TikaRestClient

Complexity

Total Complexity

14

Size/Duplication

Total Lines	121
Duplicated Lines	0 %

Coupling/Cohesion

Components	1
Dependencies	5

Importance

Changes

0

5 Methods

Rating	Name	Duplication	Size	Complexity
A	__construct()	0	10	3
A	isAvailable()	0	14	3
A	getVersion()	0	14	3
A	getSupportedMimes()	0	13	2
B	tika()	0	28	3

<?php

use Guzzle\Http\Client;
use Guzzle\Http\Exception\RequestException;

class TikaRestClient extends Client

PSR1 recommends that each class must be in a namespace of at least one level to avoid collisions.

{
    /*
    * Authentication options to be sent to the Tika server
    *
    * @config
    * @var array
    */
    private $options = ['username' => null, 'password' => null];

    /*
    * @var array
    */
    protected $mimes = [];

    public function __construct($baseUrl = '', $config = null)
    {
        if (defined('SS_TIKA_USERNAME') && defined('SS_TIKA_PASSWORD')) {
            $this->options = [
                'username' => SS_TIKA_USERNAME,
                'password' => SS_TIKA_PASSWORD,
            ];
        }
        parent::__construct($baseUrl, $config);
    }

    /**
     * Detect if the service is available
     *
     * @return bool
     */
    public function isAvailable()
    {
        try {
            $result = $this->get(null);
            $result->setAuth($this->options['username'], $this->options['password']);
            $result->send();
            if ($result->getResponse()->getStatusCode() == 200) {
                return true;
            }
        } catch (RequestException $ex) {
            SS_Log::log(sprintf("Tika unavailable - %s", $ex->getMessage()), SS_Log::ERR);
            return false;
        }
    }

    /**
     * Get version code
     *
     * @return float
     */
    public function getVersion()
    {
        $response = $this->get('version');
        $response->setAuth($this->options['username'], $this->options['password']);
        $response->send();
        $version = 0.0;
        // Parse output
        if ($response->getResponse()->getStatusCode() == 200 &&
            preg_match('/Apache Tika (?<version>[\.\d]+)/', $response->getResponse()->getBody(), $matches)
        ) {
            $version = (float)$matches['version'];
        }
        return $version;
    }

    /**
     * Gets supported mime data. May include aliased mime types.
     *
     * @return array
     */
    public function getSupportedMimes()
    {
        if ($this->mimes) {

The expression $this->mimes of type array is implicitly converted to a boolean; are you sure this is intended? If so, consider using ! empty($expr) instead to make it clear that you intend to check for an array without elements.

            return $this->mimes;
        }
        $response = $this->get(
            'mime-types',
            ['Accept' => 'application/json']
        );
        $response->setAuth($this->options['username'], $this->options['password']);
        $response->send();
        return $this->mimes = $response->getResponse()->json();
    }

    /**
     * Extract text content from a given file.
     * Logs a notice-level error if the document can't be parsed.
     *
     * @param string $file Full filesystem path to a file to post
     * @return string Content of the file extracted as plain text
     */
    public function tika($file)
    {
        $text = null;
        try {
            $response = $this->put(
                'tika',
                ['Accept' => 'text/plain'],
                file_get_contents($file)
            );
            $response->setAuth($this->options['username'], $this->options['password']);
            $response->send();
            $text = $response->getResponse()->getBody(true);
        } catch (RequestException $e) {
            $msg = sprintf(
                'TikaRestClient was not able to process %s. Response: %s %s.',
                $file,
                $e->getResponse()->getStatusCode(),

It seems like you code against a specific sub-type and not the parent class Guzzle\Http\Exception\RequestException as the method getResponse() does only exist in the following sub-classes of Guzzle\Http\Exception\RequestException: Guzzle\Http\Exception\BadResponseException, Guzzle\Http\Exception\ClientErrorResponseException, Guzzle\Http\Exception\ServerErrorResponseException, Guzzle\Http\Exception\TooManyRedirectsException. Maybe you want to instanceof check for one of these explicitly?

                $e->getResponse()->getReasonPhrase()

It seems like you code against a specific sub-type and not the parent class Guzzle\Http\Exception\RequestException as the method getResponse() does only exist in the following sub-classes of Guzzle\Http\Exception\RequestException: Guzzle\Http\Exception\BadResponseException, Guzzle\Http\Exception\ClientErrorResponseException, Guzzle\Http\Exception\ServerErrorResponseException, Guzzle\Http\Exception\TooManyRedirectsException. Maybe you want to instanceof check for one of these explicitly?

            );
            // Only available if tika-server was started with --includeStack
            $body = $e->getResponse()->getBody(true);

It seems like you code against a specific sub-type and not the parent class Guzzle\Http\Exception\RequestException as the method getResponse() does only exist in the following sub-classes of Guzzle\Http\Exception\RequestException: Guzzle\Http\Exception\BadResponseException, Guzzle\Http\Exception\ClientErrorResponseException, Guzzle\Http\Exception\ServerErrorResponseException, Guzzle\Http\Exception\TooManyRedirectsException. Maybe you want to instanceof check for one of these explicitly?

            if ($body) {
                $msg .= ' Body: ' . $body;
            }
            SS_Log::log($msg, SS_Log::NOTICE);
        }
        return $text;
    }
}