Files
jhendersonHDF 4883ee1ad8 Improve performance of h5trav interfaces for links to objects (#6400)
For objects with multiple hard links, use hash table to map between
object tokens and pathnames during traversal to avoid linear scan over
all previous objects for each hard link seen

Use separate hash table for h5trav "table" interface to map between
object tokens and an index into the table of visited objects. This
facilitates quick lookups of objects when adding hard link name aliases
for h5repack processing
2026-07-08 15:53:10 -05:00

206 lines
8.3 KiB
C

/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* Copyright by The HDF Group. *
* All rights reserved. *
* *
* This file is part of HDF5. The full HDF5 copyright notice, including *
* terms governing use, modification, and redistribution, is contained in *
* the LICENSE file, which can be found at the root of the source code *
* distribution tree, or in https://www.hdfgroup.org/licenses. *
* If you do not have access to either file, you may request a copy from *
* help@hdfgroup.org. *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
#ifndef H5TRAV_H
#define H5TRAV_H
#include "hdf5.h"
/*-------------------------------------------------------------------------
* public enum to specify type of an object
* the TYPE can be:
* H5TRAV_TYPE_UNKNOWN = -1,
* H5TRAV_TYPE_GROUP, Object is a group
* H5TRAV_TYPE_DATASET, Object is a dataset
* H5TRAV_TYPE_TYPE, Object is a named datatype
* H5TRAV_TYPE_LINK, Object is a symbolic link
* H5TRAV_TYPE_UDLINK, Object is a user-defined link
*-------------------------------------------------------------------------
*/
typedef enum {
H5TRAV_TYPE_UNKNOWN = -1, /* Unknown object type */
H5TRAV_TYPE_GROUP, /* Object is a group */
H5TRAV_TYPE_DATASET, /* Object is a dataset */
H5TRAV_TYPE_NAMED_DATATYPE, /* Object is a named datatype */
H5TRAV_TYPE_LINK, /* Object is a symbolic link */
H5TRAV_TYPE_UDLINK /* Object is a user-defined link */
} h5trav_type_t;
/*-------------------------------------------------------------------------
* public struct to store name and type of an object
*-------------------------------------------------------------------------
*/
/* Struct to keep track of symbolic link targets visited.
* Functions: symlink_visit_add() and symlink_is_visited()
*/
typedef struct symlink_trav_path_t {
H5L_type_t type;
char *file;
char *path;
} symlink_trav_path_t;
typedef struct symlink_trav_t {
size_t nalloc;
size_t nused;
symlink_trav_path_t *objs;
bool dangle_link;
} symlink_trav_t;
typedef struct trav_path_t {
char *path;
h5trav_type_t type;
H5O_token_t obj_token; /* object token */
unsigned long fileno; /* File number that object is located in */
} trav_path_t;
typedef struct trav_info_t {
size_t nalloc;
size_t nused;
const char *fname;
hid_t fid; /* File ID */
trav_path_t *paths;
symlink_trav_t symlink_visited; /* already visited symbolic links */
void *opts; /* optional data passing */
} trav_info_t;
/*-------------------------------------------------------------------------
* struct to store info about visited objects during traversal
*-------------------------------------------------------------------------
*/
typedef struct trav_seen_t {
H5O_token_t token;
char *path;
} trav_seen_t;
/*-------------------------------------------------------------------------
* keep record of hard link information
*-------------------------------------------------------------------------
*/
typedef struct trav_link_t {
char *new_name;
} trav_link_t;
/*-------------------------------------------------------------------------
* struct to store basic info needed for the h5trav table traversal algorithm
*-------------------------------------------------------------------------
*/
typedef struct trav_obj_t {
H5O_token_t obj_token; /* object token */
unsigned flags[2]; /* h5diff.object is present or not in both files*/
bool is_same_trgobj; /* same target object? no need to compare */
char *name; /* name */
h5trav_type_t type; /* type of object */
trav_link_t *links; /* array of possible link names */
size_t sizelinks; /* size of links array */
size_t nlinks; /* number of links */
} trav_obj_t;
/*-------------------------------------------------------------------------
* private struct that stores all objects
*-------------------------------------------------------------------------
*/
typedef struct trav_table_t {
hid_t fid;
size_t obj_token_size;
size_t size;
size_t nobjs;
trav_obj_t *objs;
/* Private data for this trav_table_t */
void *priv_data;
} trav_table_t;
/* Typedefs for visiting objects */
typedef herr_t (*h5trav_obj_func_t)(const char *path_name, const H5O_info2_t *oinfo, bool already_visited,
const trav_seen_t *visited_obj_info, void *udata);
typedef herr_t (*h5trav_lnk_func_t)(const char *path_name, const H5L_info2_t *linfo, void *udata);
/*-------------------------------------------------------------------------
* public functions
*-------------------------------------------------------------------------
*/
#ifdef __cplusplus
extern "C" {
#endif
/*-------------------------------------------------------------------------
* "h5trav general" public functions
*-------------------------------------------------------------------------
*/
H5TOOLS_DLL void h5trav_set_index(H5_index_t print_index_by, H5_iter_order_t print_index_order);
H5TOOLS_DLL int h5trav_visit(hid_t file_id, const char *grp_name, bool visit_start, bool recurse,
h5trav_obj_func_t visit_obj, h5trav_lnk_func_t visit_lnk, void *udata,
unsigned fields);
H5TOOLS_DLL herr_t symlink_visit_add(symlink_trav_t *visited, H5L_type_t type, const char *file,
const char *path);
H5TOOLS_DLL bool symlink_is_visited(symlink_trav_t *visited, H5L_type_t type, const char *file,
const char *path);
/*-------------------------------------------------------------------------
* "h5trav info" public functions
*-------------------------------------------------------------------------
*/
H5TOOLS_DLL int h5trav_getinfo(hid_t file_id, trav_info_t *info);
H5TOOLS_DLL ssize_t h5trav_getindex(const trav_info_t *info, const char *obj);
H5TOOLS_DLL int trav_info_visit_obj(const char *path, const H5O_info2_t *oinfo, bool already_visited,
const trav_seen_t *visited_obj_info, void *udata);
H5TOOLS_DLL int trav_info_visit_lnk(const char *path, const H5L_info2_t *linfo, void *udata);
/*-------------------------------------------------------------------------
* "h5trav table" public functions
*-------------------------------------------------------------------------
*/
H5TOOLS_DLL int h5trav_gettable(hid_t fid, trav_table_t *travt);
H5TOOLS_DLL int h5trav_getindext(const char *obj, const trav_table_t *travt);
/*-------------------------------------------------------------------------
* "h5trav print" public functions
*-------------------------------------------------------------------------
*/
H5TOOLS_DLL int h5trav_print(hid_t fid);
H5TOOLS_DLL void h5trav_set_verbose(int print_verbose);
#ifdef __cplusplus
}
#endif
/*-------------------------------------------------------------------------
* info private functions
*-------------------------------------------------------------------------
*/
H5TOOLS_DLL void trav_info_init(const char *filename, hid_t fileid, trav_info_t **info);
H5TOOLS_DLL void trav_info_free(trav_info_t *info);
H5TOOLS_DLL void trav_info_add(trav_info_t *info, const char *path, h5trav_type_t obj_type);
H5TOOLS_DLL void trav_fileinfo_add(trav_info_t *info, hid_t loc_id);
/*-------------------------------------------------------------------------
* table private functions
*-------------------------------------------------------------------------
*/
H5TOOLS_DLL void trav_table_init(hid_t fid, trav_table_t **table);
H5TOOLS_DLL void trav_table_free(trav_table_t *table);
H5TOOLS_DLL void trav_table_addflags(const unsigned *flags, char *objname, h5trav_type_t type,
trav_table_t *table);
#endif /* H5TRAV_H */