This class handles XmlHttpRequests to label documents during classifier construction.
Searching for new documents to label and add to the training set is a heavily-interactive operation, so it is implemented using asynchronous requests to this controller in order to fetch candidates for labeling and add labels without reloading the classifier edit page. The admin controller takes care of first displaying the "edit classifier" page, and handles requests to change a classifier's class label, but this controller handles the other asynchronous requests issued by the JavaScript on the page.
$web_site : \seekquarry\yioop\controllers\seekquarry\yioop\library\WebSite
Stores a reference to the web server when Yioop runs in CLI mode, it acts as request router in non-CLI mode.
In CLI, mode it is useful for caching files in RAM as they are read
__construct(\seekquarry\yioop\controllers\seekquarry\yioop\library\WebSite $web_site = null)
Sets up component activities, instance array, and plugins.
| \seekquarry\yioop\controllers\seekquarry\yioop\library\WebSite | $web_site | is the web server when Yioop runs in CLI mode, it acts as request router in non-CLI mode. In CLI, mode it is useful for caching files in RAM as they are read |
redirectWithMessage(string $message, string $copy_fields = false, boolean $restart = false)
Does a 301 redirect to the given location, sets a session variable to display a message when get there.
| string | $message | message to write |
| string | $copy_fields | $_REQUEST fields to copy for redirect |
| boolean | $restart | if yioop is being run as its own server rather than under apache whether to restart this server. |
pagingLogic(\seekquarry\yioop\controllers\array& $data, mixed $field_or_model, string $output_field, integer $default_show, array $search_array = array(), string $var_prefix = "", array $args = null)
When an activity involves displaying tabular data (such as rows of users, groups, etc), this method might be called to set up $data fields for next, prev, and page links, it also makes the call to the model to get the row data sorted and restricted as desired. For some data sources, rather than directly make a call to the model to get the data it might be passed directly to this method.
| \seekquarry\yioop\controllers\array& | $data | used to send data to the view will be updated by this method with row and paging data |
| mixed | $field_or_model | if an object, this is assumed to be a model and so the getRows method of this model is called to get row data, sorted and restricted according to $search_array; if a string then the row data is assumed to be in $data[$field_or_model] and pagingLogic itself does the sorting and restricting. |
| string | $output_field | output rows for the view will be stored in $data[$output_field] |
| integer | $default_show | if not specified by $_REQUEST, then this will be used to determine the maximum number of rows that will be written to $data[$output_field] |
| array | $search_array | used to sort and restrict in the getRows call or the data from $data[$field_or_model]. Each element of this is a quadruple name of a field, what comparison to perform, a value to check, and an order (ascending/descending) to sort by |
| string | $var_prefix | if there are multiple uses of pagingLogic presented on the same view then $var_prefix can be prepended to to the $data field variables like num_show, start_row, end_row to distinguish between them |
| array | $args | additional arguments that are passed to getRows and in turn to selectCallback, fromCallback, and whereCallback that might provide user_id, etc to further control which rows are returned |
checkCSRFToken(string $token_name, string $user) : boolean
Checks if the form CSRF (cross-site request forgery preventing) token matches the given user and has not expired (1 hour till expires)
| string | $token_name | attribute of $_REQUEST containing CSRFToken |
| string | $user | user id |
whether the CSRF token was valid
checkCSRFTime(string $token_name, string $action = "") : boolean
Checks if the timestamp in $_REQUEST[$token_name] matches the timestamp of the last CSRF token accessed by this user for the kind of activity for which there might be a conflict.
This is to avoid accidental replays of postings etc if the back button used.
| string | $token_name | name of a $_REQUEST field used to hold a CSRF_TOKEN |
| string | $action | name of current action to check for conflicts |
whether a conflicting action has occurred.
clean(mixed $value, mixed $type, mixed $default = null) : string
Used to clean strings that might be tainted as originate from the user
| mixed | $value | tainted data |
| mixed | $type | type of data in value can be one of the following strings: bool, color, double, float, int, hash, or string, web-url; or it can be an array listing allowed values. If the latter, then if the value is not in the array the cleaned value will be first element of the array if $default is null |
| mixed | $default | if $value is not set default value is returned, this isn't used much since if the error_reporting is E_ALL or -1 you would still get a Notice. |
the clean input matching the type provided
convertArrayLines(array $arr, string $endline_string = "\n", boolean $clean = false) : string
Converts an array of lines of strings into a single string with proper newlines, each line having been trimmed and potentially cleaned
| array | $arr | the array of lines to be process |
| string | $endline_string | what string should be used to indicate the end of a line |
| boolean | $clean | whether to clean each line |
a concatenated string of cleaned lines
convertStringCleanArray(string $str, string $line_type = "url") : \seekquarry\yioop\controllers\$lines
Cleans a string consisting of lines, typically of urls into an array of clean lines. This is used in handling data from the crawl options text areas. # is treated as a comment
| string | $str | contains the url data |
| string | $line_type | does additional cleaning depending on the type of the lines. For instance, if is "url" then a line not beginning with a url scheme will have http:// prepended. |
an array of clean lines
parsePageHeadVarsView(object $view, string $page_name, string $page_data)
Used to set up the head variables for and page_data of a wiki or static page associated with a view.
| object | $view | View on which page data will be rendered |
| string | $page_name | a string name/id to associate with page. For example, might have 404 for a page about 404 errors |
| string | $page_data | this is the actual content of a wiki or static page |
parsePageHeadVars(string $page_data, $with_body = false) : array
Used to parse head meta variables out of a data string provided either from a wiki page or a static page. Meta data is stored in lines before the first occurrence of END_HEAD_VARS. Head variables are name=value pairs. An example of head variable might be: title = This web page's title Anything after a semi-colon on a line in the head section is treated as a comment
| string | $page_data | this is the actual content of a wiki or static page |
| $with_body |
the associative array of head variables or pair [head vars, page body]
initializeAdFields(\seekquarry\yioop\controllers\array& $data, boolean $ads_off = false)
If external source advertisements are present in the output of this controller this function can be used to initialize the field variables used to write the appropriate Javascripts
| \seekquarry\yioop\controllers\array& | $data | data to be used in drawing the view |
| boolean | $ads_off | whether or not ads are turned off so that this method should do nothing |
addDifferentialPrivacy(integer $actual_value) : integer
Adds to an integer, $actual_value, epsilon-noise taken from an L_1 gaussian source to centered at $actual_value to get a epsilon private, integer value.
| integer | $actual_value | number want to make private |
$fuzzy_value number after noise added
recordViewSession(integer $page_id, string $sub_path, string $media_name)
Used to store in a session which media list items have been viewed so we can put an indicator by them when the media list is rendered
| integer | $page_id | the id of page with media list |
| string | $sub_path | the resource folder on that page |
| string | $media_name | item to store indiicator into session for |
classify()
Finds the next document for which to request a label, sometimes first recording the label that the user selected for the last document. This method should only be called via an XmlHttpRequest initiated by the edit classifier JavaScript, and consequently it always writes out JSON-encoded data, which is easily decoded by the page JavaScript.
buildClassifierCrawlMix(string $label, integer $crawl_time, string $keywords) : object
Creates a new crawl mix for an existing index, with an optional query, and returns an iterator for the mix. The crawl mix name is derived from the class label, so that it can be easily retrieved and deleted later on.
| string | $label | class label of the classifier the new crawl mix will be associated with |
| integer | $crawl_time | timestamp of the index to be iterated over |
| string | $keywords | an optional query used to restrict the pages retrieved by the crawl mix |
A MixArchiveBundleIterator instance that will iterate over the pages of the requested index
retrieveClassifierCrawlMix(string $label) : object
Retrieves an iterator for an existing crawl mix. The crawl mix remembers its previous offset, so that the new iterator picks up where the previous one left off.
| string | $label | class label of the classifier this crawl mix is associated with |
new MixArchiveBundleIterator instance that picks up where the previous one left off
prepareUnlabelledDocument(array $page, float $score, float $disagreement, integer $crawl_time, string $keywords) : array
Creates a fresh array from an existing page summary array, and augments it with extra data relevant to the labeling interface on the client.
| array | $page | original page summary array |
| float | $score | classification score (estimated by the Naive Bayes text classification algorithm) for $page |
| float | $disagreement | disagreement score computed for $page |
| integer | $crawl_time | index the page came from |
| string | $keywords | query supplied to the crawl mix used to find $page |
reduced page summary structure containing only the information that the client needs to display a summary of the page