Diff of Roxen.git/server/base_server/module.pike (71a2d5c3d2d784ce65edf34428276f3440db5a50 » 6f891023e4e6101175f84e224ad53ecc66309e74)

Roxen.git/server/base_server/module.pike:1: - // This file is part of Roxen WebServer. - // Copyright © 1996 - 2009, Roxen IS. - // $Id$ + - #include <module_constants.h> - #include <module.h> - #include <request_trace.h> - - constant __pragma_save_parent__ = 1; - - // Tell Pike.count_memory this is global. - constant pike_cycle_depth = 0; - - inherit "basic_defvar"; - mapping(string:array(int)) error_log=([]); - - constant is_module = 1; - // constant module_type = MODULE_ZERO; - // constant module_name = "Unnamed module"; - // constant module_doc = "Undocumented"; - constant module_unique = 1; - - //! Specifies that the module is opaque when it comes WebDAV - //! requests. Normally, recursive WebDAV requests will iterate through - //! all matching location modules even after a successful result has - //! been returned by some module. With this flag set, iteration will - //! stop after the call to this module. Useful if the module wants to - //! handle all requests for the specified location itself with no - //! fallback to other modules. - constant webdav_opaque = 0; - - private Configuration _my_configuration; - private string _module_local_identifier; - private string _module_identifier = - lambda() { - mixed init_info = roxen->bootstrap_info->get(); - if (arrayp (init_info)) { - [_my_configuration, _module_local_identifier] = init_info; - return _my_configuration->name + "/" + _module_local_identifier; - } - #ifdef DEBUG - else - error ("Got invalid bootstrap info for module: %O\n", init_info); - #endif - }(); - protected mapping _api_functions = ([]); - - class ModuleJSONLogger { - inherit Logger.BaseJSONLogger; - - void create(object parent_config) { - string name = combine_path_unix(parent_config->json_logger->logger_name, - module_local_id()); - ::create(name, UNDEFINED, parent_config->json_logger); - } - } - - // Module local JSON logger - private ModuleJSONLogger json_logger; - - string|array(string) module_creator; - string module_url; - RXML.TagSet module_tag_set; - - /* These functions exist in here because otherwise the messages in the - * event log do not always end up in the correct module/configuration. - * And the reason for that is that if the messages are logged from - * subclasses in the module, the DWIM in roxenlib.pike cannot see that - * they are logged from a module. This solution is not really all that - * beautiful, but it works. :-) - */ - void report_fatal(sprintf_format fmt, sprintf_args ... args) - { predef::report_fatal(fmt, @args); } - void report_error(sprintf_format fmt, sprintf_args ... args) - { predef::report_error(fmt, @args); } - void report_warning(sprintf_format fmt, sprintf_args ... args) - { predef::report_warning(fmt, @args); } - void report_notice(sprintf_format fmt, sprintf_args ... args) - { predef::report_notice(fmt, @args); } - void report_debug(sprintf_format fmt, sprintf_args ... args) - { predef::report_debug(fmt, @args); } - void report_warning_sparsely (sprintf_format fmt, sprintf_args ... args) - {predef::report_warning_sparsely (fmt, @args);} - void report_error_sparsely (sprintf_format fmt, sprintf_args ... args) - {predef::report_error_sparsely (fmt, @args);} - - void log_event (string facility, string action, string resource, - void|mapping(string:mixed) info) - //! Log an event. See @[Configuration.log_event] for details. - //! - //! @[facility] may be zero. The local module identifier as returned - //! by @[module_local_id] is used as facility in that case. - { - _my_configuration->log_event (facility || _module_local_identifier, - action, resource, info); - } - - void json_log_trace(string|mapping log_msg) { json_log_with_level(log_msg, Logger.BaseJSONLogger.TRACE); } - void json_log_debug(string|mapping log_msg) { json_log_with_level(log_msg, Logger.BaseJSONLogger.DBG); } - void json_log_info (string|mapping log_msg) { json_log_with_level(log_msg, Logger.BaseJSONLogger.INFO); } - void json_log_warn (string|mapping log_msg) { json_log_with_level(log_msg, Logger.BaseJSONLogger.WARN); } - void json_log_error(string|mapping log_msg) { json_log_with_level(log_msg, Logger.BaseJSONLogger.ERROR); } - void json_log_fatal(string|mapping log_msg) { json_log_with_level(log_msg, Logger.BaseJSONLogger.FATAL); } - - // Helper method to force a specific logging level - void json_log_with_level(string|mapping log_msg, int level) { - if (stringp(log_msg)) { - log_msg = ([ - "msg" : log_msg, - ]); - } - log_msg->level = level; - json_log(log_msg); - } - - // Log a message more or less verbatim via the JSON logger infrastructure - void json_log(string|mapping log_msg) { - if (stringp(log_msg)) { - log_msg = ([ - "msg" : log_msg, - ]); - } - - if (json_logger && functionp(json_logger->log)) { - json_logger->log(log_msg); - } - } - - string module_identifier() - //! Returns a string that uniquely identifies this module instance - //! within the server. The identifier is the same as - //! @[Roxen.get_module] and @[Roxen.get_modname] handles. - { - return _module_identifier; - } - - string module_local_id() - //! Returns a string that uniquely identifies this module instance - //! within the configuration. The returned string is the same as the - //! part after the first '/' in the one returned from - //! @[module_identifier]. - { - return _module_local_identifier; - } - - RoxenModule this_module() - { - return this_object(); // To be used from subclasses. - } - - //! @ignore - DECLARE_OBJ_COUNT; - //! @endignore - - string _sprintf() - { - return sprintf ("RoxenModule(%s)" + OBJ_COUNT, _module_identifier || "?"); - } - - array register_module() - { - return ({ - this_object()->module_type, - this_object()->module_name, - this_object()->module_doc, - 0, - module_unique, - this_object()->module_locked, - this_object()->module_counter, - }); - } - - string fix_cvs(string from) - { - from = replace(from, ({ "$", "Id: "," Exp $" }), ({"","",""})); - sscanf(from, "%*s,v %s", from); - return replace(from,"/","-"); - } - - int module_dependencies(Configuration configuration, - array (string) modules, - int|void now) - //! If your module depends on other modules, call this function to - //! ensure that those modules get loaded. - //! - //! @param configuration - //! The configuration for the modules. Use zero for the same as this - //! module. - //! - //! @param modules - //! An array of module identifiers. A module identifier is either - //! the name of the pike file minus extension, or a string on the - //! form that @[Roxen.get_modname] returns. In the latter case, the - //! @tt{<config name>@} and @tt{<copy>@} parts are ignored. - //! - //! @param now - //! If this flag is nonzero then any modules that aren't already - //! loaded get loaded (and started) right away, so that the code - //! following the @[module_dependencies] call can start using them. - //! - //! Otherwise the function only ensures that the modules exist in - //! the configuration and they will be loaded sometime during the - //! startup of it. - //! - //! @note - //! This function is only intended to be called from @[start]. - //! - //! @note - //! The @[now] flag does not affect calls to - //! @[ready_to_receive_requests] in the listed modules. I.e. even if - //! you have declared a dependency on a module and have @[now] set, - //! you cannot assume its @[ready_to_receive_requests] function has - //! run (in fact, you can almost safely assume it hasn't). You can - //! however assume that its @[start] function has been called. - { - modules = map (modules, - lambda (string modname) { - sscanf ((modname / "/")[-1], "%[^#]", modname); - return modname; - }); - Configuration conf = configuration || my_configuration(); - if (!conf) - report_warning ("Configuration not resolved; module(s) %s that %O " - "depend on weren't added.\n", String.implode_nicely (modules), - module_identifier() || - master()->describe_program(this_program) || - "unknown module"); - else - conf->add_modules( modules, now ); - return 1; - } - - string file_name_and_stuff() - { - return ("Loaded from: "+(roxen->filename(this_object()))+" "+ - (this_object()->cvs_version? - "CVS Version: "+ - fix_cvs(this_object()->cvs_version)+"\n":"")); - } - - - Configuration my_configuration() - //! Returns the Configuration object of the virtual server the module - //! belongs to. - { - return _my_configuration; - } - - final void set_configuration(Configuration c) - { - if(_my_configuration && _my_configuration != c) - error("set_configuration() called twice.\n"); - _my_configuration = c; - - // if configuration changes, we should reinitialize our JSON logger too! - json_logger = ModuleJSONLogger(_my_configuration); - } - - void set_module_creator(string|array(string) c) - //! Set the name and optionally email address of the author of the - //! module. Names on the format "author name <author_email>" will - //! end up as links on the module's information page in the admin - //! interface. In the case of multiple authors, an array of such - //! strings can be passed. - { - module_creator = c; - } - - void set_module_url(string to) - //! A common way of referring to a location where you maintain - //! information about your module or similar. The URL will turn up - //! on the module's information page in the admin interface, - //! referred to as the module's home page. - { - module_url = to; - } - - void free_some_sockets_please(){} - - // These functions have bodies to make module inheritance easier: An - // inheriting module can always assume that the module it inherits - // have these functions defined and properly proxy the calls to them. - // Thus an inherited module should be free to define them later on - // even if it didn't have them initially. (Modules need never proxy - // calls to the default implementations here.) - void start(int variable_save, Configuration conf, void|int newly_added) {} - void stop() {} - void ready_to_receive_requests (Configuration conf) {} - - string status() {} - - string info(Configuration conf) - { - return (this_object()->register_module()[2]); - } - - string sname( ) - { - return my_configuration()->otomod[ this_object() ]; - } - - ModuleInfo my_moduleinfo( ) - //! Returns the associated @[ModuleInfo] object - { - string f = sname(); - if( f ) return roxen.find_module( (f/"#")[0] ); - } - - void save_me() - { - my_configuration()->save_one( this_object() ); - my_configuration()->module_changed( my_moduleinfo(), this_object() ); - } - - void save() { save_me(); } - string comment() { return ""; } - - string query_internal_location() - //! Returns the internal mountpoint, where <ref>find_internal()</ref> - //! is mounted. It always ends with a '/'. - { - if(!_my_configuration) - error("Please do not call this function from create()!\n"); - return _my_configuration->query_internal_location(this_object()); - } - - string query_absolute_internal_location(RequestID id) - //! Returns the internal mountpoint as an absolute path. It always - //! ends with a '/'. - { - return (id->misc->site_prefix_path || "") + query_internal_location(); - } - - string query_location() - //! Returns the mountpoint as an absolute path. The default - //! implementation uses the "location" configuration variable in the - //! module. - { - string s; - catch{s = query("location");}; - return s; - } - - array(string) location_urls() - //! Returns an array of all locations where the module is mounted. - { - string loc = query_location(); - if (!loc) return ({}); - if(!_my_configuration) - error("Please do not call this function from create()!\n"); - array(string) urls = copy_value(_my_configuration->query("URLs")); - string hostname; - if (string world_url = _my_configuration->query ("MyWorldLocation")) - if (sizeof(world_url)) { - Standards.URI uri = Standards.URI(world_url); - hostname = uri->host; - } - if (!hostname) hostname = gethostname(); - for (int i = 0; i < sizeof (urls); i++) - { - urls[i] = (urls[i]/"#")[0]; - if (sizeof (urls[i]/"*") == 2) - urls[i] = replace(urls[i], "*", hostname); - } - return map (urls, `+, loc[1..]); - } - - string location_url() - //! Returns an http or https url including the modules mountpoint. The - //! ip-number for the corresponding port will be added to the fragment - //! of the url. An http url will be prioritized over an https url. - { - string short_array(array a) - { - return "({ " + (map(a, lambda(object o) { - return sprintf("%O", o); - })*", ") + " })"; - }; - string loc = query_location(); - if(!loc) return 0; - if(!_my_configuration) - error("Please do not call this function from create()!\n"); - string hostname; - string world_url = _my_configuration->query("MyWorldLocation"); - if (world_url && sizeof(world_url)) { - Standards.URI uri = Standards.URI(world_url); - hostname = uri->host; - } - if(!hostname) - hostname = gethostname(); - #ifdef LOCATION_URL_DEBUG - werror(" Hostname: %O\n", hostname); - #endif - Standards.URI candidate_uri; - array(string) urls = - filter(_my_configuration->registered_urls, has_prefix, "http:") + - filter(_my_configuration->registered_urls, has_prefix, "https:"); - foreach(urls, string url) - { - #ifdef LOCATION_URL_DEBUG - werror(" URL: %s\n", url); - #endif - mapping url_info = roxen.urls[url]; - if(!url_info || !url_info->port || url_info->conf != _my_configuration) - continue; - Protocol p = url_info->port; - #ifdef LOCATION_URL_DEBUG - werror(" Protocol: %s\n", p); - #endif - Standards.URI uri = Standards.URI(url); - string ip = p->ip || "127.0.0.1"; - if (ip == "::") - ip = "::1"; - uri->fragment = "ip=" + ip; - if(has_value(uri->host, "*") || has_value(uri->host, "?")) - if(glob(uri->host, hostname)) - uri->host = hostname; - else { - if(!candidate_uri) { - candidate_uri = uri; - candidate_uri->host = hostname; - } - continue; - } - uri->path += loc[1..]; - return (string)uri; - } - if(candidate_uri) { - report_warning("Warning: Could not find any suitable ports, continuing anyway. " - "Please make sure that your Primary Server URL matches " - "at least one port. Primary Server URL: %O, URLs: %s.\n", - world_url, short_array(urls)); - candidate_uri->path += loc[1..]; - return (string)candidate_uri; - } - return 0; - } - - /* By default, provide nothing. */ - multiset(string) query_provides() { return 0; } - - - function(RequestID:int|mapping) query_seclevels() - { - if(catch(query("_seclevels")) || (query("_seclevels") == 0)) - return 0; - return roxen.compile_security_pattern(query("_seclevels"),this_object()); - } - - void set_status_for_path (string path, RequestID id, int status_code, - string|void message, mixed... args) - //! Register a status to be included in the response that applies only - //! for the given path. This is used for recursive operations that can - //! yield different results for different encountered files or - //! directories. - //! - //! The status is stored in the @[MultiStatus] object returned by - //! @[id->get_multi_status]. The server will use it to make a 207 - //! Multi-Status response iff the module returns an empty mapping as - //! response. - //! - //! @param path - //! Path (below the filesystem location) to which the status applies. - //! - //! @param status_code - //! The HTTP status code. - //! - //! @param message - //! If given, it's a message to include in the response. The - //! message may contain line feeds ('\n') and ISO-8859-1 - //! characters in the ranges 32..126 and 128..255. Line feeds are - //! converted to spaces if the response format doesn't allow them. - //! - //! @param args - //! If there are more arguments after @[message] then @[message] - //! is taken as an @[sprintf] style format string which is used to - //! format @[args]. - //! - //! @note - //! This function is just a wrapper for @[id->set_status_for_path] - //! that corrects for the filesystem location. - //! - //! @seealso - //! @[RequestID.set_status_for_path], @[Roxen.http_status] - { - if (sizeof (args)) message = sprintf (message, @args); - id->set_status_for_path (query_location() + path, status_code, message); - } - - Stat stat_file(string f, RequestID id){} - array(string) find_dir(string f, RequestID id){} - mapping(string:Stat) find_dir_stat(string f, RequestID id) - { - SIMPLE_TRACE_ENTER(this, "find_dir_stat(): %O", f); - - array(string) files = find_dir(f, id); - mapping(string:Stat) res = ([]); - - foreach(files || ({}), string fname) { - SIMPLE_TRACE_ENTER(this, "stat()'ing %O", f + "/" + fname); - Stat st = stat_file(replace(f + "/" + fname, "//", "/"), id); - if (st) { - res[fname] = st; - TRACE_LEAVE("OK"); - } else { - TRACE_LEAVE("No stat info"); - } - } - - TRACE_LEAVE(""); - return(res); - } - - class DefaultPropertySet - { - inherit PropertySet; - - protected Stat stat; - - protected void create (string path, string abs_path, RequestID id, Stat stat) - { - ::create (path, abs_path, id); - this_program::stat = stat; - } - - Stat get_stat() {return stat;} - - protected mapping(string:string) response_headers; - - mapping(string:string) get_response_headers() - { - if (!response_headers) { - // Old kludge inherited from configuration.try_get_file. - if (!id->misc->common) - id->misc->common = ([]); - - RequestID sub_id = id->clone_me(); - sub_id->misc->common = id->misc->common; - - sub_id->raw_url = sub_id->not_query = query_location() + path; - if ((sub_id->raw_url != id->raw_url) && (id->raw_url != id->not_query)) { - // sub_id->raw_url = replace (id->raw_url, id->not_query, sub_id->not_query); - sub_id->raw_url = sub_id->not_query + - (({ "" }) + (id->raw_url/"?")[1..]) * "?"; - } - sub_id->method = "HEAD"; - - mapping(string:mixed)|int(-1..0)|object res = find_file (path, sub_id); - if (res == -1) res = ([]); - else if (objectp (res)) { - string ext; - if(stringp(sub_id->extension)) { - sub_id->not_query += sub_id->extension; - ext = lower_case(Roxen.extension(sub_id->not_query, sub_id)); - } - array(string) tmp=sub_id->conf->type_from_filename(sub_id->not_query, 1, ext); - if(tmp) - res = ([ "file":res, "type":tmp[0], "encoding":tmp[1] ]); - else - res = (["file": res]); - } - response_headers = sub_id->make_response_headers (res); - destruct (sub_id); - } - - return response_headers; - } - } - - //! Return the set of properties for @[path]. - //! - //! @returns - //! Returns @tt{0@} (zero) if @[path] does not exist. - //! - //! Returns an error mapping if there's some other error accessing - //! the properties. - //! - //! Otherwise returns a @[PropertySet] object. - //! - //! @seealso - //! @[query_property()] - PropertySet|mapping(string:mixed) query_property_set(string path, RequestID id) - { - SIMPLE_TRACE_ENTER (this, "Querying properties on %O", path); - Stat st = stat_file(path, id); - - if (!st) { - SIMPLE_TRACE_LEAVE ("No such file or dir"); - return 0; - } - - PropertySet res = DefaultPropertySet(path, query_location()+path, id, st); - SIMPLE_TRACE_LEAVE (""); - return res; - } - - //! Returns the value of the specified property, or an error code - //! mapping. - //! - //! @note - //! Returning a string is shorthand for returning an array - //! with a single text node. - //! - //! @seealso - //! @[query_property_set()] - string|array(Parser.XML.Tree.SimpleNode)|mapping(string:mixed) - query_property(string path, string prop_name, RequestID id) - { - mapping(string:mixed)|PropertySet properties = query_property_set(path, id); - if (!properties) { - return Roxen.http_status(Protocols.HTTP.HTTP_NOT_FOUND, - "No such file or directory."); - } - if (mappingp (properties)) - return properties; - return properties->query_property(prop_name) || - Roxen.http_status(Protocols.HTTP.HTTP_NOT_FOUND, "No such property."); - } - - //! RFC 2518 PROPFIND implementation with recursion according to - //! @[depth]. See @[PropertySet()->find_properties()] for details. - //! - //! @seealso - //! @[query_property_set()] - mapping(string:mixed) recurse_find_properties(string path, string mode, - int depth, RequestID id, - multiset(string)|void filt) - { - string prefix = map(query_location()[1..]/"/", Roxen.http_encode_url)*"/"; - MultiStatus.Prefixed result = - id->get_multi_status()->prefix (id->url_base() + prefix); - - mapping(string:mixed) recurse (string path, int depth) { - SIMPLE_TRACE_ENTER (this, "%s for %O, depth %d", - mode == "DAV:propname" ? "Listing property names" : - mode == "DAV:allprop" ? "Retrieving all properties" : - mode == "DAV:prop" ? "Retrieving specific properties" : - "Finding properties with mode " + mode, - path, depth); - mapping(string:mixed)|PropertySet properties = query_property_set(path, id); - - if (!properties) { - SIMPLE_TRACE_LEAVE ("No such file or dir"); - return 0; - } - - { - mapping(string:mixed) ret = mappingp (properties) ? - properties : properties->find_properties(mode, result, filt); - - if (ret) { - SIMPLE_TRACE_LEAVE ("Got status %d: %O", ret->error, ret->rettext); - return ret; - } - } - - if (properties->get_stat()->isdir) { - if (depth <= 0) { - SIMPLE_TRACE_LEAVE ("Not recursing due to depth limit"); - return ([]); - } - depth--; - foreach(find_dir(path, id) || ({}), string filename) { - filename = combine_path_unix(path, filename); - if (mapping(string:mixed) sub_res = recurse(filename, depth)) - if (sizeof (sub_res)) - result->add_status (filename, sub_res->error, sub_res->rettext); - } - } - - SIMPLE_TRACE_LEAVE (""); - return ([]); - }; - - return recurse (path, depth); - } - - mapping(string:mixed) patch_properties(string path, - array(PatchPropertyCommand) instructions, - RequestID id) - { - SIMPLE_TRACE_ENTER (this, "Patching properties for %O", path); - mapping(string:mixed)|PropertySet properties = query_property_set(path, id); - - if (!properties) { - SIMPLE_TRACE_LEAVE ("No such file or dir"); - return 0; - } - if (mappingp (properties)) { - SIMPLE_TRACE_LEAVE ("Got error %d from query_property_set: %O", - properties->error, properties->rettext); - return properties; - } - - mapping(string:mixed) errcode; - - if (errcode = write_access(path, 0, id)) { - SIMPLE_TRACE_LEAVE("Patching denied by write_access()."); - return errcode; - } - - if (errcode = properties->start()) { - SIMPLE_TRACE_LEAVE ("Got error %d from PropertySet.start: %O", - errcode->error, errcode->rettext); - return errcode; - } - - array(mapping(string:mixed)) results; - - mixed err = catch { - results = instructions->execute(properties); - }; - if (err) { - properties->unroll(); - throw (err); - } else { - string prefix = map((query_location()[1..] + path)/"/", - Roxen.http_encode_url)*"/"; - MultiStatus.Prefixed result = - id->get_multi_status()->prefix (id->url_base() + prefix); - int any_failed; - foreach(results, mapping(string:mixed) answer) { - if (any_failed = (answer && (answer->error >= 300))) { - break; - } - } - if (any_failed) { - // Unroll and fail any succeeded items. - int i; - mapping(string:mixed) answer = - Roxen.http_status (Protocols.HTTP.DAV_FAILED_DEP); - for(i = 0; i < sizeof(results); i++) { - if (!results[i] || results[i]->error < 300) { - result->add_property("", instructions[i]->property_name, - answer); - } else { - result->add_property("", instructions[i]->property_name, - results[i]); - } - } - properties->unroll(); - } else { - int i; - for(i = 0; i < sizeof(results); i++) { - result->add_property("", instructions[i]->property_name, - results[i]); - } - properties->commit(); - } - } - - SIMPLE_TRACE_LEAVE (""); - return 0; - } - - //! Convenience variant of @[patch_properties()] that sets a single - //! property. - //! - //! @returns - //! Returns a mapping on any error, zero otherwise. - mapping(string:mixed) set_property (string path, string prop_name, - string|array(Parser.XML.Tree.SimpleNode) value, - RequestID id) - { - mapping(string:mixed)|PropertySet properties = query_property_set(path, id); - if (!properties) return Roxen.http_status(Protocols.HTTP.HTTP_NOT_FOUND, - "File not found."); - if (mappingp (properties)) return properties; - - mapping(string:mixed) result = properties->start(); - if (result) return result; - - result = properties->set_property(prop_name, value); - if (result && result->error >= 300) { - properties->unroll(); - return result; - } - - properties->commit(); - return 0; - } - - //! Convenience variant of @[patch_properties()] that removes a single - //! property. - //! - //! @returns - //! Returns a mapping on any error, zero otherwise. - mapping(string:mixed) remove_property (string path, string prop_name, - RequestID id) - { - mapping(string:mixed)|PropertySet properties = query_property_set(path, id); - if (!properties) return Roxen.http_status(Protocols.HTTP.HTTP_NOT_FOUND, - "File not found."); - if (mappingp (properties)) return properties; - - mapping(string:mixed) result = properties->start(); - if (result) return result; - - result = properties->remove_property(prop_name); - if (result && result->error >= 300) { - properties->unroll(); - return result; - } - - properties->commit(); - return 0; - } - - string resource_id (string path, RequestID|int(0..0) id) - //! Return a string that within the filesystem uniquely identifies the - //! resource on @[path] in the given request. This is commonly @[path] - //! itself but can be extended with e.g. language, user or some form - //! variable if the path is mapped to different files according to - //! those fields. - //! - //! The important criteria here is that every unique returned string - //! corresponds to a resource that can be changed independently of - //! every other. Thus e.g. dynamic pages that evaluate to different - //! results depending on variables or cookies etc should _not_ be - //! mapped to more than one string by this function. It also means - //! that if files are stored in a filesystem which is case insensitive - //! then this function should normalize case differences. - //! - //! This function is used e.g by the default lock implementation to - //! convert paths to resources that can be locked independently of - //! each other. There's also a notion of recursive locks there, which - //! means that a recursive lock on a certain resource identifier also - //! locks every resource whose identifier it is a prefix of. Therefore - //! it's typically necessary to ensure that every identifier ends with - //! "/" so that a recursive lock on e.g. "doc/foo" doesn't lock - //! "doc/foobar". - //! - //! @param path - //! The requested path below the filesystem location. It has been - //! normalized with @[VFS.normalize_path]. - //! - //! @param id - //! The request id may have the value @expr{0@} (zero) if called - //! by @[Configuration()->expire_locks()]. - { - return has_suffix (path, "/") ? path : path + "/"; - } - - string|int authenticated_user_id (string path, RequestID id) - //! Return a value that uniquely identifies the user that the given - //! request is authenticated as. - //! - //! This function is e.g. used by the default lock implementation to - //! tell different users holding locks apart. WARNING: Due to some - //! design issues in the lock system, it's likely that it will change - //! to not use this function in the future. - //! - //! @param path - //! The requested path below the filesystem location. It has been - //! normalized with @[VFS.normalize_path]. - { - // Leave this to the standard auth system by default. - User uid = my_configuration()->authenticate (id); - return uid && uid->name(); - } - - // Mapping from resource id to a mapping from user id to the lock - // that apply to the resource. - // - // Only used internally by the default lock implementation. - protected mapping(string:mapping(mixed:DAVLock)) file_locks = ([]); - - // Mapping from resource id to a mapping from user id to the lock - // that apply recursively to the resource and all other resources - // it's a prefix of. - // - // Only used internally by the default lock implementation. - protected mapping(string:mapping(mixed:DAVLock)) prefix_locks = ([]); - - #define LOOP_OVER_BOTH(PATH, LOCKS, CODE) \ - do { \ - foreach (file_locks; PATH; LOCKS) {CODE;} \ - foreach (prefix_locks; PATH; LOCKS) {CODE;} \ - } while (0) - - //! Find some or all locks that apply to @[path]. - //! - //! @param path - //! Normalized path below the filesystem location. - //! - //! @param recursive - //! If @expr{1@} also return locks anywhere below @[path]. - //! - //! @param exclude_shared - //! If @expr{1@} do not return shared locks that are held by users - //! other than the one the request is authenticated as. (This is - //! appropriate to get the list of locks that would conflict if the - //! current user were to make a shared lock.) - //! - //! @returns - //! Returns a multiset containing all applicable locks in - //! this location module, or @expr{0@} (zero) if there are none. - //! - //! @note - //! @[DAVLock] objects may be created if the filesystem has some - //! persistent storage of them. The default implementation does not - //! store locks persistently. - //! - //! @note - //! The default implementation only handles the @expr{"DAV:write"@} - //! lock type. - multiset(DAVLock) find_locks(string path, - int(0..1) recursive, - int(0..1) exclude_shared, - RequestID id) - { - // Common case. - if (!sizeof(file_locks) && !sizeof(prefix_locks)) return 0; - - TRACE_ENTER(sprintf("find_locks(%O, %O, %O, X)", - path, recursive, exclude_shared), this); - - string rsc = resource_id (path, id); - - multiset(DAVLock) locks = (<>); - function(mapping(mixed:DAVLock):void) add_locks; - - if (exclude_shared) { - mixed auth_user = authenticated_user_id (path, id); - add_locks = lambda (mapping(mixed:DAVLock) sub_locks) { - foreach (sub_locks; string user; DAVLock lock) - if (user == auth_user || - lock->lockscope == "DAV:exclusive") - locks[lock] = 1; - }; - } - else - add_locks = lambda (mapping(mixed:DAVLock) sub_locks) { - locks |= mkmultiset (values (sub_locks)); - }; - - if (file_locks[rsc]) { - add_locks (file_locks[rsc]); - } - - foreach(prefix_locks; - string prefix; mapping(mixed:DAVLock) sub_locks) { - if (has_prefix(rsc, prefix)) { - add_locks (sub_locks); - break; - } - } - - if (recursive) { - LOOP_OVER_BOTH (string prefix, mapping(mixed:DAVLock) sub_locks, { - if (has_prefix(prefix, rsc)) { - add_locks (sub_locks); - } - }); - } - - add_locks = 0; - - TRACE_LEAVE(sprintf("Done, found %d locks.", sizeof(locks))); - - return sizeof(locks) && locks; - } - - //! Check if there are one or more locks that apply to @[path] for the - //! user the request is authenticated as. - //! - //! WARNING: This function has some design issues and will very likely - //! get a different interface. Compatibility is NOT guaranteed. - //! - //! @param path - //! Normalized path below the filesystem location. - //! - //! @param recursive - //! If @expr{1@} also check recursively under @[path] for locks. - //! - //! @returns - //! The valid return values are: - //! @mixed - //! @type DAVLock - //! The lock owned by the authenticated user that apply to - //! @[path]. (It doesn't matter if the @expr{recursive@} flag in - //! the lock doesn't match the @[recursive] argument.) - //! @type LockFlag - //! @int - //! @value LOCK_NONE - //! No locks apply. (0) - //! @value LOCK_SHARED_BELOW - //! There are only one or more shared locks held by other - //! users somewhere below @[path] (but not on @[path] - //! itself). Only returned if @[recursive] is set. (2) - //! @value LOCK_SHARED_AT - //! There are only one or more shared locks held by other - //! users on @[path]. (3) - //! @value LOCK_OWN_BELOW - //! The authenticated user has locks under @[path] (but not - //! on @[path] itself) and there are no exclusive locks held - //! by other users. Only returned if @[recursive] is set. (4) - //! @value LOCK_EXCL_BELOW - //! There are one or more exclusive locks held by other - //! users somewhere below @[path] (but not on @[path] - //! itself). Only returned if @[recursive] is set. (6) - //! @value LOCK_EXCL_AT - //! There are one or more exclusive locks held by other - //! users on @[path]. (7) - //! @endint - //! Note that the lowest bit is set for all flags that apply to - //! @[path] itself. - //! @endmixed - //! - //! @note - //! @[DAVLock] objects may be created if the filesystem has some - //! persistent storage of them. The default implementation does not - //! store locks persistently. - //! - //! @note - //! The default implementation only handles the @expr{"DAV:write"@} - //! lock type. - DAVLock|LockFlag check_locks(string path, - int(0..1) recursive, - RequestID id) - { - TRACE_ENTER(sprintf("check_locks(%O, %d, X)", path, recursive), this); - - // Common case. - if (!sizeof(file_locks) && !sizeof(prefix_locks)) { - TRACE_LEAVE ("Got no locks"); - return 0; - } - - mixed auth_user = authenticated_user_id (path, id); - path = resource_id (path, id); - - if (DAVLock lock = - file_locks[path] && file_locks[path][auth_user] || - prefix_locks[path] && prefix_locks[path][auth_user]) { - TRACE_LEAVE(sprintf("Found own lock %O.", lock->locktoken)); - return lock; - } - - LockFlag shared; - - if (mapping(mixed:DAVLock) locks = file_locks[path]) { - foreach(locks;; DAVLock lock) { - if (lock->lockscope == "DAV:exclusive") { - TRACE_LEAVE(sprintf("Found other user's exclusive lock %O.", - lock->locktoken)); - return LOCK_EXCL_AT; - } - shared = LOCK_SHARED_AT; - break; - } - } - - foreach(prefix_locks; - string prefix; mapping(mixed:DAVLock) locks) { - if (has_prefix(path, prefix)) { - if (DAVLock lock = locks[auth_user]) { - SIMPLE_TRACE_LEAVE ("Found own lock %O on %O.", lock->locktoken, prefix); - return lock; - } - if (!shared) - // If we've found a shared lock then we won't find an - // exclusive one anywhere else. - foreach(locks;; DAVLock lock) { - if (lock->lockscope == "DAV:exclusive") { - TRACE_LEAVE(sprintf("Found other user's exclusive lock %O.", - lock->locktoken)); - return LOCK_EXCL_AT; - } - shared = LOCK_SHARED_AT; - break; - } - } - } - - if (!recursive) { - SIMPLE_TRACE_LEAVE("Returning %O.", shared); - return shared; - } - - int(0..1) locked_by_auth_user; - - // We want to know if there are any locks with @[path] as prefix - // that apply to us. - LOOP_OVER_BOTH (string prefix, mapping(mixed:DAVLock) locks, { - if (has_prefix(prefix, path)) { - if (locks[auth_user]) - locked_by_auth_user = 1; - else - foreach(locks;; DAVLock lock) { - if (lock->lockscope == "DAV:exclusive") { - TRACE_LEAVE(sprintf("Found other user's exclusive lock %O.", - lock->locktoken)); - return LOCK_EXCL_BELOW; - } - if (!shared) shared = LOCK_SHARED_BELOW; - break; - } - } - }); - - SIMPLE_TRACE_LEAVE("Returning %O.", locked_by_auth_user ? LOCK_OWN_BELOW : shared); - return locked_by_auth_user ? LOCK_OWN_BELOW : shared; - } - - //! Register @[lock] on the path @[path] under the assumption that - //! there is no other lock already that conflicts with this one, i.e. - //! that @expr{check_locks(path,lock->recursive,id)@} would return - //! @expr{LOCK_NONE@} if @expr{lock->lockscope@} is - //! @expr{"DAV:exclusive"@}, or @expr{< LOCK_OWN_BELOW@} if - //! @expr{lock->lockscope@} is @expr{"DAV:shared"@}. - //! - //! This function is only provided as a helper to call from - //! @[lock_file] if the default lock implementation is to be used. - //! - //! @param path - //! Normalized path (below the filesystem location) that the lock - //! applies to. - //! - //! @param lock - //! The lock to register. - //! - //! @note - //! The default implementation only handles the @expr{"DAV:write"@} - //! lock type. It uses @[resource_id] to map paths to unique resources - //! and @[authenticated_user_id] to tell users apart. - protected void register_lock(string path, DAVLock lock, RequestID id) - { - TRACE_ENTER(sprintf("register_lock(%O, lock(%O), X).", path, lock->locktoken), - this); - ASSERT_IF_DEBUG (lock->locktype == "DAV:write"); - mixed auth_user = authenticated_user_id (path, id); - path = resource_id (path, id); - if (lock->recursive) { - if (prefix_locks[path]) { - prefix_locks[path][auth_user] = lock; - } else { - prefix_locks[path] = ([ auth_user:lock ]); - } - } else { - if (file_locks[path]) { - file_locks[path][auth_user] = lock; - } else { - file_locks[path] = ([ auth_user:lock ]); - } - } - TRACE_LEAVE("Ok."); - } - - //! Unregister @[lock] that currently is locking the resource at - //! @[path]. It's assumed that the lock is registered for exactly that - //! path. - //! - //! This function is only provided as a helper to call from - //! @[unlock_file] if the default lock implementation is to be used. - //! - //! @param path - //! Normalized path (below the filesystem location) that the lock - //! applies to. - //! - //! @param lock - //! The lock to unregister. (It must not be changed or destructed.) - //! - //! @param id - //! The request id may have the value @expr{0@} (zero) if called - //! by @[Configuration()->expire_locks()]. - protected void unregister_lock (string path, DAVLock lock, - RequestID|int(0..0) id) - { - TRACE_ENTER(sprintf("unregister_lock(%O, lock(%O), X).", path, lock->locktoken), - this); - mixed auth_user = id && authenticated_user_id (path, id); - path = resource_id (path, id); - DAVLock removed_lock; - if (lock->recursive) { - if (id) { - removed_lock = m_delete(prefix_locks[path], auth_user); - } else { - foreach(prefix_locks[path]; mixed user; DAVLock l) { - if (l == lock) { - removed_lock = m_delete(prefix_locks[path], user); - } - } - } - if (!sizeof (prefix_locks[path])) m_delete (prefix_locks, path); - } - else if (file_locks[path]) { - if (id) { - removed_lock = m_delete (file_locks[path], auth_user); - } else { - foreach(file_locks[path]; mixed user; DAVLock l) { - if (l == lock) { - removed_lock = m_delete(file_locks[path], user); - } - } - } - if (!sizeof (file_locks[path])) m_delete (file_locks, path); - } - ASSERT_IF_DEBUG (lock /*%O*/ == removed_lock /*%O*/, lock, removed_lock); - TRACE_LEAVE("Ok."); - return 0; - } - - //! Register @[lock] on the path @[path] under the assumption that - //! there is no other lock already that conflicts with this one, i.e. - //! that @expr{check_locks(path,lock->recursive,id)@} would return - //! @expr{LOCK_NONE@} if @expr{lock->lockscope@} is - //! @expr{"DAV:exclusive"@}, or @expr{<= LOCK_SHARED_AT@} if - //! @expr{lock->lockscope@} is @expr{"DAV:shared"@}. - //! - //! The implementation must at least support the @expr{"DAV:write"@} - //! lock type (RFC 2518, section 7). Briefly: An exclusive lock on a - //! file prohibits other users from changing its content. An exclusive - //! lock on a directory (aka collection) prohibits other users from - //! adding or removing files or directories in it. An exclusive lock - //! on a file or directory prohibits other users from setting or - //! deleting any of its properties. A shared lock prohibits users - //! without locks to do any of this, and it prohibits other users from - //! obtaining an exclusive lock. A resource that doesn't exist can be - //! locked, provided the directory it would be in exists (relaxed in - //! RFC 2518Bis (working draft)). The default implementation fulfills - //! these criteria. - //! - //! It's up to @[find_file] et al to actually check that the necessary - //! locks are held. It can preferably use @[write_access] for that, - //! which has a default implementation for checking - //! @expr{"DAV:write"@} locks. - //! - //! @param path - //! Normalized path (below the filesystem location) that the lock - //! applies to. - //! - //! @param lock - //! The lock to register. - //! - //! @returns - //! Returns @expr{0@} if the lock is successfully installed or if - //! locking isn't used. Returns a status mapping if an error - //! occurred. - //! - //! @note - //! To use the default lock implementation, call @[register_lock] - //! from this function. - mapping(string:mixed) lock_file(string path, - DAVLock lock, - RequestID id) - { - return 0; - } - - //! Remove @[lock] that currently is locking the resource at @[path]. - //! It's assumed that the lock is registered for exactly that path. - //! - //! @param path - //! Normalized path (below the filesystem location) that the lock - //! applies to. - //! - //! @param lock - //! The lock to unregister. (It must not be changed or destructed.) - //! - //! @param id - //! @mixed - //! @type RequestID - //! The request that attempted to unlock the lock. The function - //! should do the normal access checks before unlocking the lock - //! in this case. - //! - //! @type int(0..0) - //! The lock is unlocked internally by the server (typically due - //! to a timeout) and should be carried out without any access - //! checks. The function must succeed and return zero in this - //! case. - //! @endmixed - //! - //! @returns - //! Returns a status mapping on any error, zero otherwise. - //! - //! @note - //! To use the default lock implementation, call @[unregister_lock] - //! from this function. - mapping(string:mixed) unlock_file (string path, - DAVLock lock, - RequestID|int(0..0) id); - - //! Checks that the conditions specified by the WebDAV @expr{"If"@} - //! header are fulfilled on the given path (RFC 2518 9.4). This means - //! that locks are checked as necessary using @[check_locks]. - //! - //! WARNING: This function has some design issues and will very likely - //! get a different interface. Compatibility is NOT guaranteed. - //! - //! @param path - //! Path (below the filesystem location) that the lock applies to. - //! - //! @param recursive - //! If @expr{1@} also check write access recursively under @[path]. - //! - //! @returns - //! Returns @expr{0@} (zero) on success, a status mapping on - //! failure, or @expr{1@} if @[recursive] is set and write access is - //! allowed on this level but maybe not somewhere below. The caller - //! should in the last case do the operation on this level if - //! possible and then handle each member in the directory - //! recursively with @[write_access] etc. - mapping(string:mixed)|int(0..1) check_if_header(string relative_path, - int(0..1) recursive, - RequestID id) - { - SIMPLE_TRACE_ENTER(this, "Checking \"If\" header for %O", - relative_path); - - int/*LockFlag*/|DAVLock lock = check_locks(relative_path, recursive, id); - - int(0..1) got_sublocks; - if (lock && intp(lock)) { - if (lock & 1) { - TRACE_LEAVE("Locked by other user."); - return Roxen.http_status(Protocols.HTTP.DAV_LOCKED); - } - else if (recursive) - // This is set for LOCK_OWN_BELOW too since it might be - // necessary to come back here and check the If header for - // those locks. - got_sublocks = 1; - } - - string path = relative_path; - if (!has_suffix (path, "/")) path += "/"; // get_if_data always adds a "/". - path = query_location() + path; // No need for fancy combine_path stuff here. - - mapping(string:array(array(array(string)))) if_data = id->get_if_data(); - array(array(array(string))) condition; - if (!if_data || !sizeof(condition = if_data[path] || if_data[0])) { - if (lock) { - TRACE_LEAVE("Locked, no if header."); - return Roxen.http_status(Protocols.HTTP.DAV_LOCKED); - } - SIMPLE_TRACE_LEAVE("No lock and no if header - ok%s.", - got_sublocks ? " (this level only)" : ""); - return got_sublocks; // No condition and no lock -- Ok. - } - - string|int(-1..0) etag; - - int(0..1) locked_fail = !!lock; - next_condition: - foreach(condition, array(array(string)) sub_cond) { - SIMPLE_TRACE_ENTER(this, - "Trying condition ( %{%s:%O %})...", sub_cond); - int negate; - DAVLock locked = lock; - foreach(sub_cond, array(string) token) { - switch(token[0]) { - case "not": - negate = !negate; - break; - case "etag": - if (!etag) { - // Get the etag for this resource (if any). - // FIXME: We only support straight strings as etag properties. - if (!stringp(etag = query_property(relative_path, - "DAV:getetag", id))) { - etag = -1; - } - } - if (etag != token[1]) { - // No etag available for this resource, or mismatch. - if (!negate) { - TRACE_LEAVE("Etag mismatch."); - continue next_condition; - } - } else if (negate) { - // Etag match with negated expression. - TRACE_LEAVE("Matched negated etag."); - continue next_condition; - } - negate = 0; - break; - case "key": - // The user has specified a key, so don't fail with DAV_LOCKED. - locked_fail = 0; - if (negate) { - if (lock && lock->locktoken == token[1]) { - TRACE_LEAVE("Matched negated lock."); - continue next_condition; // Fail. - } - } else if (!lock || lock->locktoken != token[1]) { - // Lock mismatch. - TRACE_LEAVE("Lock mismatch."); - continue next_condition; // Fail. - } else { - locked = 0; - } - negate = 0; - break; - } - } - if (!locked) { - TRACE_LEAVE("Found match."); - SIMPLE_TRACE_LEAVE("Ok%s.", - got_sublocks ? " (this level only)" : ""); - return got_sublocks; // Found matching sub-condition. - } - SIMPLE_TRACE_LEAVE("Conditional ok, but still locked (locktoken: %O).", - lock->locktoken); - locked_fail = 1; - } - - TRACE_LEAVE("Failed."); - return Roxen.http_status(locked_fail ? - Protocols.HTTP.DAV_LOCKED : - Protocols.HTTP.HTTP_PRECOND_FAILED); - } - - //! Used by some default implementations to check if we may perform a - //! write access to @[path]. It should at least call - //! @[check_if_header] to check DAV locks. It takes the same arguments - //! and has the same return value as that function. - //! - //! WARNING: This function has some design issues and will very likely - //! get a different interface. Compatibility is NOT guaranteed. - //! - //! A filesystem module should typically put all needed write access - //! checks here and then use this from @[find_file()], - //! @[delete_file()] etc. - protected mapping(string:mixed)|int(0..1) write_access(string relative_path, - int(0..1) recursive, - RequestID id) - { - return check_if_header (relative_path, recursive, id); - } - - mapping(string:mixed)|int(-1..0)|Stdio.File find_file(string path, - RequestID id); - - //! Used by the default @[recurse_delete_files] implementation to - //! delete a file or an empty directory. - //! - //! @returns - //! Returns a 2xx series status mapping on success (typically 204 No - //! Content). Returns 0 if the file doesn't exist. Returns an - //! appropriate status mapping for any other error. - //! - //! @note - //! The default implementation falls back to @[find_file()]. - protected mapping(string:mixed) delete_file(string path, RequestID id) - { - // Fall back to find_file(). - RequestID tmp_id = id->clone_me(); - tmp_id->not_query = query_location() + path; - tmp_id->method = "DELETE"; - // FIXME: Logging? - return find_file(path, tmp_id) || - tmp_id->misc->error_code && Roxen.http_status (tmp_id->misc->error_code); - } - - //! Delete @[path] recursively. - //! - //! The default implementation handles the recursion and calls - //! @[delete_file] for each file and empty directory. - //! - //! @returns - //! Returns a 2xx series status mapping on success (typically 204 No - //! Content). Returns 0 if the file doesn't exist. Returns an - //! appropriate status mapping for any other error. That includes an - //! empty mapping in case some subparts couldn't be deleted, to - //! signify a 207 Multi-Status response using the info in - //! @[id->get_multi_status()]. - mapping(string:mixed) recurse_delete_files(string path, - RequestID id, - void|MultiStatus.Prefixed stat) - { - SIMPLE_TRACE_ENTER (this, "Deleting %O recursively", path); - if (!stat) - stat = id->get_multi_status()->prefix (id->url_base() + - query_location()[1..]); - - Stat st = stat_file(path, id); - if (!st) { - SIMPLE_TRACE_LEAVE ("No such file or directory"); - return 0; - } - - mapping(string:mixed) recurse (string path, Stat st) - { - // Note: Already got an extra TRACE_ENTER level on entry here. - - if (st->isdir) { - // RFC 2518 8.6.2 - // The DELETE operation on a collection MUST act as if a - // "Depth: infinity" header was used on it. - int fail; - if (!has_suffix(path, "/")) path += "/"; - foreach(find_dir(path, id) || ({}), string fname) { - fname = path + fname; - if (Stat sub_stat = stat_file (fname, id)) { - SIMPLE_TRACE_ENTER (this, "Deleting %O", fname); - if (mapping(string:mixed) sub_res = recurse(fname, sub_stat)) { - // RFC 2518 8.6.2 - // Additionally 204 (No Content) errors SHOULD NOT be returned - // in the 207 (Multi-Status). The reason for this prohibition - // is that 204 (No Content) is the default success code. - if (sizeof (sub_res) && sub_res->error != 204) { - stat->add_status(fname, sub_res->error, sub_res->rettext); - } - if (!sizeof (sub_res) || sub_res->error >= 300) fail = 1; - } - } - } - if (fail) { - SIMPLE_TRACE_LEAVE ("Partial failure"); - return ([]); - } - } - - SIMPLE_TRACE_LEAVE (""); - return delete_file (path, id); - }; - - return recurse(path, st) || Roxen.http_status(204); - } - - //! Make a new collection (aka directory) at @[path]. - //! - //! @returns - //! Returns a 2xx series status on success (typically 201 Created). - //! Returns @expr{0@} (zero) if there's no directory to create the - //! new one in. Returns other result mappings on failure. - mapping(string:mixed) make_collection(string path, RequestID id) - { - // Fall back to find_file(). - RequestID tmp_id = id->clone_me(); - tmp_id->not_query = query_location() + path; - tmp_id->method = "MKCOL"; - // FIXME: Logging? - return find_file(path, tmp_id); - } - - //! Used by the default @[copy_collection] implementation to copy all - //! properties at @[source] to @[destination]. - //! - //! @param source - //! Source path below the filesystem location. - //! - //! @param destination - //! Destination path below the filesystem location. - //! - //! @param behavior - //! Specifies how to copy properties. See the @[PropertyBehavior] - //! type for details. - //! - //! @returns - //! @expr{0@} (zero) on success or an appropriate status mapping for - //! any error. - protected mapping(string:mixed) copy_properties( - string source, string destination, PropertyBehavior behavior, RequestID id) - { - SIMPLE_TRACE_ENTER(this, "copy_properties(%O, %O, %O, %O)", - source, destination, behavior, id); - PropertySet source_properties = query_property_set(source, id); - PropertySet destination_properties = query_property_set(destination, id); - - multiset(string) property_set = source_properties->query_all_properties(); - mapping(string:mixed) res; - foreach(property_set; string property_name;) { - SIMPLE_TRACE_ENTER(this, "Copying the property %O.", property_name); - string|array(Parser.XML.Tree.SimpleNode)|mapping(string:mixed) source_val = - source_properties->query_property(property_name); - if (mappingp(source_val)) { - TRACE_LEAVE("Reading of property failed. Skipped."); - continue; - } - string|array(Parser.XML.Tree.SimpleNode)|mapping(string:mixed) dest_val = - destination_properties->query_property(property_name); - if (dest_val == source_val) { - TRACE_LEAVE("Destination already has the correct value."); - continue; - } - mapping(string:mixed) subres = - destination_properties->set_property(property_name, source_val); - if (!behavior) { - TRACE_LEAVE("Omit verify."); - continue; - } - if ((intp(behavior) || behavior[property_name]) && (subres->error < 300)) { - // FIXME: Check that if the property was live in source, - // it is still live in destination. - // This is likely already so, since we're in the same module. - } - if ((subres->error < 300) || - (subres->error == Protocols.HTTP.HTTP_CONFLICT)) { - // Ok, or read-only property. - TRACE_LEAVE("Copy ok or read-only property."); - continue; - } - if (!subres) { - // Copy failed, but attempt to copy the rest. - res = Roxen.http_status(Protocols.HTTP.HTTP_PRECOND_FAILED); - } - TRACE_LEAVE("Copy failed."); - } - TRACE_LEAVE(res?"Failed.":"Ok."); - return res; - } - - //! Used by the default @[recurse_copy_files] and @[move_collection] - //! to copy a collection (aka directory) and its properties but not - //! its contents. - //! - //! @param source - //! Source path below the filesystem location. - //! - //! @param destination - //! Destination path below the filesystem location. - //! - //! @param behavior - //! Specifies how to copy properties. See the @[PropertyBehavior] - //! type for details. - //! - //! @param overwrite - //! Specifies how to handle the situation if the destination already - //! exists. See the @[Overwrite] type for details. - //! - //! @param result - //! A @[MultiStatus.Prefixed] to collect status mappings if some - //! subparts fail. It's prefixed with the URL to the filesystem - //! location. - //! - //! @returns - //! Returns a 2xx series status mapping on success (typically 201 - //! Created if the destination didn't exist before, or 204 No - //! Content otherwise). Returns 0 if the source doesn't exist. - //! Returns an appropriate status mapping for any other error. That - //! includes an empty mapping in case there's a failure on some - //! subpart or at the destination, to signify a 207 Multi-Status - //! response using the info in @[id->get_multi_status()]. - protected mapping(string:mixed) copy_collection( - string source, string destination, PropertyBehavior behavior, - Overwrite overwrite, MultiStatus.Prefixed result, RequestID id) - { - SIMPLE_TRACE_ENTER(this, "copy_collection(%O, %O, %O, %O, %O, %O).", - source, destination, behavior, overwrite, result, id); - Stat st = stat_file(destination, id); - if (st) { - // Destination exists. Check the overwrite header. - switch(overwrite) { - case DO_OVERWRITE: - // RFC 2518 8.8.4 - // If a resource exists at the destination, and the Overwrite - // header is "T" then prior to performing the copy the server - // MUST perform a DELETE with "Depth: infinity" on the - // destination resource. - TRACE_ENTER("Destination exists and overwrite is on.", this); - mapping(string:mixed) res = recurse_delete_files(destination, id, result); - if (res && (!sizeof (res) || res->error >= 300)) { - // Failed to delete something. - TRACE_LEAVE("Deletion failed."); - TRACE_LEAVE("Copy collection failed."); - // RFC 2518 9.6 says: - // - // If a COPY or MOVE is not performed due to the value of - // the Overwrite header, the method MUST fail with a 412 - // (Precondition Failed) status code. - // - // That can perhaps be interpreted as that we should return - // 412 here. But otoh, in RFC 2518 8.8.5 COPY status codes: - // - // 412 (Precondition Failed) - /.../ the Overwrite header - // is "F" and the state of the destination resource is - // non-null. - // - // That clearly doesn't include this case. Also, common sense - // says that the error from the failed delete is more useful - // to the client. - #if 0 - return Roxen.http_status(Protocols.HTTP.HTTP_PRECOND_FAILED); - #else - if (sizeof (res)) { - // RFC 2518 8.8.3: - // If an error in executing the COPY method occurs with a - // resource other than the resource identified in the - // Request-URI then the response MUST be a 207 - // (Multi-Status). - // - // So if the failure was on the root destination resource we - // have to convert it to a multi-status. - result->add_status (destination, res->error, res->rettext); - } - return ([]); - #endif - } - TRACE_LEAVE("Deletion ok."); - break; - case NEVER_OVERWRITE: - TRACE_LEAVE("Destination already exists."); - return Roxen.http_status(Protocols.HTTP.HTTP_PRECOND_FAILED); - case MAYBE_OVERWRITE: - // No overwrite header. - // Be nice, and fail only if we don't already have a collection. - if (st->isdir) { - TRACE_LEAVE("Destination exists and is a directory."); - return copy_properties(source, destination, behavior, id); - } - TRACE_LEAVE("Destination exists and is not a directory."); - return Roxen.http_status(Protocols.HTTP.HTTP_PRECOND_FAILED); - } - } - // Create the new collection. - TRACE_LEAVE("Make a new collection."); - mapping(string:mixed) res = make_collection(destination, id); - if (res && res->error >= 300) return res; - return copy_properties(source, destination, behavior, id) || res; - } - - //! Used by the default @[recurse_copy_files] to copy a single file - //! along with its properties. - //! - //! @param source - //! Source path below the filesystem location. - //! - //! @param destination - //! Destination path below the filesystem location. - //! - //! @param behavior - //! Specifies how to copy properties. See the @[PropertyBehavior] - //! type for details. - //! - //! @param overwrite - //! Specifies how to handle the situation if the destination already - //! exists. See the @[Overwrite] type for details. - //! - //! @returns - //! Returns a 2xx series status mapping on success (typically 201 - //! Created if the destination didn't exist before, or 204 No - //! Content otherwise). Returns 0 if the source doesn't exist. - //! Returns an appropriate status mapping for any other error. - protected mapping(string:mixed) copy_file(string source, string destination, - PropertyBehavior behavior, - Overwrite overwrite, RequestID id) - { - SIMPLE_TRACE_ENTER(this, "copy_file(%O, %O, %O, %O, %O)\n", - source, destination, behavior, overwrite, id); - TRACE_LEAVE("Not implemented."); - return Roxen.http_status (Protocols.HTTP.HTTP_NOT_IMPL); - } - - //! Copy a resource recursively from @[source] to @[destination]. - //! - //! @param source - //! Source path below the filesystem location. - //! - //! @param destination - //! Destination path below the filesystem location. - //! - //! @param behavior - //! Specifies how to copy properties. See the @[PropertyBehavior] - //! type for details. - //! - //! @param overwrite - //! Specifies how to handle the situation if the destination already - //! exists. See the @[Overwrite] type for details. - //! - //! @returns - //! Returns a 2xx series status mapping on success (typically 201 - //! Created if the destination didn't exist before, or 204 No - //! Content otherwise). Returns 0 if the source doesn't exist. - //! Returns an appropriate status mapping for any other error. That - //! includes an empty mapping in case there's a failure on some - //! subpart or at the destination, to signify a 207 Multi-Status - //! response using the info in @[id->get_multi_status()]. - mapping(string:mixed) recurse_copy_files(string source, string destination, - PropertyBehavior behavior, - Overwrite overwrite, RequestID id) - { - SIMPLE_TRACE_ENTER(this, "Recursive copy from %O to %O (%s)", - source, destination, - overwrite == DO_OVERWRITE ? "replace" : - overwrite == NEVER_OVERWRITE ? "no overwrite" : - "overlay"); - string src_tmp = has_suffix(source, "/")?source:(source+"/"); - string dst_tmp = has_suffix(destination, "/")?destination:(destination+"/"); - if ((src_tmp == dst_tmp) || - has_prefix(src_tmp, dst_tmp) || - has_prefix(dst_tmp, src_tmp)) { - TRACE_LEAVE("Source and destination overlap."); - return Roxen.http_status(403, "Source and destination overlap."); - } - - string prefix = map(query_location()[1..]/"/", Roxen.http_encode_url)*"/"; - MultiStatus.Prefixed result = - id->get_multi_status()->prefix (id->url_base() + prefix); - - mapping(string:mixed) recurse(string source, string destination) { - // Note: Already got an extra TRACE_ENTER level on entry here. - - Stat st = stat_file(source, id); - if (!st) { - TRACE_LEAVE("Source not found."); - return 0; - } - // FIXME: Check destination? - if (st->isdir) { - mapping(string:mixed) res = - copy_collection(source, destination, behavior, overwrite, result, id); - if (res && (!sizeof (res) || res->error >= 300)) { - // RFC 2518 8.8.3 and 8.8.8 (error minimization). - TRACE_LEAVE("Copy of collection failed."); - return res; - } - foreach(find_dir(source, id), string filename) { - string subsrc = combine_path_unix(source, filename); - string subdst = combine_path_unix(destination, filename); - SIMPLE_TRACE_ENTER(this, "Copy from %O to %O\n", subsrc, subdst); - mapping(string:mixed) sub_res = recurse(subsrc, subdst); - if (sub_res && !(<0, 201, 204>)[sub_res->error]) { - result->add_status(subdst, sub_res->error, sub_res->rettext); - } - } - TRACE_LEAVE(""); - return res; - } else { - TRACE_LEAVE(""); - return copy_file(source, destination, behavior, overwrite, id); - } - }; - - int start_ms_size = id->multi_status_size(); - mapping(string:mixed) res = recurse (source, destination); - if (res && res->error != 204 && res->error != 201) - return res; - else if (id->multi_status_size() != start_ms_size) - return ([]); - else - return res; - } - - //! Used by the default @[recurse_move_files] to move a file (and not - //! a directory) from @[source] to @[destination]. - //! - //! The default implementation tries to call @[find_file] with the - //! MOVE method. If that returns 501 Not Implemented then it copies - //! the file and deletes the source afterwards. - //! - //! @param source - //! Source path below the filesystem location. - //! - //! @param destination - //! Destination path below the filesystem location. - //! - //! @param behavior - //! Specifies how to move properties. See the @[PropertyBehavior] - //! type for details. - //! - //! @param overwrite - //! Specifies how to handle the situation if the destination already - //! exists. See the @[Overwrite] type for details. - //! - //! @returns - //! Returns a 2xx series status mapping on success (typically 201 - //! Created if the destination didn't exist before, or 204 No - //! Content otherwise). Returns 0 if the source doesn't exist. - //! Returns an appropriate status mapping for any other error. That - //! includes an empty mapping in case there's a failure on some - //! subpart or at the destination, to signify a 207 Multi-Status - //! response using the info in @[id->get_multi_status()]. - protected mapping(string:mixed) move_file(string source, string destination, - PropertyBehavior behavior, - Overwrite overwrite, RequestID id) - { - // Fall back to find_file(). - RequestID tmp_id = id->clone_me(); - tmp_id->not_query = query_location() + source; - tmp_id->misc["new-uri"] = query_location() + destination; - tmp_id->request_headers->destination = - id->url_base() + query_location()[1..] + destination; - tmp_id->method = "MOVE"; - mapping(string:mixed) res = find_file(source, tmp_id); - if (!res || res->error != 501) return res; - // Not implemented. Fall back to COPY + DELETE. - res = copy_file(source, destination, behavior, overwrite, id); - if (res && res->error >= 300) { - // Copy failed. - return res; - } - return delete_file(source, id); - } - - //! Used by the default @[recurse_move_files] to move a collection - //! (aka directory) and all its content from @[source] to - //! @[destination]. - //! - //! The default implementation tries to call @[find_file] with the - //! MOVE method. If that returns 501 Not Implemented then it copies - //! the collection, moves each directory entry recursively, and - //! deletes the source collection afterwards. - //! - //! @param source - //! Source path below the filesystem location. - //! - //! @param destination - //! Destination path below the filesystem location. - //! - //! @param behavior - //! Specifies how to move properties. See the @[PropertyBehavior] - //! type for details. - //! - //! @param overwrite - //! Specifies how to handle the situation if the destination already - //! exists. See the @[Overwrite] type for details. - //! - //! @returns - //! Returns a 2xx series status mapping on success (typically 201 - //! Created if the destination didn't exist before, or 204 No - //! Content otherwise). Returns 0 if the source doesn't exist. - //! Returns an appropriate status mapping for any other error. That - //! includes an empty mapping in case there's a failure on some - //! subpart or at the destination, to signify a 207 Multi-Status - //! response using the info in @[id->get_multi_status()]. - //! - //! @note - //! The function must be prepared to recurse to check DAV locks - //! properly. - protected mapping(string:mixed) move_collection( - string source, string destination, PropertyBehavior behavior, - Overwrite overwrite, RequestID id) - { - // Fall back to find_file(). - RequestID tmp_id = id->clone_me(); - tmp_id->not_query = query_location() + source; - tmp_id->misc["new-uri"] = query_location() + destination; - tmp_id->request_headers->destination = - id->url_base() + query_location()[1..] + destination; - tmp_id->method = "MOVE"; - mapping(string:mixed) res = find_file(source, tmp_id); - if (!res || res->error != 501) return res; - // Not implemented. Fall back to COPY + DELETE. - string prefix = map(query_location()[1..]/"/", Roxen.http_encode_url)*"/"; - MultiStatus.Prefixed result = - id->get_multi_status()->prefix (id->url_base() + prefix); - res = copy_collection(source, destination, behavior, overwrite, result, id); - if (res && (res->error >= 300 || !sizeof(res))) { - // Copy failed. - return res; - } - int fail; - foreach(find_dir(source, id), string filename) { - string subsrc = combine_path_unix(source, filename); - string subdst = combine_path_unix(destination, filename); - SIMPLE_TRACE_ENTER(this, "Recursive move from %O to %O\n", - subsrc, subdst); - if (mapping(string:mixed) sub_res = - recurse_move_files(subsrc, subdst, behavior, overwrite, id)) { - if (!(<0, 201, 204>)[sub_res->error]) { - result->add_status(subdst, sub_res->error, sub_res->rettext); - } - if (!sizeof (sub_res) || sub_res->error >= 300) { - // Failed to move some content. - fail = 1; - } - } - } - if (fail) return ([]); - return delete_file(source, id); - } - - //! Move a resource from @[source] to @[destination]. - //! - //! @param source - //! Source path below the filesystem location. - //! - //! @param destination - //! Destination path below the filesystem location. - //! - //! @param behavior - //! Specifies how to move properties. See the @[PropertyBehavior] - //! type for details. - //! - //! @param overwrite - //! Specifies how to handle the situation if the destination already - //! exists. See the @[Overwrite] type for details. - //! - //! @returns - //! Returns a 2xx series status mapping on success (typically 201 - //! Created if the destination didn't exist before, or 204 No - //! Content otherwise). Returns 0 if the source doesn't exist. - //! Returns an appropriate status mapping for any other error. That - //! includes an empty mapping in case there's a failure on some - //! subpart or at the destination, to signify a 207 Multi-Status - //! response using the info in @[id->get_multi_status()]. - mapping(string:mixed) recurse_move_files(string source, string destination, - PropertyBehavior behavior, - Overwrite overwrite, RequestID id) - { - Stat st = stat_file(source, id); - if (!st) return 0; - - if (st->isdir) { - return move_collection(source, destination, behavior, overwrite, id); - } - return move_file(source, destination, behavior, overwrite, id); - } - - string real_file(string f, RequestID id){} - - void add_api_function( string name, function f, void|array(string) types) - { - _api_functions[name] = ({ f, types }); - } - - mapping api_functions() - { - return _api_functions; - } - - #if ROXEN_COMPAT <= 1.4 - mapping(string:function) query_tag_callers() - //! Compat - { - mapping(string:function) m = ([]); - foreach(glob("tag_*", indices( this_object())), string q) - if(functionp( this_object()[q] )) - m[replace(q[4..], "_", "-")] = this_object()[q]; - return m; - } - - mapping(string:function) query_container_callers() - //! Compat - { - mapping(string:function) m = ([]); - foreach(glob("container_*", indices( this_object())), string q) - if(functionp( this_object()[q] )) - m[replace(q[10..], "_", "-")] = this_object()[q]; - return m; - } - #endif - - mapping(string:array(int|function)) query_simpletag_callers() - { - mapping(string:array(int|function)) m = ([]); - foreach(glob("simpletag_*", indices(this_object())), string q) - if(functionp(this_object()[q])) - m[replace(q[10..],"_","-")] = - ({ intp (this_object()[q + "_flags"]) && this_object()[q + "_flags"], - this_object()[q] }); - return m; - } - - mapping(string:array(int|function)) query_simple_pi_tag_callers() - { - mapping(string:array(int|function)) m = ([]); - foreach (glob ("simple_pi_tag_*", indices (this_object())), string q) - if (functionp (this_object()[q])) - m[replace (q[sizeof ("simple_pi_tag_")..], "_", "-")] = - ({(intp (this_object()[q + "_flags"]) && this_object()[q + "_flags"]) | - RXML.FLAG_PROC_INSTR, this_object()[q]}); - return m; - } - - RXML.TagSet query_tag_set() - { - if (!module_tag_set) { - array(function|program|object) tags = - filter (rows (this_object(), - glob ("Tag*", indices (this_object()))), - lambda(mixed x) { return functionp(x)||programp(x); }); - for (int i = 0; i < sizeof (tags); i++) - if (programp (tags[i])) - if (!tags[i]->is_RXML_Tag) tags[i] = 0; - else tags[i] = tags[i](); - else { - tags[i] = tags[i](); - // Bogosity: The check is really a little too late here.. - if (!tags[i]->is_RXML_Tag) tags[i] = 0; - } - tags -= ({0}); - module_tag_set = - (this_object()->ModuleTagSet || RXML.TagSet) (this_object(), "", tags); - } - return module_tag_set; - } - - mixed get_value_from_file(string path, string index, void|string pre) - { - Stdio.File file=Stdio.File(); - if(!file->open(path,"r")) return 0; - if(has_suffix(index, "()")) - index = index[..sizeof(index) - 3]; - - // Pass path to original file so that include statements for local files - // work correctly. - return compile_string((pre || "") + file->read(), path)[index]; - } - - #if constant(roxen.FSGarbWrapper) - //! Register a filesystem path for automatic garbage collection. - //! - //! @param path - //! Path in the real filesystem to garbage collect. - //! - //! @param max_age - //! Maximum allowed age in seconds for files. - //! - //! @param max_size - //! Maximum total size in bytes for all files under the path. - //! Zero to disable the limit. - //! - //! @param max_files - //! Maximum number of files under the path. - //! Zero to disable the limit. - //! - //! @returns - //! Returns a roxen.FSGarbWrapper object. The garbage collector - //! will be removed when this object is destructed (eg via - //! refcount-garb). - roxen.FSGarbWrapper register_fsgarb(string path, int max_age, - int|void max_size, int|void max_files, - string|void quarantine) - { - return roxen.register_fsgarb(module_identifier(), path, max_age, - max_size, max_files, quarantine); - } - #endif - - private mapping __my_tables = ([]); - - array(mapping(string:mixed)) sql_query( string query, mixed ... args ) - //! Do a SQL-query using @[get_my_sql], the table names in the query - //! should be written as &table; instead of table. As an example, if - //! the tables 'meta' and 'data' have been created with create_tables - //! or get_my_table, this query will work: - //! - //! SELECT &meta;.id AS id, &data;.data as DATA - //! FROM &data;, &meta; WHERE &my.meta;.xsize=200 - //! - { - return get_my_sql()->query( replace( query, __my_tables ), @args ); - } - - object sql_big_query( string query, mixed ... args ) - //! Identical to @[sql_query], but the @[Sql.Sql()->big_query] method - //! will be used instead of the @[Sql.Sql()->query] method. - { - return get_my_sql()->big_query( replace( query, __my_tables ), @args ); - } - - array(mapping(string:mixed)) sql_query_ro( string query, mixed ... args ) - //! Do a read-only SQL-query using @[get_my_sql], the table names in the query - //! should be written as &table; instead of table. As an example, if - //! the tables 'meta' and 'data' have been created with create_tables - //! or get_my_table, this query will work: - //! - //! SELECT &meta;.id AS id, &data;.data as DATA - //! FROM &data;, &meta; WHERE &my.meta;.xsize=200 - //! - { - return get_my_sql(1)->query( replace( query, __my_tables ), @args ); - } - - object sql_big_query_ro( string query, mixed ... args ) - //! Identical to @[sql_query_ro], but the @[Sql.Sql()->big_query] method - //! will be used instead of the @[Sql.Sql()->query] method. - { - return get_my_sql(1)->big_query( replace( query, __my_tables ), @args ); - } - - protected int create_sql_tables( mapping(string:array(string)) definitions, - string|void comment, int|void no_unique_names ) - //! Create multiple tables in one go. See @[get_my_table] - //! Returns the number of tables that were actually created. - { - int ddc; - if( !no_unique_names ) - foreach( indices( definitions ), string t ) - ddc+=get_my_table( t, definitions[t], comment, 1 ); - else - { - Sql.Sql sql = get_my_sql(); - foreach( indices( definitions ), string t ) - { - if( !catch { - sql->query("CREATE TABLE "+t+" ("+definitions[t]*","+")" ); - } ) - ddc++; - DBManager.is_module_table( this_object(), my_db, t, comment ); - } - } - return ddc; - } - - protected string sql_table_exists( string name ) - //! Return the real name of the table 'name' if it exists. - { - if(strlen(name)) - name = "_"+name; - - string res = hash(_my_configuration->name)->digits(36) - + "_" + replace(sname(), ({ "#","-" }), ({ "_","_" })) + name; - - return catch(get_my_sql()->query( "SELECT * FROM "+res+" LIMIT 1" ))?0:res; - } - - - protected string|int get_my_table( string|array(string) name, - void|array(string)|string definition, - string|void comment, int|void flag ) - //! @decl string get_my_table( string name, array(string) types ) - //! @decl string get_my_table( string name, string definition ) - //! @decl string get_my_table( string definition ) - //! @decl string get_my_table( array(string) definition ) - //! - //! Returns the name of a table in the 'shared' database that is - //! unique for this module. It is possible to select another database - //! by using @[set_my_db] before calling this function. - //! - //! You can use @[create_sql_tables] instead of this function if you want - //! to create more than one table in one go. - //! - //! If @[flag] is true, return 1 if a table was created, and 0 otherwise. - //! - //! In the first form, @[name] is the (postfix of) the name of the - //! table, and @[types] is an array of definitions, as an example: - //! - //! - //! @code - //! cache_table = get_my_table( "cache", ({ - //! "id INT UNSIGNED AUTO_INCREMENT", - //! "data BLOB", - //! }) ); - //! @endcode - //! - //! In the second form, the whole table definition is instead sent as - //! a string. The cases where the name is not included (the third and - //! fourth form) is equivalent to the first two cases with the name "" - //! - //! If the table does not exist in the datbase, it is created. - //! - //! @note - //! This function may not be called from create - // - // If it exists, but it's definition is different, the table will be - // altered with a ALTER TABLE call to conform to the definition. This - // might not work if the database the table resides in is not a MySQL - // database (normally it is, but it's possible, using @[set_my_db], - // to change this). - { - string oname; - int ddc; - if( !definition ) - { - definition = name; - oname = name = ""; - } - else if(strlen(name)) - name = "_"+(oname = name); - - Sql.Sql sql = get_my_sql(); - - string res = hash(_my_configuration->name)->digits(36) - + "_" + replace(sname(),({ "#","-" }), ({ "_","_" })) + name; - - if( !sql ) - { - report_error("Failed to get SQL handle, permission denied for "+my_db+"\n"); - return 0; - } - if( arrayp( definition ) ) - definition *= ", "; - - if( catch(sql->query( "SELECT * FROM "+res+" LIMIT 1" )) ) - { - ddc++; - mixed error = - catch - { - get_my_sql()->query( "CREATE TABLE "+res+" ("+definition+")" ); - DBManager.is_module_table( this_object(), my_db, res, - oname+"\0"+comment ); - }; - if( error ) - { - if( strlen( name ) ) - name = " "+name; - report_error( "Failed to create table"+name+": "+ - describe_error( error ) ); - return 0; - } - if( flag ) - { - __my_tables[ "&"+oname+";" ] = res; - return ddc; - } - return __my_tables[ "&"+oname+";" ] = res; - } - // // Update definition if it has changed. - // mixed error = - // catch - // { - // get_my_sql()->query( "ALTER TABLE "+res+" ("+definition+")" ); - // }; - // if( error ) - // { - // if( strlen( name ) ) - // name = " for "+name; - // report_notice( "Failed to update table definition"+name+": "+ - // describe_error( error ) ); - // } - if( flag ) - { - __my_tables[ "&"+oname+";" ] = res; - return ddc; - } - return __my_tables[ "&"+oname+";" ] = res; - } - - protected string my_db = "local"; - - protected void set_my_db( string to ) - //! Select the database in which tables will be created with - //! get_my_table, and also the one that will be returned by - //! @[get_my_sql] - { - my_db = to; - } - - Sql.Sql get_my_sql( int|void read_only, void|string charset ) - //! Return a SQL-object for the database set with @[set_my_db], - //! defaulting to the 'shared' database. If @[read_only] is specified, - //! the database will be opened in read_only mode. @[charset] may be - //! used to specify a charset for the connection if the database - //! supports it. - //! - //! See also @[DBManager.get] - { - return DBManager.cached_get( my_db, _my_configuration, read_only, charset ); - } - - // Callback used by the DB browser, if defined, for custom formatting - // of database fields. - int|string format_db_browser_value (string db_name, string table_name, - string column_name, array(string) col_names, - array(string) col_types, array(string) row, - RequestID id); + Newline at end of file removed.

Roxen.git / server / base_server / module.pike