X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Fdata%2Fvariable.c;h=121a57517cd4fecc4ce0cbe1cc6b25b9609092e6;hb=e9cbda0449c8c50c2295e14737b875cc34751d55;hp=2ceeb0d0e9610a6396d58246db942ed8c0302ac1;hpb=81579d9e9f994fb2908f50af41c3eb033d216e58;p=pspp diff --git a/src/data/variable.c b/src/data/variable.c index 2ceeb0d0e9..121a57517c 100644 --- a/src/data/variable.c +++ b/src/data/variable.c @@ -31,6 +31,7 @@ #include "libpspp/assertion.h" #include "libpspp/compiler.h" #include "libpspp/hash-functions.h" +#include "libpspp/i18n.h" #include "libpspp/message.h" #include "libpspp/misc.h" #include "libpspp/str.h" @@ -132,7 +133,7 @@ var_clone (const struct variable *old_var) var_set_print_format (new_var, var_get_print_format (old_var)); var_set_write_format (new_var, var_get_write_format (old_var)); var_set_value_labels (new_var, var_get_value_labels (old_var)); - var_set_label (new_var, var_get_label (old_var)); + var_set_label (new_var, var_get_label (old_var), false); var_set_measure (new_var, var_get_measure (old_var)); var_set_display_width (new_var, var_get_display_width (old_var)); var_set_alignment (new_var, var_get_alignment (old_var)); @@ -163,109 +164,27 @@ var_destroy (struct variable *v) /* Variable names. */ -/* Return variable V's name. */ +/* Return variable V's name, as a UTF-8 encoded string. */ const char * var_get_name (const struct variable *v) { return v->name; } -/* Sets V's name to NAME. +/* Sets V's name to NAME, a UTF-8 encoded string. Do not use this function for a variable in a dictionary. Use dict_rename_var instead. */ void var_set_name (struct variable *v, const char *name) { assert (!var_has_vardict (v)); - assert (var_is_plausible_name (name, false)); + assert (id_is_plausible (name, false)); free (v->name); v->name = xstrdup (name); dict_var_changed (v); } -/* Returns true if NAME is an acceptable name for a variable, - false otherwise. If ISSUE_ERROR is true, issues an - explanatory error message on failure. */ -bool -var_is_valid_name (const char *name, bool issue_error) -{ - bool plausible; - size_t length, i; - - /* Note that strlen returns number of BYTES, not the number of - CHARACTERS */ - length = strlen (name); - - plausible = var_is_plausible_name(name, issue_error); - - if ( ! plausible ) - return false; - - - if (!lex_is_id1 (name[0])) - { - if (issue_error) - msg (SE, _("Character `%c' (in %s) may not appear " - "as the first character in a variable name."), - name[0], name); - return false; - } - - - for (i = 0; i < length; i++) - { - if (!lex_is_idn (name[i])) - { - if (issue_error) - msg (SE, _("Character `%c' (in %s) may not appear in " - "a variable name."), - name[i], name); - return false; - } - } - - return true; -} - -/* Returns true if NAME is an plausible name for a variable, - false otherwise. If ISSUE_ERROR is true, issues an - explanatory error message on failure. - This function makes no use of LC_CTYPE. -*/ -bool -var_is_plausible_name (const char *name, bool issue_error) -{ - size_t length; - - /* Note that strlen returns number of BYTES, not the number of - CHARACTERS */ - length = strlen (name); - if (length < 1) - { - if (issue_error) - msg (SE, _("Variable name cannot be empty string.")); - return false; - } - else if (length > VAR_NAME_LEN) - { - if (issue_error) - msg (SE, _("Variable name %s exceeds %d-character limit."), - name, (int) VAR_NAME_LEN); - return false; - } - - if (lex_id_to_token (ss_cstr (name)) != T_ID) - { - if (issue_error) - msg (SE, _("`%s' may not be used as a variable name because it " - "is a reserved word."), name); - return false; - } - - return true; -} - /* Returns VAR's dictionary class. */ enum dict_class var_get_dict_class (const struct variable *var) @@ -508,9 +427,11 @@ alloc_value_labels (struct variable *v) v->val_labs = val_labs_create (v->width); } -/* Attempts to add a value label with the given VALUE and LABEL - to V. Returns true if successful, false otherwise (probably - due to an existing label). */ +/* Attempts to add a value label with the given VALUE and UTF-8 encoded LABEL + to V. Returns true if successful, false otherwise (probably due to an + existing label). + + In LABEL, the two-byte sequence "\\n" is interpreted as a new-line. */ bool var_add_value_label (struct variable *v, const union value *value, const char *label) @@ -519,9 +440,10 @@ var_add_value_label (struct variable *v, return val_labs_add (v->val_labs, value, label); } -/* Adds or replaces a value label with the given VALUE and LABEL +/* Adds or replaces a value label with the given VALUE and UTF-8 encoded LABEL to V. -*/ + + In LABEL, the two-byte sequence "\\n" is interpreted as a new-line. */ void var_replace_value_label (struct variable *v, const union value *value, const char *label) @@ -537,8 +459,8 @@ var_clear_value_labels (struct variable *v) var_set_value_labels (v, NULL); } -/* Returns the label associated with VALUE for variable V, - or a null pointer if none. */ +/* Returns the label associated with VALUE for variable V, as a UTF-8 string in + a format suitable for output, or a null pointer if none. */ const char * var_lookup_value_label (const struct variable *v, const union value *value) { @@ -644,33 +566,60 @@ var_get_label (const struct variable *v) return v->label; } -/* Sets V's variable label to LABEL, stripping off leading and - trailing white space and truncating to 255 characters. - If LABEL is a null pointer or if LABEL is an empty string - (after stripping white space), then V's variable label (if - any) is removed. */ -void -var_set_label (struct variable *v, const char *label) +/* Sets V's variable label to UTF-8 encoded string LABEL, stripping off leading + and trailing white space. If LABEL is a null pointer or if LABEL is an + empty string (after stripping white space), then V's variable label (if any) + is removed. + + Variable labels are limited to 255 bytes in V's encoding (as returned by + var_get_encoding()). If LABEL fits within this limit, this function returns + true. Otherwise, the variable label is set to a truncated value, this + function returns false and, if ISSUE_WARNING is true, issues a warning. */ +bool +var_set_label (struct variable *v, const char *label, bool issue_warning) { + bool truncated = false; + free (v->label); v->label = NULL; if (label != NULL) { + const char *dict_encoding = var_get_encoding (v); struct substring s = ss_cstr (label); + size_t trunc_len; + + if (dict_encoding != NULL) + { + enum { MAX_LABEL_LEN = 255 }; + + trunc_len = utf8_encoding_trunc_len (label, dict_encoding, + MAX_LABEL_LEN); + if (ss_length (s) > trunc_len) + { + if (issue_warning) + msg (SW, _("Truncating variable label for variable `%s' to %d " + "bytes."), var_get_name (v), MAX_LABEL_LEN); + ss_truncate (&s, trunc_len); + truncated = true; + } + } + ss_trim (&s, ss_cstr (CC_SPACES)); - ss_truncate (&s, 255); if (!ss_is_empty (s)) v->label = ss_xstrdup (s); } + dict_var_changed (v); + + return truncated; } /* Removes any variable label from V. */ void var_clear_label (struct variable *v) { - var_set_label (v, NULL); + var_set_label (v, NULL, false); } /* Returns true if V has a variable V, @@ -832,14 +781,16 @@ var_get_short_name (const struct variable *var, size_t idx) return idx < var->short_name_cnt ? var->short_names[idx] : NULL; } -/* Sets VAR's short name with the given IDX to SHORT_NAME, - truncating it to SHORT_NAME_LEN characters and converting it - to uppercase in the process. Specifying a null pointer for - SHORT_NAME clears the specified short name. */ +/* Sets VAR's short name with the given IDX to the UTF-8 string SHORT_NAME. + The caller must already have checked that, in the dictionary encoding, + SHORT_NAME is no more than SHORT_NAME_LEN bytes long. The new short name + will be converted to uppercase. + + Specifying a null pointer for SHORT_NAME clears the specified short name. */ void var_set_short_name (struct variable *var, size_t idx, const char *short_name) { - assert (short_name == NULL || var_is_plausible_name (short_name, false)); + assert (short_name == NULL || id_is_plausible (short_name, false)); /* Clear old short name numbered IDX, if any. */ if (idx < var->short_name_cnt) @@ -861,7 +812,7 @@ var_set_short_name (struct variable *var, size_t idx, const char *short_name) for (i = old_cnt; i < var->short_name_cnt; i++) var->short_names[i] = NULL; } - var->short_names[idx] = xstrndup (short_name, MAX_SHORT_STRING); + var->short_names[idx] = xstrdup (short_name); str_uppercase (var->short_names[idx]); }