X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Flanguage%2Fxforms%2Frecode.c;h=275408de90e7d8a3dbd3fde47628d3ac540d3e27;hb=e1335ea67114822554e0d51bd5e0b49401b17dc4;hp=5cfad0ed766ec2376429c32b3dc1bbb83e212271;hpb=a258e53c63a08b0ec48aea8f03808eb651729424;p=pspp diff --git a/src/language/xforms/recode.c b/src/language/xforms/recode.c index 5cfad0ed76..275408de90 100644 --- a/src/language/xforms/recode.c +++ b/src/language/xforms/recode.c @@ -1,5 +1,5 @@ /* PSPP - a program for statistical analysis. - Copyright (C) 1997-9, 2000, 2009, 2010 Free Software Foundation, Inc. + Copyright (C) 1997-9, 2000, 2009, 2010, 2011 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 @@ -22,9 +22,9 @@ #include "data/case.h" #include "data/data-in.h" +#include "data/dataset.h" #include "data/dictionary.h" #include "data/format.h" -#include "data/procedure.h" #include "data/transformations.h" #include "data/variable.h" #include "language/command.h" @@ -34,6 +34,7 @@ #include "libpspp/assertion.h" #include "libpspp/cast.h" #include "libpspp/compiler.h" +#include "libpspp/i18n.h" #include "libpspp/message.h" #include "libpspp/pool.h" #include "libpspp/str.h" @@ -84,8 +85,6 @@ struct recode_trns { struct pool *pool; - - /* Variable types, for convenience. */ enum val_type src_type; /* src_vars[*] type. */ enum val_type dst_type; /* dst_vars[*] type. */ @@ -105,25 +104,28 @@ struct recode_trns }; static bool parse_src_vars (struct lexer *, struct recode_trns *, const struct dictionary *dict); -static bool parse_mappings (struct lexer *, struct recode_trns *); +static bool parse_mappings (struct lexer *, struct recode_trns *, + const char *dict_encoding); static bool parse_dst_vars (struct lexer *, struct recode_trns *, const struct dictionary *dict); static void add_mapping (struct recode_trns *, size_t *map_allocated, const struct map_in *); static bool parse_map_in (struct lexer *lexer, struct map_in *, struct pool *, - enum val_type src_type, size_t max_src_width); + enum val_type src_type, size_t max_src_width, + const char *dict_encoding); static void set_map_in_generic (struct map_in *, enum map_in_type); static void set_map_in_num (struct map_in *, enum map_in_type, double, double); static void set_map_in_str (struct map_in *, struct pool *, - const struct string *, size_t width); + struct substring, size_t width, + const char *dict_encoding); static bool parse_map_out (struct lexer *lexer, struct pool *, struct map_out *); static void set_map_out_num (struct map_out *, double); static void set_map_out_str (struct map_out *, struct pool *, - const struct string *); + struct substring); -static void enlarge_dst_widths (struct recode_trns *); +static bool enlarge_dst_widths (struct recode_trns *); static void create_dst_vars (struct recode_trns *, struct dictionary *); static trns_proc_func recode_trns_proc; @@ -137,15 +139,16 @@ cmd_recode (struct lexer *lexer, struct dataset *ds) { do { + struct dictionary *dict = dataset_dict (ds); struct recode_trns *trns = pool_create_container (struct recode_trns, pool); /* Parse source variable names, then input to output mappings, then destintation variable names. */ - if (!parse_src_vars (lexer, trns, dataset_dict (ds) ) - || !parse_mappings (lexer, trns) - || !parse_dst_vars (lexer, trns, dataset_dict (ds))) + if (!parse_src_vars (lexer, trns, dict) + || !parse_mappings (lexer, trns, dict_get_encoding (dict)) + || !parse_dst_vars (lexer, trns, dict)) { recode_trns_free (trns); return CMD_FAILURE; @@ -154,22 +157,28 @@ cmd_recode (struct lexer *lexer, struct dataset *ds) /* Ensure that all the output strings are at least as wide as the widest destination variable. */ if (trns->dst_type == VAL_STRING) - enlarge_dst_widths (trns); + { + if ( ! enlarge_dst_widths (trns)) + { + recode_trns_free (trns); + return CMD_FAILURE; + } + } /* Create destination variables, if needed. This must be the final step; otherwise we'd have to delete destination variables on failure. */ - trns->dst_dict = dataset_dict (ds); + trns->dst_dict = dict; if (trns->src_vars != trns->dst_vars) - create_dst_vars (trns, dataset_dict (ds)); + create_dst_vars (trns, dict); /* Done. */ add_transformation (ds, recode_trns_proc, recode_trns_free, trns); } - while (lex_match (lexer, '/')); + while (lex_match (lexer, T_SLASH)); - return lex_end_of_command (lexer); + return CMD_SUCCESS; } /* Parses a set of variables to recode into TRNS->src_vars and @@ -191,7 +200,8 @@ parse_src_vars (struct lexer *lexer, into TRNS->mappings and TRNS->map_cnt. Sets TRNS->dst_type. Returns true if successful, false on parse error. */ static bool -parse_mappings (struct lexer *lexer, struct recode_trns *trns) +parse_mappings (struct lexer *lexer, struct recode_trns *trns, + const char *dict_encoding) { size_t map_allocated; bool have_dst_type; @@ -211,7 +221,7 @@ parse_mappings (struct lexer *lexer, struct recode_trns *trns) trns->map_cnt = 0; map_allocated = 0; have_dst_type = false; - if (!lex_force_match (lexer, '(')) + if (!lex_force_match (lexer, T_LPAREN)) return false; do { @@ -231,12 +241,13 @@ parse_mappings (struct lexer *lexer, struct recode_trns *trns) struct map_in in; if (!parse_map_in (lexer, &in, trns->pool, - trns->src_type, trns->max_src_width)) + trns->src_type, trns->max_src_width, + dict_encoding)) return false; add_mapping (trns, &map_allocated, &in); - lex_match (lexer, ','); + lex_match (lexer, T_COMMA); } - while (!lex_match (lexer, '=')); + while (!lex_match (lexer, T_EQUALS)); if (!parse_map_out (lexer, trns->pool, &out)) return false; @@ -276,10 +287,10 @@ parse_mappings (struct lexer *lexer, struct recode_trns *trns) trns->dst_type = dst_type; have_dst_type = true; - if (!lex_force_match (lexer, ')')) + if (!lex_force_match (lexer, T_RPAREN)) return false; } - while (lex_match (lexer, '(')); + while (lex_match (lexer, T_LPAREN)); return true; } @@ -291,7 +302,8 @@ parse_mappings (struct lexer *lexer, struct recode_trns *trns) false on parse error. */ static bool parse_map_in (struct lexer *lexer, struct map_in *in, struct pool *pool, - enum val_type src_type, size_t max_src_width) + enum val_type src_type, size_t max_src_width, + const char *dict_encoding) { if (lex_match_id (lexer, "ELSE")) @@ -318,12 +330,13 @@ parse_map_in (struct lexer *lexer, struct map_in *in, struct pool *pool, return false; else { - set_map_in_str (in, pool, lex_tokstr (lexer), max_src_width); + set_map_in_str (in, pool, lex_tokss (lexer), max_src_width, + dict_encoding); lex_get (lexer); if (lex_token (lexer) == T_ID - && lex_id_match (ss_cstr ("THRU"), ss_cstr (lex_tokid (lexer)))) + && lex_id_match (ss_cstr ("THRU"), lex_tokss (lexer))) { - msg (SE, _("THRU is not allowed with string variables.")); + msg (SE, _("%s is not allowed with string variables."), "THRU"); return false; } } @@ -370,13 +383,16 @@ set_map_in_num (struct map_in *in, enum map_in_type type, double x, double y) right to WIDTH characters long. */ static void set_map_in_str (struct map_in *in, struct pool *pool, - const struct string *string, size_t width) + struct substring string, size_t width, + const char *dict_encoding) { + char *s = recode_string (dict_encoding, "UTF-8", + ss_data (string), ss_length (string)); in->type = MAP_SINGLE; value_init_pool (pool, &in->x, width); value_copy_buf_rpad (&in->x, width, - CHAR_CAST_BUG (uint8_t *, ds_data (string)), - ds_length (string), ' '); + CHAR_CAST (uint8_t *, s), strlen (s), ' '); + free (s); } /* Parses a mapping output value into OUT, allocating memory from @@ -393,7 +409,7 @@ parse_map_out (struct lexer *lexer, struct pool *pool, struct map_out *out) set_map_out_num (out, SYSMIS); else if (lex_is_string (lexer)) { - set_map_out_str (out, pool, lex_tokstr (lexer)); + set_map_out_str (out, pool, lex_tokss (lexer)); lex_get (lexer); } else if (lex_match_id (lexer, "COPY")) @@ -421,10 +437,10 @@ set_map_out_num (struct map_out *out, double value) /* Sets OUT as a string mapping output with the given VALUE. */ static void set_map_out_str (struct map_out *out, struct pool *pool, - const struct string *value) + const struct substring value) { - const char *string = ds_data (value); - size_t length = ds_length (value); + const char *string = ss_data (value); + size_t length = ss_length (value); if (length == 0) { @@ -518,26 +534,46 @@ parse_dst_vars (struct lexer *lexer, struct recode_trns *trns, /* Ensures that all the output values in TRNS are as wide as the widest destination variable. */ -static void +static bool enlarge_dst_widths (struct recode_trns *trns) { size_t i; - + const struct variable *narrow_var = NULL; + int min_dst_width = INT_MAX; trns->max_dst_width = 0; + for (i = 0; i < trns->var_cnt; i++) { const struct variable *v = trns->dst_vars[i]; if (var_get_width (v) > trns->max_dst_width) trns->max_dst_width = var_get_width (v); + + if (var_get_width (v) < min_dst_width) + { + min_dst_width = var_get_width (v); + narrow_var = v; + } } for (i = 0; i < trns->map_cnt; i++) { struct map_out *out = &trns->mappings[i].out; if (!out->copy_input) - value_resize_pool (trns->pool, &out->value, - out->width, trns->max_dst_width); + { + if (out->width > min_dst_width) + { + msg (ME, + _("Cannot recode because the variable %s would require a width of %d bytes or greater, but it has a width of only %d bytes."), + var_get_name (narrow_var), out->width, min_dst_width); + return false; + } + + value_resize_pool (trns->pool, &out->value, + out->width, trns->max_dst_width); + } } + + return true; } /* Creates destination variables that don't already exist. */ @@ -632,7 +668,7 @@ find_src_string (struct recode_trns *trns, const uint8_t *value, char *error; error = data_in (ss_buffer (CHAR_CAST_BUG (char *, value), width), - LEGACY_NATIVE, FMT_F, &uv, 0, encoding); + C_ENCODING, FMT_F, &uv, 0, encoding); match = error == NULL; free (error);