/* $Id: hashmap.c,v 1.13 2003-07-31 23:38:28 rjkaes Exp $
 *
 * A hashmap implementation.  The keys are case-insensitive NULL terminated
 * strings, and the data is arbitrary lumps of data.  Copies of both the
 * key and the data in the hashmap itself, so you must free the original
 * key and data to avoid a memory leak.  The hashmap returns a pointer
 * to the data when a key is searched for, so take care in modifying the
 * data as it's modifying the data stored in the hashmap.  (In other words,
 * don't try to free the data, or realloc the memory. :)
 *
 * Copyright (C) 2002  Robert James Kaes (rjkaes@flarenet.com)
 *
 * 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
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */

#include "tinyproxy.h"

#include "hashmap.h"
#include "heap.h"

/*
 * These structures are the storage for the hashmap.  Entries are stored in
 * struct hashentry_s (the key, data, and length), and all the "buckets" are
 * grouped together in hashmap_s.  The hashmap_s.size member is for
 * internal use.  It stores the number of buckets the hashmap was created
 * with.
 */
struct hashentry_s {
	char *key;
	void *data;
	size_t len;

	struct hashentry_s *prev, *next;
};
struct hashmap_s {
	unsigned int size;
	hashmap_iter end_iterator;

	struct hashentry_s **buckets;
};

/*
 * A NULL terminated string is passed to this function and a "hash" value
 * is produced within the range of [0 .. size)  (In other words, 0 to one
 * less than size.)
 * The contents of the key are converted to lowercase, so this function
 * is not case-sensitive.
 *
 * If any of the arguments are invalid a negative number is returned.
 */
static int
hashfunc(const char *key, unsigned int size)
{
	uint32_t hash;

	if (key == NULL)
		return -EINVAL;
	if (size == 0)
		return -ERANGE;
     
	for (hash = tolower(*key++); *key != '\0'; key++) {
		uint32_t bit =
			(hash & 1) ? (1 << (sizeof(uint32_t) - 1)) : 0;
		hash >>= 1;

		hash += tolower(*key) + bit;
	}

	/* Keep the hash within the table limits */
	return hash % size;
}

/*
 * Create a hashmap with the requested number of buckets.  If "nbuckets" is
 * not greater than zero a NULL is returned; otherwise, a _token_ to the
 * hashmap is returned.
 *
 * NULLs are also returned if memory could not be allocated for hashmap.
 */
hashmap_t
hashmap_create(unsigned int nbuckets)
{
	struct hashmap_s* ptr;

	if (nbuckets == 0)
		return NULL;

	ptr = (struct hashmap_s*)safecalloc(1, sizeof(struct hashmap_s));
	if (!ptr)
		return NULL;

	ptr->size = nbuckets;
	ptr->buckets =
		(struct hashentry_s**)safecalloc(nbuckets,
						 sizeof(struct hashentry_s *));
	if (!ptr->buckets) {
		safefree(ptr);
		return NULL;
	}

	/* This points to "one" past the end of the hashmap. */
	ptr->end_iterator = 0;

	return ptr;
}

/*
 * Follow the chain of hashentries and delete them (including the data and
 * the key.)
 *
 * Returns: 0 if the function completed successfully
 *          negative number is returned if "entry" was NULL
 */
static inline int
delete_hashbucket(struct hashentry_s* entry)
{
	struct hashentry_s *nextptr;
	struct hashentry_s *ptr;

	if (entry == NULL)
		return -EINVAL;

	ptr = entry;
	while (ptr) {
		nextptr = ptr->next;
		
		safefree(ptr->key);
		safefree(ptr->data);
		safefree(ptr);

		ptr = nextptr;
	}

	return 0;
}

/*
 * Deletes a hashmap.  All the key/data pairs are also deleted.
 *
 * Returns: 0 on success
 *          negative if a NULL "map" was supplied
 */
int
hashmap_delete(hashmap_t map)
{
	unsigned int i;
   
	if (map == NULL)
		return -EINVAL;

	for (i = 0; i != map->size; i++) {
		if (map->buckets[i] != NULL) {
			delete_hashbucket(map->buckets[i]);
			map->buckets[i] = NULL;
		}
	}

	safefree(map->buckets);
	safefree(map);

	return 0;
}

/*
 * Inserts a NULL terminated string (as the key), plus any arbitrary "data"
 * of "len" bytes.  Both the key and the data are copied, so the original
 * key/data must be freed to avoid a memory leak.
 * The "data" must be non-NULL and "len" must be greater than zero.  You
 * cannot insert NULL data in association with the key.
 *
 * Returns: 0 on success
 *          negative number if there are errors
 */
int
hashmap_insert(hashmap_t map, const char *key,
	       const void *data, size_t len)
{
	struct hashentry_s *ptr;
	int hash;
	char *key_copy;
	void *data_copy;

	assert(map != NULL);
	assert(key != NULL);
	assert(data != NULL);
	assert(len > 0);

	if (map == NULL || key == NULL)
		return -EINVAL;
	if (!data || len < 1)
		return -ERANGE;

	hash = hashfunc(key, map->size);
	if (hash < 0)
		return hash;

	/*
	 * First make copies of the key and data in case there is a memory
	 * problem later.
	 */
	key_copy = safestrdup(key);
	if (!key_copy)
		return -ENOMEM;

	if (data) {
		data_copy = safemalloc(len);
		if (!data_copy) {
			safefree(key_copy);
			return -ENOMEM;
		}
		memcpy(data_copy, data, len);
	} else {
		data_copy = NULL;
	}

	ptr = (struct hashentry_s*)safemalloc(sizeof(struct hashentry_s));
	if (!ptr) {
		safefree(key_copy);
		safefree(data_copy);
		return -ENOMEM;
	}

	ptr->key = key_copy;
	ptr->data = data_copy;
	ptr->len = len;

	/*
	 * Put the entry at the beginning of the chain.  This is a constant
	 * time insert.  Thanks to Justin Guyett for the code.
	 */
	ptr->prev = NULL;
	ptr->next = map->buckets[hash];
	map->buckets[hash] = ptr;
	if (ptr->next)
		ptr->next->prev = ptr;

	map->end_iterator++;
	return 0;
}

/*
 * Get an iterator to the first entry.
 *
 * Returns: an negative value upon error.
 */
hashmap_iter
hashmap_first(hashmap_t map)
{
	assert(map != NULL);

	if (!map)
		return -EINVAL;

	if (map->end_iterator == 0)
		return -1;
	else
		return 0;
}

/*
 * Checks to see if the iterator is pointing at the "end" of the entries.
 *
 * Returns: 1 if it is the end
 *          0 otherwise
 */
int
hashmap_is_end(hashmap_t map, hashmap_iter iter)
{
	assert(map != NULL);
	assert(iter >= 0);

	if (!map || iter < 0)
		return -EINVAL;

	if (iter == map->end_iterator)
		return 1;
	else
		return 0;
}

/*
 * Return a "pointer" to the first instance of the particular key.  It can
 * be tested against hashmap_is_end() to see if the key was not found.
 *
 * Returns: negative upon an error
 *          an "iterator" pointing at the first key
 *          an "end-iterator" if the key wasn't found
 */
hashmap_iter
hashmap_find(hashmap_t map, const char* key)
{
	unsigned int i;
	hashmap_iter iter = 0;
	struct hashentry_s* ptr;

	assert(map != NULL);
	assert(key != NULL);

	if (!map || !key)
		return -EINVAL;

	/*
	 * Loop through all the keys and look for the first occurrence
	 * of a particular key.
	 */
	for (i = 0; i != map->size; i++) {
		ptr = map->buckets[i];

		while (ptr) {
			if (strcasecmp(ptr->key, key) == 0) {
				/* Found it, so return the current count */
				return iter;
			}

			iter++;
			ptr = ptr->next;
		}
	}

	return iter;
}

/*
 * Retrieve the data associated with a particular iterator.
 *
 * Returns: the length of the data block upon success
 *          negative upon error
 */
ssize_t
hashmap_return_entry(hashmap_t map, hashmap_iter iter,
		     char** key, void** data)
{
	unsigned int i;
	struct hashentry_s* ptr;
	hashmap_iter count = 0;

	assert(map != NULL);
	assert(iter >= 0);
	assert(iter != map->end_iterator);
	assert(key != NULL);
	assert(data != NULL);

	if (!map || iter < 0 || !key || !data)
		return -EINVAL;

	for (i = 0; i != map->size; i++) {
		ptr = map->buckets[i];
		while (ptr) {
			if (count == iter) {
				/* This is the data so return it */
				*key = ptr->key;
				*data = ptr->data;
				return ptr->len;
			}

			ptr = ptr->next;
			count++;
		}
	}

	return -EFAULT;
}

/*
 * Searches for _any_ occurrences of "key" within the hashmap.
 * 
 * Returns: negative upon an error
 *          zero if no key is found
 *          count found
 */
ssize_t
hashmap_search(hashmap_t map, const char *key)
{
	int hash;
	struct hashentry_s* ptr;
	ssize_t count = 0;

	if (map == NULL || key == NULL)
		return -EINVAL;

	hash = hashfunc(key, map->size);
	if (hash < 0)
		return hash;

	ptr = map->buckets[hash];

	/* All right, there is an entry here, now see if it's the one we want */
	while (ptr) {
		if (strcasecmp(ptr->key, key) == 0)
			++count;

		/* This entry didn't contain the key; move to the next one */
		ptr = ptr->next;
	}

	return count;
}

/*
 * Get the first entry (assuming there is more than one) for a particular
 * key.  The data MUST be non-NULL.
 *
 * Returns: negative upon error
 *          zero if no entry is found
 *          length of data for the entry
 */
ssize_t
hashmap_entry_by_key(hashmap_t map, const char* key, void** data)
{
	int hash;
	struct hashentry_s* ptr;

	if (!map || !key || !data)
		return -EINVAL;
       
	hash = hashfunc(key, map->size);
	if (hash < 0)
		return hash;

	ptr = map->buckets[hash];

	while (ptr) {
		if (strcasecmp(ptr->key, key) == 0) {
			*data = ptr->data;
			return ptr->len;
		}

		ptr = ptr->next;
	}

	return 0;
}

/*
 * Go through the hashmap and remove the particular key.
 * NOTE: This will invalidate any iterators which have been created.
 *
 * Remove: negative upon error
 *         0 if the key was not found
 *         positive count of entries deleted
 */
ssize_t
hashmap_remove(hashmap_t map, const char *key)
{
	int hash;
	struct hashentry_s* ptr;
	short int deleted = 0;

	if (map == NULL || key == NULL)
		return -EINVAL;

	hash = hashfunc(key, map->size);
	if (hash < 0)
		return hash;

	ptr = map->buckets[hash];
	while (ptr) {
		if (strcasecmp(ptr->key, key) == 0) {
			/*
			 * Found the data, now need to remove everything
			 * and update the hashmap.
			 */
			struct hashentry_s* prevptr = ptr->prev;
			if (prevptr != NULL) {
				prevptr->next = ptr->next;
				if (ptr->next)
					ptr->next->prev = prevptr;
			} else {
				/* Entry was first in map */
				map->buckets[hash] = ptr->next;
				if (ptr->next)
					ptr->next->prev = NULL;
			}

			safefree(ptr->key);
			safefree(ptr->data);
			safefree(ptr);
			
			++deleted;
			--map->end_iterator;

			if (prevptr)
				ptr = prevptr;
			else
				ptr = map->buckets[hash];

			continue;
		}

		/* This entry didn't contain the key; move to the next one */
		ptr = ptr->next;
	}

	/* The key was not found, so return 0 */
	return deleted;
}