/* ==================================================================== * The Apache Software License, Version 1.1 * * Copyright (c) 2000 The Apache Software Foundation. All rights * reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in * the documentation and/or other materials provided with the * distribution. * * 3. The end-user documentation included with the redistribution, * if any, must include the following acknowledgment: * "This product includes software developed by the * Apache Software Foundation (http://www.apache.org/)." * Alternately, this acknowledgment may appear in the software itself, * if and wherever such third-party acknowledgments normally appear. * * 4. The names "Apache" and "Apache Software Foundation" must * not be used to endorse or promote products derived from this * software without prior written permission. For written * permission, please contact apache@apache.org. * * 5. Products derived from this software may not be called "Apache", * nor may "Apache" appear in their name, without prior written * permission of the Apache Software Foundation. * * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE * DISCLAIMED. IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * ==================================================================== * * This software consists of voluntary contributions made by many * individuals on behalf of the Apache Software Foundation. For more * information on the Apache Software Foundation, please see * . */ #ifndef __AP_CACHE_H__ #define __AP_CACHE_H__ #include #include "httpd.h" #include "apr_file_io.h" #include "apr_network_io.h" #include "apr_pools.h" #include "apr_hooks.h" #include "httpd.h" /** * @package Apache Caching Module API */ /* Interface to caching modules * This interface will allow access to special modules that will * do actual caching calls and maintain elements appropriatly. * * To date there is only a file version and a shared memory of these caching * backends. Clients of thie API need not know where their data will go, in * general there are several calls (marked A) that will work on the database * as a whole. From those points you will go onto (B) where you can seek, create * and remove records. * Upon seeking or creating records you will have an active ap_cache_el. You may * continue down into (C) section. * * A cache element has two distinct parts, the header (D) and the data (E). * One of the uses of the header section will, in fact, be internal to the * cache backends to manage expiration. For example, ap_cache modules may * use the "Cache-Control" header entry for ap_cache_garbage_collect(). * All data portions or headers may be used for any purpose, and are not * actually used by the API, though some headers may have special meanings * to certain backends. */ /* ********************* * Example client usage: * * ap_cache_handle_t *my_cache; * ap_cache_el *element; * apr_file_t *element_buff; * * ap_cache_create(&my_cache, "Cache of Farm Animals"); * * ap_cache_push(my_cache, "Pig", &element); * ap_cache_el_header_add(element, "Sound", "Oink"); * ap_cache_el_data(element, &element_buff); * ap_bputs("I smell bacon!\n", element_buff); * ap_cache_el_finalize(element); * * ap_cache_seek(my_cache", "Cow", &element); * ap_cache_el_header_walk(my_cache, some_func, NULL, "Sound", NULL); * .... * * ap_cache_close(my_cache); * * A client can do anything it wants to an "active" cache_el however it * must guarantee that when it is done with the cache element it will * be finalized. In this way an element in a cache can only be active one * time (and any cache_seek for this element will fail), for this reason * one shouldn't stay open for long ammounts of time. The client is also * responsible for calling garbage_collect periodically to give the cache a * chance to clean up for itself if this is the behaviour it wants. */ /* Types used by clients of this interface */ typedef struct ap_cache_handle_t ap_cache_handle_t; typedef struct ap_cache_el { ap_cache_handle_t *cache; const char *name; } ap_cache_el; /* A) Works on the cache database as a whole */ /** * This will initialize a cache_handle. This is the main entry point into the * caching API, from this point active caching modules will be asked to fill * in the cache_handle. * @param Where to put the handle * @param A descriptive unique string for your client, this description could * used by caching modules to determine if the their backend is suitable * for this client. * @param Current server_rec, this will be used for retreiving configuration, * and various other necesar server pieces. * @deffunc apr_status_t ap_cache_init(ap_cache_handle_t **h, const char *desc, server_rec *r) */ apr_status_t ap_cache_init(ap_cache_handle_t **h, const char *desc, server_rec *r); /** * This function will finalize a cache_handle, after this call the handle will * no longer be usable. * @param The handle to close * @deffunc apr_status_t ap_cache_close(ap_cache_handle_t *) */ apr_status_t ap_cache_close(ap_cache_handle_t *); /** * Force a garbage collection of the cache_handle, the client should call this periodically, * the caching module will not do this on its own, however it isn't required to actually * garbage collect anything, and may defer the call until later. * @param The handle to force a garbage collection. * @deffunc apr_status_t ap_cache_garbage_collect(ap_cache_handle_t *h) */ apr_status_t ap_cache_garbage_collect(ap_cache_handle_t *h); /* B) insertion and query into database */ /** * Seek for a given element in an open cache. This call will fail if the requested element * is already "in use" by previous call to ap_cache_seek or ap_cache_create. * When finished with the element you must call ap_cache_el_finalize immediatly so the * element is no longer locked. * @param The cache to search in. * @param The name of the record you are looking for * @param Where to put the cache element if a seek succeeds. * @deffunc apr_status_t ap_cache_seek(ap_cache_handle_t *h, const char *name, ap_cache_el **) */ apr_status_t ap_cache_seek(ap_cache_handle_t *h, const char *name, ap_cache_el **); /** * Create a new element inside of a cache, you must call this first function to put * something new into a cache; after calling you may use the cache_el passed in as * you would use one retrieved from an ap_cache_seek. The element will be locked after * this call. * When finished with the element you must call ap_cache_el_finalize immediatly so the * element is no longer locked. * @param The cache to create this element in. * @param The name to give this new record * @param Where to put this new element. * @deffunc apr_status_t ap_cache_create(ap_cache_handle_t *h, const char *name, ap_cache_el **) */ apr_status_t ap_cache_create(ap_cache_handle_t *h, const char *name, ap_cache_el **); /** * Remove a record from a cache. This call will fail if the requested element * is already "in use" by previous call to ap_cache_seek or ap_cache_create. * When finished with the element you must call ap_cache_el_finalize immediatly so the * element is no longer locked. * @param The cache to remove this record from. * @param The name of the record to remove from the cache. * @deffunc apr_status_t ap_cache_remove(ap_cache_handle_t *h, const char *name) */ apr_status_t ap_cache_remove(ap_cache_handle_t *h, const char *name); /* (C) Works on an actual element */ /* (D) Works on the header section */ /** * This will retrieve a header value from the element. * @param A previously ap-cache_seek()'d or ap_cache_create()'d element. * @param Header name looking to retrieve, must be null terminated. * @param Where to put the value * @deffunc apr_status_t ap_cache_el_header(ap_cache_el *el, const char *hdr, char **val) */ apr_status_t ap_cache_el_header(ap_cache_el *el, const char *hdr, char **val); /** * Walk through all the headers for given values. This function is synonymous with * ap_table_walk. * @param The element to walk through. * @param The callback function to use for each element. The paramaters for this function: * 1) Client defined data, as passed in by the next paramater to ap-cache_el_header_walk * 2) The name of current header that forced this callback. * 3) The value of the current header. * @param User defined data passed back to the callback as argument 1. * @param NULL terminated list of headers to walk through. If the first value * of this list is NULL then ALL element will be walked over. * @deffunc apr_status_t ap_cache_el_header_walk(ap_cache_el *el, * int (*comp)(void *, const char *, const char *), void *rec, ...); */ apr_status_t ap_cache_el_header_walk(ap_cache_el *el, int (*comp)(void *, const char *, const char *), void *rec, ...); /** * This will merge an existing apr_table_t into a cache_el's header section. * @param The cache element to merge onto. * @param The filled in apr_table_t to merge in. * @deffunc apr_status_t ap_cache_el_header_merge(ap_cache_el *el, apr_table_t *tbl) */ apr_status_t ap_cache_el_header_merge(ap_cache_el *el, apr_table_t *tbl); /** * This will set the current value of a header name to a given value. Using this function * the same as first ap_cache_el_header_remove, and then ap-cache_el_header_add. * @param The cache element to modify * @param The name of the header to change * @param The value to assign to the given name. * @deffunc apr_status_t ap_cache_el_header_set(ap_cache_el *el, const char *hdrname, const char *hdrval) */ apr_status_t ap_cache_el_header_set(ap_cache_el *el, const char *hdrname, const char *hdrval); /** * Each header may have more than one value, you may call this function repeatedly and it will * continue adding values onto a header element. If you want to assign a single value to a * header you must use ap_cache_el_header_set instead. * @param The cache element to add to * @param The name of the header to append values to. * @param The value to append to the given header name. * @deffunc apr_status_t ap_cache_el_header_add(ap_cache_el *el, const char *hdrname, const char *hdrval) */ apr_status_t ap_cache_el_header_add(ap_cache_el *el, const char *hdrname, const char *hdrval); /** * This will remove all headers of a given name. * @param The cache element to remove headers from * @param The name of the header to remove. This will remove ALL values assigned to this * header (via the ap_cache_el_header_add call). * @deffunc apr_status_t ap_cache_el_header_remove(ap_cache_el *el, const char *hdr) */ apr_status_t ap_cache_el_header_remove(ap_cache_el *el, const char *hdr); /** * This will clear out an entire header section. You may use this if you are intending * to change the entire value of the header section of a cache element. * @param The element to clear * @deffunc apr_status_t ap_cache_el_header_clear(ap_cache_el *el) */ apr_status_t ap_cache_el_header_clear(ap_cache_el *el); /* (E) Works on the data section */ /** * Retrieve a apr_file_t for a given cache element, where this data goes is opaque to all * clients of this API. You can do all operations on the apr_file_t and trust the underlying * caching module will accept the data and put it in the appropriate place. * @param The element to retrieve data * @param Where to put the apr_file_t structure when it comes back. In some cases this * will be a normal buff that will either write to a network, or disk - but * you should not rely on it going anywhere in a caching module as the destination * for all data is opaque. * @deffunc apr_status_t ap_cache_el_data(ap_cache_el *el, apr_file_t **) */ apr_status_t ap_cache_el_data(ap_cache_el *el, apr_file_t **); /** * Convenience function to put an existing apr_file_t into a cache_el's data section. This * function will probably not be fully optimal - and will actually just pipe one apr_file_t * to another. * @param The element to append to * @param An existing apr_file_t to append onto the ap_cache_el's stream of data. * @deffunc apr_status_t ap_cache_el_data_append(ap_cache_el *el, apr_file_t *data) */ apr_status_t ap_cache_el_data_append(ap_cache_el *el, apr_file_t *data); /** * Clear the data section of an existing cache_el. You may use this if you are * intending to change the entire value of the data section of a cache element. * @param The element to clear * @deffunc apr_status_t ap_cache_el_data_clear(ap_cache_el *el) */ apr_status_t ap_cache_el_data_clear(ap_cache_el *el); /** * This will complete an open element. When you are done working on a caching * element you must call this so the object will be unlocked and all data will * be finalized, in some cases that means certain data won't make it into the * destination backend until this call is made. Each module may decide how much * this function actually does but you MUST call this function immediatly after * completing a cache record. * @param The element to finalize, after calling this function the caching * element is no longer valid and you must ap_cache_seek for it again if * you want to make any further changes to it. * @deffunc apr_status_t ap_cache_el_finalize(ap_cache_el *el) */ apr_status_t ap_cache_el_finalize(ap_cache_el *el); /* ****************************************************************************/ /* ****************************************************************************/ /* This section is internal entirely, but it is exposed because * implementors of caching modules will need to use some of this. Clients * of the library are NEVER to use this interface however, and should use * the above accessors to the cache. */ /* This is how a cache module can grab control. This will be fired once the * ap_cache_init call is made, each paramater will coorespond to the paramaters * passed into ap_cache_init. If your cache wants to reject a hook call return * APR_ENOTIMPL from your hook and the next caching module will be tried. */ AP_DECLARE_HOOK(apr_status_t, cache_init, (ap_cache_handle_t **, const char *desc, server_rec *t)) /* These are various enum's passed into call back functions (as defined below) */ typedef enum { AP_CACHE_SEEK, AP_CACHE_CREATE, AP_CACHE_CHANGE, AP_CACHE_REMOVE } ap_cache_query; typedef enum { AP_CACHE_DATA, AP_CACHE_HEADER } ap_cache_part; /* These are the callback functions filled in by handler of the cache_init hook. * function may be NULL and will in turn return APR_ENOTIMPL by any of the various * calls in the API. */ typedef struct ap_cache_methods { apr_status_t (*cache_close)(ap_cache_handle_t *h); apr_status_t (*cache_garbage_coll)(ap_cache_handle_t *h); apr_status_t (*cache_element)(ap_cache_handle_t *h, const char *name, ap_cache_el **, ap_cache_query flag); apr_status_t (*cache_el_header_walk)(ap_cache_el *el, int (*comp)(void *, const char *, const char *), void *rec, va_list); apr_status_t (*cache_el_hdr)(ap_cache_el *el, const char *name, const char *val, ap_cache_query flag); apr_status_t (*cache_el_data)(ap_cache_el *el, apr_file_t **); apr_status_t (*cache_el_reset)(ap_cache_el *, ap_cache_part flag); apr_status_t (*cache_el_final)(ap_cache_el *el); } ap_cache_methods; /* This is declared here because modules need to fill this in, however * clients of the library should NEVER use this */ struct ap_cache_handle_t { apr_pool_t *pool; /* pool for alloc's */ server_rec *server; /* access to configurations, used on init */ ap_cache_methods meth; }; #endif /* __AP_CACHE_H__ */