X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Flanguage%2Fexpressions%2Fparse.c;h=0e5f6abbb6323514c3b349f845a962caaacdbbe3;hb=93ec42221da8b677420bf11435e0d24d0503601b;hp=c1035611f5befcd232319bcc7f54edcf41fc6b72;hpb=691c25e36fd1ee722dd35419d6110e3876b99f9c;p=pspp diff --git a/src/language/expressions/parse.c b/src/language/expressions/parse.c index c1035611f5..0e5f6abbb6 100644 --- a/src/language/expressions/parse.c +++ b/src/language/expressions/parse.c @@ -1,5 +1,5 @@ /* PSPP - a program for statistical analysis. - Copyright (C) 1997-9, 2000, 2006, 2010 Free Software Foundation, Inc. + Copyright (C) 1997-9, 2000, 2006, 2010, 2011, 2012, 2014 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 @@ -23,22 +23,24 @@ #include #include -#include "helpers.h" -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include "xalloc.h" +#include "data/case.h" +#include "data/dictionary.h" +#include "data/settings.h" +#include "data/variable.h" +#include "language/expressions/helpers.h" +#include "language/lexer/format-parser.h" +#include "language/lexer/lexer.h" +#include "language/lexer/variable-parser.h" +#include "libpspp/array.h" +#include "libpspp/assertion.h" +#include "libpspp/i18n.h" +#include "libpspp/message.h" +#include "libpspp/misc.h" +#include "libpspp/pool.h" +#include "libpspp/str.h" + +#include "gl/c-strcase.h" +#include "gl/xalloc.h" /* Declarations. */ @@ -261,8 +263,9 @@ type_check (struct expression *e, atom_type_name (actual_type)); return false; } - if (actual_type == OP_number && expected_type == OP_boolean) - *n = expr_allocate_unary (e, OP_NUM_TO_BOOLEAN, *n); + if (actual_type == OP_number && expected_type == EXPR_BOOLEAN) + *n = expr_allocate_binary (e, OP_NUM_TO_BOOLEAN, *n, + expr_allocate_string (e, ss_empty ())); break; case EXPR_STRING: @@ -499,16 +502,14 @@ match_operator (struct lexer *lexer, const struct operator ops[], size_t op_cnt, const struct operator *op; for (op = ops; op < ops + op_cnt; op++) - { - if (op->token == T_DASH) - lex_negative_to_dash (lexer); - if (lex_match (lexer, op->token)) - { - if (operator != NULL) - *operator = op; - return true; - } - } + if (lex_token (lexer) == op->token) + { + if (op->token != T_NEG_NUM) + lex_get (lexer); + if (operator != NULL) + *operator = op; + return true; + } if (operator != NULL) *operator = NULL; return false; @@ -710,6 +711,7 @@ parse_add (struct lexer *lexer, struct expression *e) { { T_PLUS, OP_ADD, "addition (`+')" }, { T_DASH, OP_SUB, "subtraction (`-')" }, + { T_NEG_NUM, OP_ADD, "subtraction (`-')" }, }; return parse_binary_operators (lexer, e, parse_mul (lexer, e), @@ -752,8 +754,21 @@ parse_exp (struct lexer *lexer, struct expression *e) "That is, `a**b**c' equals `(a**b)**c', not as `a**(b**c)'. " "To disable this warning, insert parentheses."); - return parse_binary_operators (lexer, e, parse_primary (lexer, e), &op, 1, - parse_primary, chain_warning); + union any_node *lhs, *node; + bool negative = false; + + if (lex_token (lexer) == T_NEG_NUM) + { + lhs = expr_allocate_number (e, -lex_tokval (lexer)); + negative = true; + lex_get (lexer); + } + else + lhs = parse_primary (lexer, e); + + node = parse_binary_operators (lexer, e, lhs, &op, 1, + parse_primary, chain_warning); + return negative ? expr_allocate_unary (e, OP_NEG, node) : node; } /* Parses system variables. */ @@ -773,12 +788,14 @@ parse_sysvar (struct lexer *lexer, struct expression *e) time_t last_proc_time = time_of_last_procedure (e->ds); struct tm *time; char temp_buf[10]; + struct substring s; time = localtime (&last_proc_time); sprintf (temp_buf, "%02d %s %02d", abs (time->tm_mday) % 100, months[abs (time->tm_mon) % 12], abs (time->tm_year) % 100); - return expr_allocate_string_buffer (e, temp_buf, strlen (temp_buf)); + ss_alloc_substring (&s, ss_cstr (temp_buf)); + return expr_allocate_string (e, s); } else if (lex_match_id (lexer, "$TRUE")) return expr_allocate_boolean (e, 1.0); @@ -812,7 +829,7 @@ parse_sysvar (struct lexer *lexer, struct expression *e) return expr_allocate_number (e, settings_get_viewwidth ()); else { - msg (SE, _("Unknown system variable %s."), lex_tokid (lexer)); + msg (SE, _("Unknown system variable %s."), lex_tokcstr (lexer)); return NULL; } } @@ -824,22 +841,22 @@ parse_primary (struct lexer *lexer, struct expression *e) switch (lex_token (lexer)) { case T_ID: - if (lex_look_ahead (lexer) == T_LPAREN) + if (lex_next_token (lexer, 1) == T_LPAREN) { /* An identifier followed by a left parenthesis may be a vector element reference. If not, it's a function call. */ - if (e->ds != NULL && dict_lookup_vector (dataset_dict (e->ds), lex_tokid (lexer)) != NULL) + if (e->ds != NULL && dict_lookup_vector (dataset_dict (e->ds), lex_tokcstr (lexer)) != NULL) return parse_vector_element (lexer, e); else return parse_function (lexer, e); } - else if (lex_tokid (lexer)[0] == '$') + else if (lex_tokcstr (lexer)[0] == '$') { /* $ at the beginning indicates a system variable. */ return parse_sysvar (lexer, e); } - else if (e->ds != NULL && dict_lookup_var (dataset_dict (e->ds), lex_tokid (lexer))) + else if (e->ds != NULL && dict_lookup_var (dataset_dict (e->ds), lex_tokcstr (lexer))) { /* It looks like a user variable. (It could be a format specifier, but we'll assume @@ -860,7 +877,7 @@ parse_primary (struct lexer *lexer, struct expression *e) return expr_allocate_format (e, &fmt); /* All attempts failed. */ - msg (SE, _("Unknown identifier %s."), lex_tokid (lexer)); + msg (SE, _("Unknown identifier %s."), lex_tokcstr (lexer)); return NULL; } break; @@ -875,8 +892,17 @@ parse_primary (struct lexer *lexer, struct expression *e) case T_STRING: { - union any_node *node = expr_allocate_string_buffer ( - e, ds_cstr (lex_tokstr (lexer) ), ds_length (lex_tokstr (lexer) )); + const char *dict_encoding; + union any_node *node; + char *s; + + dict_encoding = (e->ds != NULL + ? dict_get_encoding (dataset_dict (e->ds)) + : "UTF-8"); + s = recode_string_pool (dict_encoding, "UTF-8", lex_tokcstr (lexer), + ss_length (lex_tokss (lexer)), e->expr_pool); + node = expr_allocate_string (e, ss_cstr (s)); + lex_get (lexer); return node; } @@ -906,7 +932,7 @@ parse_vector_element (struct lexer *lexer, struct expression *e) /* Find vector, skip token. The caller must already have verified that the current token is the name of a vector. */ - vector = dict_lookup_vector (dataset_dict (e->ds), lex_tokid (lexer)); + vector = dict_lookup_vector (dataset_dict (e->ds), lex_tokcstr (lexer)); assert (vector != NULL); lex_get (lexer); @@ -981,7 +1007,7 @@ compare_names (const char *test, const char *name, bool abbrev_ok) static int compare_strings (const char *test, const char *name, bool abbrev_ok UNUSED) { - return strcasecmp (test, name); + return c_strcasecmp (test, name); } static bool @@ -1021,7 +1047,7 @@ lookup_function (const char *name, } static int -extract_min_valid (char *s) +extract_min_valid (const char *s) { char *p = strrchr (s, '.'); if (p == NULL @@ -1197,11 +1223,11 @@ parse_function (struct lexer *lexer, struct expression *e) union any_node *n; - ds_init_string (&func_name, lex_tokstr (lexer)); - min_valid = extract_min_valid (ds_cstr (lex_tokstr (lexer))); - if (!lookup_function (ds_cstr (lex_tokstr (lexer)), &first, &last)) + ds_init_substring (&func_name, lex_tokss (lexer)); + min_valid = extract_min_valid (lex_tokcstr (lexer)); + if (!lookup_function (lex_tokcstr (lexer), &first, &last)) { - msg (SE, _("No function or vector named %s."), ds_cstr (lex_tokstr (lexer))); + msg (SE, _("No function or vector named %s."), lex_tokcstr (lexer)); ds_destroy (&func_name); return NULL; } @@ -1219,7 +1245,7 @@ parse_function (struct lexer *lexer, struct expression *e) for (;;) { if (lex_token (lexer) == T_ID - && toupper (lex_look_ahead (lexer)) == T_ID) + && lex_next_token (lexer, 1) == T_TO) { const struct variable **vars; size_t var_cnt; @@ -1244,8 +1270,7 @@ parse_function (struct lexer *lexer, struct expression *e) break; else if (!lex_match (lexer, T_COMMA)) { - lex_error (lexer, _("expecting `,' or `)' invoking %s function"), - first->name); + lex_error_expecting (lexer, "`,'", "`)'", NULL_SENTINEL); goto fail; } } @@ -1267,13 +1292,14 @@ parse_function (struct lexer *lexer, struct expression *e) msg (SW, _("%s is a PSPP extension."), f->prototype); if (f->flags & OPF_UNIMPLEMENTED) { - msg (SE, _("%s is not yet implemented."), f->prototype); + msg (SE, _("%s is not available in this version of PSPP."), + f->prototype); goto fail; } if ((f->flags & OPF_PERM_ONLY) && proc_in_temporary_transformations (e->ds)) { - msg (SE, _("%s may not appear after TEMPORARY."), f->prototype); + msg (SE, _("%s may not appear after %s."), f->prototype, "TEMPORARY"); goto fail; } @@ -1461,18 +1487,6 @@ expr_allocate_vector (struct expression *e, const struct vector *vector) return n; } -union any_node * -expr_allocate_string_buffer (struct expression *e, - const char *string, size_t length) -{ - union any_node *n = pool_alloc (e->expr_pool, sizeof n->string); - n->type = OP_string; - if (length > MAX_STRING) - length = MAX_STRING; - n->string.s = copy_string (e, string, length); - return n; -} - union any_node * expr_allocate_string (struct expression *e, struct substring s) {