Interface to Write-Once hash table for string to string mapping. More...

#include <stdbool.h>

Include dependency graph for hashtable.h:

This graph shows which files directly or indirectly include this file:

Functions
struct hash_table *	hash_create (unsigned int chains)
	Create a new hash table. More...

void	hash_destroy (struct hash_table *ht)
	Destroys a hash table. More...

bool	hash_add (struct hash_table ht, const char key, const char *value)
	Adds a key/value pair to a hash table. More...

const char *	hash_get (struct hash_table ht, const char key)
	Looks up a the value associated with with a key from a specific hash table. More...

nserror	hash_add_file (struct hash_table ht, const char path)
	Add key/value pairs to a hash table with data from a file. More...

nserror	hash_add_inline (struct hash_table ht, const uint8_t data, size_t size)
	Add key/value pairs to a hash table with data from a memory buffer. More...

Detailed Description

Interface to Write-Once hash table for string to string mapping.

Definition in file hashtable.h.

Function Documentation

◆ hash_add()

bool hash_add	(	struct hash_table *	ht,
		const char *	key,
		const char *	value
	)

Adds a key/value pair to a hash table.

If the key you're adding is already in the hash table, it does not replace it, but it does take precedent over it. The old key/value pair will be inaccessable but still in memory until hash_destroy() is called on the hash table.

Parameters

ht	The hash table context to add the key/value pair to.
key	The key to associate the value with. A copy is made.
value	The value to associate the key with. A copy is made.

Returns: true if the add succeeded, false otherwise. (Failure most likely indicates insufficent memory to make copies of the key and value.

Definition at line 303 of file hashtable.c.

References hash_table::chain, hash_string_fnv(), hash_entry::key_length, hash_table::nchains, hash_entry::next, NSLOG, and hash_entry::pairing.

Referenced by gtk_fetch_filetype_init(), messages_create_ctx(), monkey_fetch_filetype_init(), and process_line().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hash_add_file()

nserror hash_add_file	(	struct hash_table *	ht,
		const char *	path
	)

Add key/value pairs to a hash table with data from a file.

The file should be formatted as a series of lines terminated with newline character. Each line should contain a key/value pair separated by a colon. If a line is empty or starts with a # character it will be ignored.

The file may be optionally gzip compressed.

Parameters

ht	The hash table context to add the key/value pairs to.
path	Path to file with key/value pairs in.

Returns: NSERROR_OK on success else error code

Definition at line 363 of file hashtable.c.

References LINE_BUFFER_SIZE, NSERROR_BAD_PARAMETER, NSERROR_NOT_FOUND, NSERROR_OK, NSLOG, path(), and process_line().

Referenced by messages_load_ctx(), and nsgtk_accelerator_init().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hash_add_inline()

nserror hash_add_inline	(	struct hash_table *	ht,
		const uint8_t *	data,
		size_t	size
	)

Add key/value pairs to a hash table with data from a memory buffer.

The data format is the same as in hash_add_file() but held in memory

The data may optionally be gzip compressed.

Parameters

ht	The hash table context to add the key/value pairs to.
data	Source of key/value pairs
size	length of data

Returns: NSERROR_OK on success else error code

Definition at line 399 of file hashtable.c.

References hash_add_inline_gzip(), and hash_add_inline_plain().

Referenced by messages_add_from_inline(), and nsgtk_accelerator_init().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hash_create()

struct hash_table * hash_create ( unsigned int chains )

Create a new hash table.

Allocate a new hash table and return a context for it. The memory consumption of a hash table is approximately 8 + (nchains * 12) bytes if it is empty.

Parameters

chains Number of chains/buckets this hash table will have. This should be a prime number, and ideally a prime number just over a power of two, for best performance and distribution.