API:EPrints/Repository

From EPrints Documentation
Revision as of 14:01, 25 February 2010 by Tdb01r (talk | contribs) (parse_xml)
Jump to: navigation, search

EPrints 3 Reference: Directory Structure - Metadata Fields - Repository Configuration - XML Config Files - XML Export Format - EPrints data structure - Core API - Data Objects


API: Core API

Latest Source Code (3.4, 3.3) | Revision Log | Before editing this page please read Pod2Wiki


Contents

NAME

EPrints::Repository - Single connection to the EPrints system

User Comments


DESCRIPTION

This module is not really a session. The name is out of date, but hard to change.

EPrints::Repository represents a connection to the EPrints system. It connects to a single EPrints repository, and the database used by that repository. Thus it has an associated EPrints::Database and EPrints::Repository object.

Each "session" has a "current language". If you are running in a multilingual mode, this is used by the HTML rendering functions to choose what language to return text in.

The "session" object also knows about the current apache connection, if there is one, including the CGI parameters.

If the connection requires a username and password then it can also give access to the EPrints::DataObj::User object representing the user who is causing this request.

The session object also provides many methods for creating XHTML results which can be returned via the web interface.

User Comments


METHODS

User Comments


new

$repository = EPrints::Repository->new( $mode, [$repository_id], [$noise], [$nocheckdb] )

Create a connection to an EPrints repository which provides access to the database and to the repository configuration.

This method can be called in two modes. Setting $mode to 0 means this is a connection via a CGI web page. $repository_id is ignored, instead the value is taken from the "PerlSetVar EPrints_ArchiveID" option in the apache configuration for the current directory.

If this is being called from a command line script, then $mode should be 1, and $repository_id should be the ID of the repository we want to connect to.

$mode : mode = 0 - We are online (CGI script) mode = 1 - We are offline (bin script) $repository_id is repository_id mode = 2 - We are online, but don't create a CGI query (so we

don't consume the data).

$noise is the level of debugging output. 0 - silent 1 - quietish 2 - noisy 3 - debug all SQL statements 4 - debug database connection

Under normal conditions use "0" for online and "1" for offline.

$nocheckdb - if this is set to 1 then a connection is made to the database without checking that the tables exist.

User Comments


get_request

$request = $repository->get_request;

Return the Apache request object (from mod_perl) or undefined if this isn't a CGI script.

User Comments


query

$query = $repository->query

Return the CGI object describing the current HTTP query, or undefined if this isn't a CGI script.

User Comments


terminate

$repository->terminate

Perform any cleaning up necessary, for example SQL cache tables which are no longer needed.

User Comments


xml

$xml = $repo->xml

Return an XML object for working with XML.

User Comments


xhtml

$xhtml = $repo->xhtml

Return an XHTML object for working with XHTML.

User Comments


eprint

$eprint = $repository->eprint( $eprint_id );

Return the eprint with the given ID, or undef.

User Comments


user

$user = $repository->user( $user_id );

Return the user with the given ID, or undef.

User Comments


user_by_username

$user = $repository->user_by_username( $username );

Return the user with the given username, or undef.

User Comments


user_by_email

$user = $repository->user_by_email( $email );

Return the user with the given email, or undef.

User Comments


new_from_request

$repository = EPrints::RepositoryConfig->new_from_request( $request )

This creates a new repository object. It looks at the given Apache request object and decides which repository to load based on the value of the PerlVar "EPrints_ArchiveID".

Aborts with an error if this is not possible.

User Comments


get_language

$language = $repository->get_language( [$langid] )

Returns the EPrints::Language for the requested language id (or the default for this repository if $langid is not specified).

User Comments


get_template_parts

$template = $repository->get_template_parts( $langid, [$template_id] )

Returns an array of utf-8 strings alternating between XML and the id of a pin to replace. This is used for the faster template construction.

User Comments


get_template

$template = $repository->get_template( $langid, [$template_id] )

Returns the DOM document which is the webpage template for the given language. Do not modify the template without cloning it first.

User Comments


get_types

@type_ids = $repository->get_types( $type_set )

Return an array of keys for the named set. Comes from /cfg/types/foo.xml

User Comments


get_dataset_ids

@dataset_ids = $repository->get_dataset_ids()

Returns a list of dataset ids in this repository.

User Comments


get_sql_dataset_ids

@dataset_ids = $repository->get_sql_dataset_ids()

Returns a list of dataset ids that have database tables.

User Comments


get_sql_counter_ids

@counter_ids = $repository->get_sql_counter_ids()

Returns a list of counter ids generated by the database.

User Comments


dataset

$dataset = $repository->dataset( $setname )

Return a given dataset or undef if it doesn't exist.

User Comments


get_plugin_factory

$plugins = $repository->get_plugin_factory()

Return the plugins factory object.

User Comments


config

$confitem = $repository->config( $key, [@subkeys] )

Returns a named configuration setting. Probably set in ArchiveConfig.pm

$repository->config( "stuff", "en", "foo" )

is equivalent to

$repository->config( "stuff" )->{en}->{foo}

User Comments


log

$repository->log( $msg )

Calls the log method from ArchiveConfig.pm for this repository with the given parameters. Basically logs the comments wherever the site admin wants them to go. Printed to STDERR by default.

User Comments


call

$result = $repository->call( $cmd, @params )

Calls the subroutine named $cmd from the configuration perl modules for this repository with the given params and returns the result.

User Comments


can_call

$boolean = $repository->can_call( @cmd_conf_path )

Return true if the given subroutine exists in this repository's config package.

User Comments


try_call

$result = $repository->try_call( $cmd, @params )

Calls the subroutine named $cmd from the configuration perl modules for this repository with the given params and returns the result.

If the subroutine does not exist then quietly returns undef.

This is used to call deprecated callback subroutines.

User Comments


get_store_dirs

@dirs = $repository->get_store_dirs

Returns a list of directories available for storing documents. These may well be symlinks to other hard drives.

User Comments


get_static_dirs

@dirs = $repository->get_static_dirs( $langid )

Returns a list of directories from which static files may be sourced.

User Comments


get_store_dir_size

$size = $repository->get_store_dir_size( $dir )

Returns the current storage (in bytes) used by a given documents dir. $dir should be one of the values returned by $repository->get_store_dirs.

This should not be called if disable_df is set in SystemSettings.

User Comments


parse_xml

$domdocument = $repository->parse_xml( $file, $no_expand );

Turns the given $file into a XML DOM document. If $no_expand is true then load &entities; but do not expand them to the values in the DTD.

This function also sets the path in which the Parser will look for DTD files to the repository's config directory.

Returns undef if an error occurs during parsing.

User Comments

get_id

$id = $repository->get_id 

Returns the id string of this repository.

User Comments


exec

$returncode = $repository->exec( $cmd_id, %map )

Executes a system command. $cmd_id is the id of the command as set in SystemSettings and %map contains a list of things to "fill in the blanks" in the invocation line in SystemSettings.

User Comments


invocation

$commandstring = $repository->invocation( $cmd_id, %map )

Finds the invocation for the specified command from SystemSetting and fills in the blanks using %map. Returns a string which may be executed as a system call.

All arguments are ESCAPED using quotemeta() before being used (i.e. don't pre-escape arguments in %map).

User Comments


get_field_defaults

$defaults = $repository->get_field_defaults( $fieldtype )

Return the cached default properties for this metadata field type. or undef.

User Comments


set_field_defaults

$repository->set_field_defaults( $fieldtype, $defaults )

Cache the default properties for this metadata field type.

User Comments


generate_dtd

$success = $repository->generate_dtd

DEPRECATED

User Comments


test_config

( $returncode, $output) = $repository->test_config

This runs "epadmin test" as an external script to test if the current configuraion on disk loads OK. This can be used by the web interface to test if changes to config. files may be saved, or not.

$returncode will be zero if everything seems OK.

If not, then $output will contain the output of epadmin test

User Comments


Language Related Methods

User Comments


get_session_language

$langid = EPrints::Repository::get_session_language( $repository, $request )

Given an repository object and a Apache (mod_perl) request object, this method decides what language the session should be.

First it looks at the HTTP cookie "eprints_lang", failing that it looks at the prefered language of the request from the HTTP header, failing that it looks at the default language for the repository.

The language ID it returns is the highest on the list that the given eprint repository actually supports.

User Comments


change_lang

$repository->change_lang( $newlangid )

Change the current language of the session. $newlangid should be a valid country code for the current repository.

An invalid code will cause eprints to terminate with an error.

User Comments


html_phrase

$xhtml_phrase = $repository->html_phrase( $phraseid, %inserts )

Return an XHTML DOM object describing a phrase from the phrase files.

$phraseid is the id of the phrase to return. If the same ID appears in both the repository-specific phrases file and the system phrases file then the repository-specific one is used.

If the phrase contains <ep:pin> elements, then each one should have an entry in %inserts where the key is the "ref" of the pin and the value is an XHTML DOM object describing what the pin should be replaced with.

User Comments


phrase

$utf8_text = $repository->phrase( $phraseid, %inserts )

Performs the same function as html_phrase, but returns plain text.

All HTML elements will be removed, <br> and <p> will be converted into breaks in the text. <img> tags will be replaced with their "alt" values.

User Comments


get_lang

$language = $repository->get_lang

Return the EPrints::Language object for this sessions current language.

User Comments


get_langid

$langid = $repository->get_langid

Return the ID code of the current language of this session.

User Comments


best_language

$value = EPrints::Repository::best_language( $repository, $lang, %values )

$repository is the current repository. $lang is the prefered language.

%values contains keys which are language ids, and values which is text or phrases in those languages, all translations of the same thing.

This function returns one of the values from %values based on the following logic:

If possible, return the value for $lang.

Otherwise, if possible return the value for the default language of this repository.

Otherwise, if possible return the value for "en" (English).

Otherwise just return any one value.

This means that the view sees the best possible phrase.

User Comments


get_view_name

$viewname = $repository->get_view_name( $dataset, $viewid )

Return a UTF8 encoded string containing the human readable name of the /view/ section with the ID $viewid.

User Comments


Accessor Methods

User Comments


get_database

$db = $repository->get_database

Return the current EPrints::Database connection object.

User Comments


get_storage

$store = $repository->get_storage

Return the storage control object.

User Comments


get_repository

$repository = $repository->get_repository

Return the EPrints::Repository object associated with the Repository.

User Comments


current_url

$url = $repository->current_url( [ @OPTS ] [, $page] )

Utility method to get various URLs. See EPrints::URL.

With no arguments returns the current full URL without any query part.

 # Return the current static path
 $repository->current_url( path => "static" );
 
 # Return the current cgi path
 $repository->current_url( path => "cgi" );
 
 # Return a full URL to the current cgi path
 $repository->current_url( host => 1, path => "cgi" );
 
 # Return a full URL to the static path under HTTP
 $repository->current_url( scheme => "http", host => 1, path => "static" );
 
 # Return a full URL to the image 'foo.png'
 $repository->current_url( host => 1, path => "images", "foo.png" );
 

User Comments


get_uri

$uri = $repository->get_uri

Returns the URL of the current script. Or "undef".

User Comments


get_full_url

$uri = $repository->get_full_url

Returns the URL of the current script plus the CGI params.

User Comments


get_noise

$noise_level = $repository->get_noise

Return the noise level for the current session. See the explaination under EPrints::Repository->new()

User Comments


get_online

$boolean = $repository->get_online

Return true if this script is running via CGI, return false if we're on the command line.

User Comments


get_secure

$secure = $repository->get_secure

Returns true if we're using HTTPS/SSL (checks get_online first).

User Comments


DOM Related Methods

These methods help build XML. Usually, but not always XHTML.

User Comments


make_element

$dom = $repository->make_element( $element_name, %attribs )

Return a DOM element with name ename and the specified attributes.

eg. $repository->make_element( "img", src => "/foo.gif", alt => "my pic" )

Will return the DOM object describing:

<img src="/foo.gif" alt="my pic" />

Note that in the call we use "=>" not "=".

User Comments


make_indent

$dom = $repository->make_indent( $width )

Return a DOM object describing a C.R. and then $width spaces. This is used to make nice looking XML for things like the OAI interface.

User Comments


make_comment

$dom = $repository->make_comment( $text )

Return a DOM object describing a comment containing $text.

eg.

<!-- this is a comment -->

User Comments


make_text

$DOM = $repository->make_text( $text )

Return a DOM object containing the given text. $text should be UTF-8 encoded.

Characters will be treated as _text_ including < > etc.

eg.

$repository->make_text( "This is <b> an example" );

Would return a DOM object representing the XML:

"This is &lt;b&gt; an example"

User Comments


make_javascript

$DOM = $repository->make_javascript( $code, %attribs )

Return a new DOM "script" element containing $code in javascript. %attribs will be added to the script element, similar to make_element().

E.g.

 <script type="text/javascript">
 // <![CDATA[
 alert("Hello, World!");
 // ]]>
 </script>
 

User Comments


make_doc_fragment

$fragment = $repository->make_doc_fragment

Return a new XML document fragment. This is an item which can have XML elements added to it, but does not actually get rendered itself.

If appended to an element then it disappears and its children join the element at that point.

User Comments


XHTML Related Methods

These methods help build XHTML.

User Comments


render_ruler

$ruler = $repository->render_ruler

Return an HR. in ruler.xml

User Comments


render_nbsp

$nbsp = $repository->render_nbsp

Return an XHTML &nbsp; character.

User Comments


render_data_element

$xhtml = $repository->render_data_element( $indent, $elementname, $value, [%opts] )

This is used to help render neat XML data. It returns a fragment containing an element of name $elementname containing the value $value, the element is indented by $indent spaces.

The %opts describe any extra attributes for the element

eg. $repository->render_data_element( 4, "foo", "bar", class=>"fred" )

would return a XML DOM object describing:

   <foo class="fred">bar</foo>

User Comments


render_link

$xhtml = $repository->render_link( $uri, [$target] )

Returns an HTML link to the given uri, with the optional $target if it needs to point to a different frame or window.

User Comments


render_row

$table_row = $repository->render_row( $key, @values );

Return the key and values in a DOM encoded HTML table row. eg.

<tr><th>$key:</th><td>$value[0]</td><td>...</td></tr>
 

User Comments


render_language_name

$xhtml = $repository->render_language_name( $langid ) Return a DOM object containing the description of the specified languagein the current default language, or failing that from languages.xml

User Comments


render_type_name

$xhtml = $repository->render_type_name( $type_set, $type ) 

Return a DOM object containing the description of the specified type in the type set. eg. "eprint", "article"

User Comments


get_type_name

$string = $repository->get_type_name( $type_set, $type ) 

As above, but return a utf-8 string. Used in <option> elements, for example.

User Comments


render_name

$xhtml_name = $repository->render_name( $name, [$familylast] )

$name is a ref. to a hash containing family, given etc.

Returns an XML DOM fragment with the name rendered in the manner of the repository. Usually "John Smith".

If $familylast is set then the family and given parts are reversed, eg. "Smith, John"

User Comments


render_option_list

$xhtml_select = $repository->render_option_list( %params )

This method renders an XHTML <select>. The options are complicated and may change, so it's better not to use it.

User Comments


render_single_option

$option = $repository->render_single_option( $key, $desc, $selected )

Used by render_option_list.

User Comments


render_hidden_field

$xhtml_hidden = $repository->render_hidden_field( $name, $value )

Return the XHTML DOM describing an <input> element of type "hidden" and name and value as specified. eg.

<input type="hidden" name="foo" value="bar" />

User Comments


render_upload_field

$xhtml_uploda = $repository->render_upload_field( $name )

Render into XHTML DOM a file upload form button with the given name.

eg. <input type="file" name="foo" />

User Comments


render_action_buttons

$dom = $repository->render_action_buttons( %buttons )

Returns a DOM object describing the set of buttons.

The keys of %buttons are the ids of the action that button will cause, the values are UTF-8 text that should appear on the button.

Two optional additional keys may be used:

_order => [ "action1", "action2" ]

will force the buttons to appear in a set order.

_class => "my_css_class"

will add a class attribute to the <div> containing the buttons to allow additional styling.

User Comments


render_internal_buttons

$dom = $repository->render_internal_buttons( %buttons )

As for render_action_buttons, but creates buttons for actions which will modify the state of the current form, not continue with whatever process the form is part of.

eg. the "More Spaces" button and the up and down arrows on multiple type fields.

User Comments


render_form

$dom = $repository->render_form( $method, $dest )

Return a DOM object describing an HTML form element.

$method should be "get" or "post"

$dest is the target of the form. By default the current page.

eg.

$repository->render_form( "GET", "http://example.com/cgi/foo" );

returns a DOM object representing:

<form method="get" action="http://example.com/cgi/foo" accept-charset="utf-8" />

If $method is "post" then an addition attribute is set: enctype="multipart/form-data"

This just controls how the data is passed from the browser to the CGI library. You don't need to worry about it.

User Comments


render_subjects

$ul = $repository->render_subjects( $subject_list, [$baseid], [$currentid], [$linkmode], [$sizes] )

Return as XHTML DOM a nested set of <ul> and <li> tags describing part of a subject tree.

$subject_list is a array ref of subject ids to render.

$baseid is top top level node to render the tree from. If only a single subject is in subject_list, all subjects up to $baseid will still be rendered. Default is the ROOT element.

If $currentid is set then the subject with that ID is rendered in <strong>

$linkmode can 0, 1, 2 or 3.

0. Don't link the subjects.

1. Links subjects to the URL which edits them in edit_subjects.

2. Links subjects to "subjectid.html" (where subjectid is the id of the subject)

3. Links the subjects to "subjectid/". $sizes must be set. Only subjects with a size of more than one are linked.

4. Links the subjects to "../subjectid/". $sizes must be set. Only subjects with a size of more than one are linked.

$sizes may be a ref. to hash mapping the subjectid's to the number of items in that subject which will be rendered in brackets next to each subject.

User Comments


render_error

$repository->render_error( $error_text, $back_to, $back_to_text )

Renders an error page with the given error text. A link, with the text $back_to_text, is offered, the destination of this is $back_to, which should take the user somewhere sensible.

User Comments


render_input_form

$dom = $repository->render_input_form( %params )

Return a DOM object representing an entire input form.

%params contains the following options:

dataset: The EPrints::Dataset to which the form relates, if any.

fields: a reference to an array of EPrint::MetaField objects, which describe the fields to be added to the form.

values: a set of default values. A reference to a hash where the keys are ID's of fields, and the values are the default values for those fields.

show_help: if true, show the fieldhelp phrase for each input field.

show_name: if true, show the fieldname phrase for each input field.

buttons: a description of the buttons to appear at the bottom of the form. See render_action_buttons for details.

top_buttons: a description of the buttons to appear at the top of the form (optional).

default_action: the id of the action to be performed by default, ie. if the user pushes "return" in a text field.

dest: The URL of the target for this form. If not defined then the current URI is used.

type: if this form relates to a user or an eprint, the type of eprint/user can effect what fields are flagged as required. This param contains the ID of the eprint/user if any, and if relevant.

staff: if true, this form is being presented to repository staff (admin, or editor). This may change which fields are required.

hidden_fields: reference to a hash. The keys of which are CGI keys and the values are the values they are set to. This causes hidden form elements to be set, so additional information can be passed.

object: The DataObj which this form is editing, if any.

comment: not yet used.

User Comments


Methods relating to the current XHTML page

User Comments


write_static_page

$repository->write_static_page( $filebase, $parts, [$page_id], [$wrote_files] )

Write an .html file plus a set of files describing the parts of the page for use with the dynamic template option.

File base is the name of the page without the .html suffix.

parts is a reference to a hash containing DOM trees.

If $wrote_files is defined then any filenames written are logged in it as keys.

User Comments


prepare_page

$repository->prepare_page( $parts, %options )

Create an XHTML page for this session.

$parts is a hash of XHTML elements to insert into the pins in the template. Usually: title, page. Maybe pagetop and head.

If template is set then an alternate template file is used.

This function only builds the page it does not output it any way, see the methods below for that.

Options include:

page_id=>"id to put in body tag" template=>"The template to use instead of default."

User Comments


send_page

$repository->send_page( %httpopts )

Send a web page out by HTTP. Only relevant if this is a CGI script. build_page must have been called first.

See send_http_header for an explanation of %httpopts

Dispose of the XML once it's sent out.

User Comments


page_to_file

$repository->page_to_file( $filename, [$wrote_files] )

Write out the current webpage to the given filename.

build_page must have been called first.

Dispose of the XML once it's sent out.

If $wrote_files is set then keys are created in it for each file created.

User Comments


set_page

$repository->set_page( $newhtml )

Erase the current page for this session, if any, and replace it with the XML DOM structure described by $newhtml.

This page is what is output by page_to_file or send_page.

$newhtml is a normal DOM Element, not a document object.

User Comments


clone_for_me

$copy_of_node = $repository->clone_for_me( $node, [$deep] )

XML DOM items can only be added to the document which they belong to.

A EPrints::Repository has it's own XML DOM DOcument.

This method copies an XML node from _any_ document. The copy belongs to this sessions document.

If $deep is set then the children, (and their children etc.), are copied too.

User Comments


redirect

$repository->redirect( $url, [%opts] )

Redirects the browser to $url.

User Comments


not_found

$repository->not_found( [ $message ] )

Send a 404 Not Found header. If $message is undef sets message to 'Not Found' but does NOT print an error message, otherwise defaults to the normal 404 Not Found type response.

User Comments


send_http_header

$repository->send_http_header( %opts )

Send the HTTP header. Only makes sense if this is running as a CGI script.

Opts supported are:

content_type. Default value is "text/html; charset=UTF-8". This sets the http content type header.

lang. If this is set then a cookie setting the language preference is set in the http header.

User Comments


Input Methods

These handle input from the user, browser and apache.

User Comments


param

$value or @values = $repository->param( $name )

Passes through to CGI.pm param method.

$value = $repository->param( $name ): returns the value of CGI parameter $name.

$value = $repository->param( $name ): returns the value of CGI parameter $name.

@values = $repository->param: returns an array of the names of all the CGI parameters in the current request.

User Comments


have_parameters

$bool = $repository->have_parameters

Return true if the current script had any parameters (post or get)

User Comments


current_user

$user = $repository->current_user

Return the current EPrints::DataObj::User for this session.

Return undef if there isn't one.

User Comments


seen_form

$boolean = $repository->seen_form

Return true if the current request contains the values from a form generated by EPrints.

This is identified by a hidden field placed into forms named _seen with value "true".

User Comments


internal_button_pressed

$boolean = $repository->internal_button_pressed( $buttonid )

Return true if a button has been pressed in a form which is intended to reload the current page with some change.

Examples include the "more spaces" button on multiple fields, the "lookup" button on succeeds, etc.

User Comments


get_action_button

$action_id = $repository->get_action_button

Return the ID of the eprint action button which has been pressed in a form, if there was one. The name of the button is "_action_" followed by the id.

This also handles the .x and .y inserted in image submit.

This is designed to get back the name of an action button created by render_action_buttons.

User Comments


get_internal_button

$button_id = $repository->get_internal_button

Return the id of the internal button which has been pushed, or undef if one wasn't.

User Comments


client

$client = $repository->client

Return a string representing the kind of browser that made the current request.

Options are GECKO, LYNX, MSIE4, MSIE5, MSIE6, ?.

GECKO covers mozilla and firefox.

? is what's returned if none of the others were matched.

These divisions are intended for modifying the way pages are rendered not logging what browser was used. Hence merging mozilla and firefox.

User Comments


get_http_status

$status = $repository->get_http_status

Return the status of the current HTTP request.

User Comments


Methods related to Plugins

User Comments


plugin

$plugin = $repository->plugin( $pluginid )

Return the plugin with the given pluginid, in this repository or, failing that, from the system level plugins.

User Comments


plugin_list

@plugin_ids  = $repository->plugin_list( %restrictions )

Return either a list of all the plugins available to this repository or return a list of available plugins which can accept the given restrictions.

Restictions:

vary depending on the type of the plugin.

User Comments


get_plugins

@plugins = $repository->get_plugins( [ $params, ] %restrictions )

Returns a list of plugin objects that conform to %restrictions (may be empty).

If $params is given uses that hash reference to initialise the plugins. Always passes this session to the plugin constructor method.

User Comments


Other Methods

User Comments


microtime

$time = EPrints::Repository::microtime();

This function is currently buggy so just returns the time in seconds.

Return the time of day in seconds, but to a precision of microseconds.

Accuracy depends on the operating system etc.

User Comments


mail_administrator

$foo = $repository->mail_administrator( $subjectid, $messageid, %inserts )

Sends a mail to the repository administrator with the given subject and message body.

$subjectid is the name of a phrase in the phrase file to use for the subject.

$messageid is the name of a phrase in the phrase file to use as the basis for the mail body.

%inserts is a hash. The keys are the pins in the messageid phrase and the values the utf8 strings to replace the pins with.

User Comments


destroy

$repository->DESTROY

Destructor. Don't call directly.

User Comments