X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Fvar.h;h=6d5cad7938ea2589cb6047b55ac6a3d8b21af14c;hb=97d6c6f6b1922621ca013668eba9a9a9f71d60fe;hp=38f0d74a84c3323d1683be2228a108c192441289;hpb=4944c86a9318bc5b5578ab145a95c116ffd2c9fd;p=pspp diff --git a/src/var.h b/src/var.h index 38f0d74a84..6d5cad7938 100644 --- a/src/var.h +++ b/src/var.h @@ -20,63 +20,11 @@ #if !var_h #define var_h 1 +#include #include "format.h" +#include "t-test.h" +#include "val.h" -/* Values. */ - -/* Definition of the max length of a short string value, generally - eight characters. */ -#define MAX_SHORT_STRING ((SIZEOF_DOUBLE)>=8 ? ((SIZEOF_DOUBLE)+1)/2*2 : 8) -#define MIN_LONG_STRING (MAX_SHORT_STRING+1) - -/* FYI: It is a bad situation if sizeof(flt64) < MAX_SHORT_STRING: - then short string missing values can be truncated in system files - because there's only room for as many characters as can fit in a - flt64. */ -#if MAX_SHORT_STRING > 8 -#error MAX_SHORT_STRING must be less than 8. -#endif - -/* Special values. */ -#define SYSMIS (-DBL_MAX) -#define LOWEST second_lowest_value -#define HIGHEST DBL_MAX - -/* Describes one value, which is either a floating-point number or a - short string. */ -union value - { - /* A numeric value. */ - double f; - - /* A short-string value. */ - unsigned char s[MAX_SHORT_STRING]; - - /* This member is used by data-in.c to return a string result, - since it may need to return a long string. As currently - implemented, it's a pointer to a static internal buffer in - data-in.c. - - Also used by evaluate_expression() to return a string result. - As currently implemented, it's a pointer to a dynamic buffer in - the appropriate expression. - - Also used by the AGGREGATE procedure in handling string - values. */ - unsigned char *c; - - /* Sometimes we insert value's in a hash table. */ - unsigned long hash[SIZEOF_DOUBLE / SIZEOF_LONG]; - }; - -/* Describes one value label. */ -struct value_label - { - union value v; /* The value being labeled. */ - char *s; /* Pointer to malloc()'d label. */ - int ref_count; /* Reference count. */ - }; - /* Frequency tables. */ /* Frequency table entry. */ @@ -99,7 +47,7 @@ struct freq_tab int mode; /* FRQM_GENERAL or FRQM_INTEGER. */ /* General mode. */ - struct avl_tree *tree; /* Undifferentiated data. */ + struct hsh_table *data; /* Undifferentiated data. */ /* Integer mode. */ double *vector; /* Frequencies proper. */ @@ -108,7 +56,7 @@ struct freq_tab double sysmis; /* Sum of weights of SYSMIS values. */ /* All modes. */ - struct freq *valid; /* Valid freqs. */ + struct freq *valid; /* Valid freqs. */ int n_valid; /* Number of total freqs. */ struct freq *missing; /* Missing freqs. */ @@ -118,14 +66,6 @@ struct freq_tab double total_cases; /* Sum of weights of all cases. */ double valid_cases; /* Sum of weights of valid cases. */ }; - -/* A complete set of 3 frequency tables. */ -struct freq_tab_set - { - struct freq_tab miss; /* Includes user-missing values. */ - struct freq_tab no_miss; /* Excludes user-missing values. */ - struct freq_tab sel; /* Identical to either miss or no_miss. */ - }; /* Procedures' private per-variable data. */ @@ -143,6 +83,7 @@ struct crosstab_proc int count; /* max - min. */ }; + /* FREQUENCIES private data. */ enum { @@ -153,7 +94,9 @@ enum struct frequencies_proc { - /* General mode. */ + int used; /* 1=This variable already used. */ + + /* Freqency table. */ struct freq_tab tab; /* Frequencies table to use. */ /* Percentiles. */ @@ -205,28 +148,7 @@ struct descriptives_proc /* GET private data. */ struct get_proc { - int fv, nv; /* First, last, # of values. */ - }; - -/* Sort order. */ -enum - { - SRT_ASCEND, /* A, B, C, ..., X, Y, Z. */ - SRT_DESCEND /* Z, Y, X, ..., C, B, A. */ - }; - -/* SORT CASES private data. */ -struct sort_cases_proc - { - int order; /* SRT_ASCEND or SRT_DESCEND. */ - }; - -/* MODIFY VARS private data. */ -struct modify_vars_proc - { - char new_name[9]; /* Variable's new name. */ - int drop_this_var; /* 0=keep this var, 1=drop this var. */ - struct variable *next; /* Next in linked list. */ + int fv, nv; /* First, # of values. */ }; /* MEANS private data. */ @@ -292,22 +214,17 @@ enum MISSING_COUNT }; -/* A variable's dictionary entry. Note: don't reorder name[] from the - first element; a pointer to `variable' should be a pointer to - member `name'.*/ +/* A variable's dictionary entry. */ struct variable { - /* Required by parse_variables() to be in this order. */ char name[9]; /* As a string. */ int index; /* Index into its dictionary's var[]. */ - int type; /* NUMERIC or ALPHA. */ - int foo; /* Used for temporary storage. */ + int type; /* NUMERIC or ALPHA. */ - /* Also important but parse_variables() doesn't need it. Still, - check before reordering. */ int width; /* Size of string variables in chars. */ int fv, nv; /* Index into `value's, number of values. */ - int left; /* 0=do not LEAVE, 1=LEAVE. */ + unsigned init : 1; /* 1=VFM must init and possibly reinit. */ + unsigned reinit : 1; /* Cases are: 1=reinitialized; 0=left. */ /* Missing values. */ int miss_type; /* One of the MISSING_* constants. */ @@ -318,10 +235,11 @@ struct variable struct fmt_spec write; /* Default format for WRITE. */ /* Labels. */ - struct avl_tree *val_lab; /* Avltree of value_label structures. */ + struct val_labs *val_labs; char *label; /* Variable label. */ /* Per-procedure info. */ + void *aux; struct get_proc get; union { @@ -330,13 +248,35 @@ struct variable struct frequencies_proc frq; struct list_proc lst; struct means_proc mns; - struct sort_cases_proc srt; - struct modify_vars_proc mfv; struct matrix_data_proc mxd; struct match_files_proc mtf; + struct t_test_proc t_t; } p; }; + +int compare_variables (const void *, const void *, void *); +unsigned hash_variable (const void *, void *); + +/* Classes of variables. */ +enum dict_class + { + DC_ORDINARY, /* Ordinary identifier. */ + DC_SYSTEM, /* System variable. */ + DC_SCRATCH /* Scratch variable. */ + }; + +enum dict_class dict_class_from_id (const char *name); +const char *dict_class_to_name (enum dict_class dict_class); + +/* Vector of variables. */ +struct vector + { + int idx; /* Index for dict_get_vector(). */ + char name[9]; /* Name. */ + struct variable **var; /* Vector of variables. */ + int cnt; /* Number of variables. */ + }; /* Cases. */ @@ -346,38 +286,93 @@ struct ccase { union value data[1]; }; + +/* Linked list of cases. */ +struct case_list + { + struct case_list *next; + struct ccase c; + }; /* Dictionary. */ /* Complete dictionary state. */ -struct dictionary - { - struct variable **var; /* Variable descriptions. */ - struct avl_tree *var_by_name; /* Variables arranged by name. */ - int nvar; /* Number of variables. */ - - int N; /* Current case limit (N command). */ - int nval; /* Number of value structures per case. */ - - int n_splits; /* Number of SPLIT FILE variables. */ - struct variable **splits; /* List of SPLIT FILE vars. */ - - char *label; /* File label. */ - - int n_documents; /* Number of lines of documents. */ - char *documents; /* Documents; 80*n_documents bytes in size. */ - - int weight_index; /* `value' index of $WEIGHT, or -1 if none. - Call update_weighting() before using! */ - char weight_var[9]; /* Name of WEIGHT variable. */ - - char filter_var[9]; /* Name of FILTER variable. */ - /* Do not make another field the last field! or see - temporary.c:restore_dictionary() before doing so! */ - }; +struct dictionary; + +struct dictionary *dict_create (void); +struct dictionary *dict_clone (const struct dictionary *); +void dict_clear (struct dictionary *); +void dict_destroy (struct dictionary *); + +size_t dict_get_var_cnt (const struct dictionary *); +struct variable *dict_get_var (const struct dictionary *, size_t idx); +void dict_get_vars (const struct dictionary *, + struct variable ***vars, size_t *cnt, + unsigned exclude_classes); + +struct variable *dict_create_var (struct dictionary *, const char *, + int width); +struct variable *dict_create_var_assert (struct dictionary *, const char *, + int width); +struct variable *dict_clone_var (struct dictionary *, const struct variable *, + const char *); +void dict_rename_var (struct dictionary *, struct variable *, const char *); + +struct variable *dict_lookup_var (const struct dictionary *, const char *); +struct variable *dict_lookup_var_assert (const struct dictionary *, + const char *); +int dict_contains_var (const struct dictionary *, const struct variable *); +void dict_delete_var (struct dictionary *, struct variable *); +void dict_delete_vars (struct dictionary *, + struct variable *const *, size_t count); +void dict_reorder_vars (struct dictionary *, + struct variable *const *, size_t count); +int dict_rename_vars (struct dictionary *, + struct variable **, char **new_names, + size_t count, char **err_name); + +struct variable *dict_get_weight (const struct dictionary *); +double dict_get_case_weight (const struct dictionary *, const struct ccase *); +void dict_set_weight (struct dictionary *, struct variable *); + +struct variable *dict_get_filter (const struct dictionary *); +void dict_set_filter (struct dictionary *, struct variable *); + +int dict_get_case_limit (const struct dictionary *); +void dict_set_case_limit (struct dictionary *, int); + +int dict_get_next_value_idx (const struct dictionary *); +size_t dict_get_case_size (const struct dictionary *); + +void dict_compact_values (struct dictionary *); +size_t dict_get_compacted_value_cnt (const struct dictionary *); +int *dict_get_compacted_idx_to_fv (const struct dictionary *); + +struct variable *const *dict_get_split_vars (const struct dictionary *); +size_t dict_get_split_cnt (const struct dictionary *); +void dict_set_split_vars (struct dictionary *, + struct variable *const *, size_t cnt); + +const char *dict_get_label (const struct dictionary *); +void dict_set_label (struct dictionary *, const char *); + +const char *dict_get_documents (const struct dictionary *); +void dict_set_documents (struct dictionary *, const char *); + +int dict_create_vector (struct dictionary *, + const char *name, + struct variable **, size_t cnt); +const struct vector *dict_get_vector (const struct dictionary *, + size_t idx); +size_t dict_get_vector_cnt (const struct dictionary *); +const struct vector *dict_lookup_vector (const struct dictionary *, + const char *name); +void dict_clear_vectors (struct dictionary *); + +void discard_variables (void); /* This is the active file dictionary. */ -extern struct dictionary default_dict; +extern struct dictionary *default_dict; /* Transformation state. */ @@ -410,26 +405,8 @@ void cancel_temporary (void); /* Functions. */ -int is_varname (const char *); -int is_dict_varname (const struct dictionary *, const char *); - -/* Flags for passing to fill_all_vars(). */ -enum - { - FV_NONE = 0, /* No flags. */ - FV_NO_SYSTEM = 001, /* Don't include system variables. */ - FV_NO_SCRATCH = 002 /* Don't include scratch variables. */ - }; - -void fill_all_vars (struct variable ***, int *, int flags); - -int val_lab_cmp (const void *, const void *, void *); -char *get_val_lab (const struct variable *, union value, int); -void free_val_lab (void *, void *); -void free_value_label (struct value_label *); -struct avl_tree *copy_value_labels (struct avl_tree *); - void dump_split_vars (const struct ccase *); +typedef int (* is_missing_func )(const union value *, const struct variable *); int is_num_user_missing (double, const struct variable *); int is_str_user_missing (const unsigned char[], const struct variable *); @@ -437,62 +414,19 @@ int is_missing (const union value *, const struct variable *); int is_system_missing (const union value *, const struct variable *); int is_user_missing (const union value *, const struct variable *); void copy_missing_values (struct variable *dest, const struct variable *src); - -int cmp_variable (const void *, const void *, void *); - -#if GLOBAL_DEBUGGING -struct variable *force_create_variable (struct dictionary *, const char *name, - int type, int width); -struct variable *force_dup_variable (struct dictionary *, - const struct variable *src, - const char *name); -#else -#define force_create_variable(A, B, C, D) \ - create_variable (A, B, C, D) -#define force_dup_variable(A, B, C) \ - dup_variable (A, B, C) -#endif - -struct variable *create_variable (struct dictionary *, const char *name, - int type, int width); -void delete_variable (struct dictionary *, struct variable *v); -struct variable *find_variable (const char *name); -struct variable *find_dict_variable (const struct dictionary *, - const char *name); -void init_variable (struct dictionary *, struct variable *, const char *name, - int type, int width); -void replace_variable (struct variable *, const char *name, - int type, int width); -void clear_variable (struct dictionary *, struct variable *); -void rename_variable (struct dictionary *, struct variable *v, - const char *new_name); -void discard_variables (void); -void clear_default_dict (void); -void copy_variable (struct variable *dest, const struct variable *src); -struct variable *dup_variable (struct dictionary *dict, - const struct variable *src, const char *name); - -struct variable *update_weighting (struct dictionary *); -void stop_weighting (struct dictionary *); - -struct dictionary *save_dictionary (void); -void restore_dictionary (struct dictionary *); -void free_dictionary (struct dictionary *); -struct dictionary *new_dictionary (int copy); /* Transformations. */ +struct trns_header; +typedef int trns_proc_func (struct trns_header *, struct ccase *, int); +typedef void trns_free_func (struct trns_header *); + /* Header for all transformations. */ struct trns_header { - /* Index into t_trns[]. */ - int index; - - /* Transformation proc. */ - int (*proc) (struct trns_header *, struct ccase *); - - /* Garbage collector proc. */ - void (*free) (struct trns_header *); + int index; /* Index into t_trns[]. */ + trns_proc_func *proc; /* Transformation proc. */ + trns_free_func *free; /* Garbage collector proc. */ }; /* Array of transformations */ @@ -508,28 +442,38 @@ extern int f_trns; void add_transformation (struct trns_header *trns); void cancel_transformations (void); +struct var_set; + +struct var_set *var_set_create_from_dict (struct dictionary *d); +struct var_set *var_set_create_from_array (struct variable **var, size_t); + +size_t var_set_get_cnt (struct var_set *vs); +struct variable *var_set_get_var (struct var_set *vs, size_t idx); +struct variable *var_set_lookup_var (struct var_set *vs, const char *name); +void var_set_destroy (struct var_set *vs); + /* Variable parsers. */ -/* Only parse_variables() supports options other than PV_APPEND, - PV_SINGLE. */ enum { PV_NONE = 0, /* No options. */ - PV_SINGLE = 0001, /* Restrict to a single varname or TO use. */ + PV_SINGLE = 0001, /* Restrict to a single name or TO use. */ PV_DUPLICATE = 0002, /* Don't merge duplicates. */ PV_APPEND = 0004, /* Append to existing list. */ PV_NO_DUPLICATE = 0010, /* Error on duplicates. */ PV_NUMERIC = 0020, /* Vars must be numeric. */ PV_STRING = 0040, /* Vars must be string. */ PV_SAME_TYPE = 00100, /* All vars must be the same type. */ - PV_NO_SCRATCH = 00200 /* Disallow scratch variables. */ + PV_NO_SCRATCH = 00200, /* Disallow scratch variables. */ }; struct variable *parse_variable (void); struct variable *parse_dict_variable (struct dictionary *); -int parse_variables (struct dictionary *dict, struct variable ***v, - int *nv, int pv_opts); -int parse_DATA_LIST_vars (char ***names, int *nnames, int pv_opts); -int parse_mixed_vars (char ***names, int *nnames, int pv_opts); +int parse_variables (struct dictionary *, struct variable ***, int *, + int opts); +int parse_var_set_vars (struct var_set *, struct variable ***, int *, + int opts); +int parse_DATA_LIST_vars (char ***names, int *cnt, int opts); +int parse_mixed_vars (char ***names, int *cnt, int opts); #endif /* !var_h */