheet->dict; + gnumeric_unref (spreadsheet); + } + else if ( odf_read_support && 0 == strncasecmp (tok, "ODS", 3)) + { + struct spreadsheet *spreadsheet = ods_probe (filename, true); + if (spreadsheet == NULL) + goto error; + reader = ods_make_reader (spreadsheet, &opts); + dict = spreadsheet->dict; + ods_unref (spreadsheet); + } + + free (filename); if (reader) { dataset_set_dict (ds, dict); dataset_set_source (ds, reader); - destroy_spreadsheet_read_info (sri); free (tok); + destroy_spreadsheet_read_info (&opts); return CMD_SUCCESS; } - destroy_spreadsheet_read_info (sri); } else msg (SE, _("Unsupported TYPE %s."), tok); + error: + destroy_spreadsheet_read_info (&opts); free (tok); return CMD_FAILURE; } @@ -178,13 +232,15 @@ parse_get_psql (struct lexer *lexer, struct dataset *ds) return CMD_FAILURE; } -static struct spreadsheet_read_info * -parse_spreadsheet (struct lexer *lexer) +static bool +parse_spreadsheet (struct lexer *lexer, char **filename, + struct spreadsheet_read_options *opts) { - struct spreadsheet_read_info *sri = xzalloc (sizeof *sri); - sri->sheet_index = 1; - sri->read_names = true; - sri->asw = -1; + opts->sheet_index = 1; + opts->sheet_name = NULL; + opts->cell_range = NULL; + opts->read_names = true; + opts->asw = -1; lex_force_match (lexer, T_SLASH); @@ -196,7 +252,7 @@ parse_spreadsheet (struct lexer *lexer) if (!lex_force_string (lexer)) goto error; - sri->file_name = utf8_to_filename (lex_tokcstr (lexer)); + *filename = utf8_to_filename (lex_tokcstr (lexer)); lex_get (lexer); @@ -205,7 +261,7 @@ parse_spreadsheet (struct lexer *lexer) if ( lex_match_id (lexer, "ASSUMEDSTRWIDTH")) { lex_match (lexer, T_EQUALS); - sri->asw = lex_integer (lexer); + opts->asw = lex_integer (lexer); lex_get (lexer); } else if (lex_match_id (lexer, "SHEET")) @@ -216,18 +272,27 @@ parse_spreadsheet (struct lexer *lexer) if ( ! lex_force_string (lexer) ) goto error; - sri->sheet_name = ss_xstrdup (lex_tokss (lexer)); - sri->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")) { - sri->sheet_index = lex_integer (lexer); + opts->sheet_index = lex_integer (lexer); + if (opts->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")) { @@ -235,18 +300,22 @@ parse_spreadsheet (struct lexer *lexer) if (lex_match_id (lexer, "FULL")) { - sri->cell_range = NULL; + opts->cell_range = NULL; } else if (lex_match_id (lexer, "RANGE")) { if ( ! lex_force_string (lexer) ) goto error; - sri->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")) { @@ -254,14 +323,18 @@ parse_spreadsheet (struct lexer *lexer) if ( lex_match_id (lexer, "ON")) { - sri->read_names = true; + opts->read_names = true; } else if (lex_match_id (lexer, "OFF")) { - sri->read_names = false; + opts->read_names = false; } else - goto error; + { + msg (SE, _("%s must be followed by either \"%s\" or \"%s\"."), + "/READNAMES", "ON", "OFF"); + goto error; + } } else { @@ -270,11 +343,10 @@ parse_spreadsheet (struct lexer *lexer) } } - return sri; + return true; error: - destroy_spreadsheet_read_info (sri); - return NULL; + return false; } @@ -306,6 +378,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) 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; @@ -326,6 +399,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) data_parser_set_type (parser, DP_DELIMITED); data_parser_set_span (parser, false); data_parser_set_quotes (parser, ss_empty ()); + data_parser_set_quote_escape (parser, true); data_parser_set_empty_line_has_field (parser, true); for (;;) @@ -333,7 +407,18 @@ 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; @@ -345,7 +430,7 @@ 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) @@ -358,7 +443,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) goto error; if (lex_integer (lexer) < 1) { - msg (SE, _("Value of FIRSTCASE must be 1 or greater.")); + msg (SE, _("Value of %s must be 1 or greater."), "FIRSTCASE"); goto error; } data_parser_set_skip (parser, lex_integer (lexer) - 1); @@ -383,7 +468,7 @@ 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; } } @@ -396,7 +481,7 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) goto error; if (lex_integer (lexer) < 1) { - msg (SE, _("Value of FIXCASE must be at least 1.")); + msg (SE, _("Value of %s must be 1 or greater."), "FIXCASE"); goto error; } data_parser_set_records (parser, lex_integer (lexer)); @@ -407,33 +492,22 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) lex_match (lexer, T_EQUALS); if (lex_match (lexer, T_ALL)) { - data_parser_set_case_limit (parser, -1); - data_parser_set_case_percent (parser, 100); + /* Nothing to do. */ } else if (lex_match_id (lexer, "FIRST")) { if (!lex_force_int (lexer)) goto error; - if (lex_integer (lexer) < 1) - { - msg (SE, _("Value of FIRST must be at least 1.")); - goto error; - } - data_parser_set_case_limit (parser, lex_integer (lexer)); lex_get (lexer); } else if (lex_match_id (lexer, "PERCENT")) { if (!lex_force_int (lexer)) goto error; - if (lex_integer (lexer) < 1 || lex_integer (lexer) > 100) - { - msg (SE, _("Value of PERCENT must be between 1 and 100.")); - goto error; - } - data_parser_set_case_percent (parser, lex_integer (lexer)); lex_get (lexer); } + msg (SW, _("Ignoring obsolete IMPORTCASES subcommand. (N OF CASES " + "or SAMPLE may be used to substitute.)")); } else if (lex_match_id_n (lexer, "DELIMITERS", 4)) { @@ -487,14 +561,11 @@ parse_get_txt (struct lexer *lexer, struct dataset *ds) data_parser_set_quotes (parser, lex_tokss (lexer)); lex_get (lexer); } - else if (settings_get_syntax () == ENHANCED - && lex_match_id (lexer, "ESCAPE")) - data_parser_set_quote_escape (parser, true); else if (lex_match_id (lexer, "VARIABLES")) break; else { - lex_error (lexer, _("expecting %s"), "VARIABLES"); + lex_error_expecting (lexer, "VARIABLES", NULL_SENTINEL); goto error; } } @@ -605,12 +676,13 @@ 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: @@ -618,18 +690,14 @@ 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) +destroy_spreadsheet_read_info (struct spreadsheet_read_options *opts) { - if ( NULL == sri) - return; - - free (sri->sheet_name); - free (sri->cell_range); - free (sri->file_name); - free (sri); + free (opts->cell_range); + free (opts->sheet_name); }