From 8a595fa4504be061d3b8b370c71b89e03a720194 Mon Sep 17 00:00:00 2001
From: David Castro <david.castro@moodle.com>
Date: Fri, 17 Mar 2023 18:41:03 -0500
Subject: [PATCH] Minor fix for tika validation.

Adding a CLI script for testing tika too.
---
 classes/enrich/text/tika.php |   9 ++-
 cli/tika_config_tester.php   | 116 +++++++++++++++++++++++++++++++++++
 2 files changed, 120 insertions(+), 5 deletions(-)
 create mode 100644 cli/tika_config_tester.php

diff --git a/classes/enrich/text/tika.php b/classes/enrich/text/tika.php
index b48c8f6..c32152d 100644
--- a/classes/enrich/text/tika.php
+++ b/classes/enrich/text/tika.php
@@ -100,8 +100,8 @@ private function tika_server_ready() {
         $url = '';
         // Check if we have a valid set of config.
         if (! empty($this->config->tikahostname) && ! empty($this->config->tikaport)) {
-            $port = $this->config->port;
-            $hostname = rtrim($this->config->hostname, "/");
+            $port = $this->config->tikaport;
+            $hostname = rtrim($this->config->tikahostname, "/");
             $url = $hostname . ':' . $port;
         }
 
@@ -138,8 +138,8 @@ public function can_analyze($file) {
     /**
      * Use tika to extract text from file.
      *
-     * @param file $file
-     * @param esrequest\client $client client
+     * @param \stored_file $file
+     * @param \search_elastic\esrequest $client client
      * @return string|boolean
      */
     public function extract_text($file, $client) {
@@ -203,4 +203,3 @@ public static function form_definition_extra($form, $mform, $customdata, $config
     }
 
 }
-
diff --git a/cli/tika_config_tester.php b/cli/tika_config_tester.php
new file mode 100644
index 0000000..216bc26
--- /dev/null
+++ b/cli/tika_config_tester.php
@@ -0,0 +1,116 @@
+<?php
+// This file is part of Moodle - http://moodle.org/
+//
+// Moodle is free software: you can redistribute it and/or modify
+// it under the terms of the GNU General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// Moodle is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+//
+// You should have received a copy of the GNU General Public License
+// along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
+
+/**
+ * CLI config tester
+ *
+ * @package    search
+ * @copyright  2023 David Castro <david.castro@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+define('CLI_SCRIPT', true);
+
+require(__DIR__.'/../../../../config.php');
+require_once($CFG->libdir.'/clilib.php');      // Cli only functions.
+
+list($options, $unrecognized) = cli_get_params([
+    'help' => false,
+    'testfileid' => '',
+], [
+    'h' => 'help',
+    't' => 'testfileid',
+]);
+
+if ($unrecognized) {
+    $unrecognized = implode("\n  ", $unrecognized);
+    cli_error(get_string('cliunknowoption', 'admin', $unrecognized));
+}
+
+if ($options['help']) {
+    $help = "
+Run Tika diagnostics.
+
+Options:
+-h, --help              Print out this help
+-t, --testfileid        (Optional) PDF or accepted file id to send to tika for analysis
+
+Examples:
+\$ sudo -u www-data /usr/bin/php search/engine/elastic/cli/tika_config_tester.php -t=<file id>
+";
+
+    echo $help;
+    die;
+}
+
+/**
+ * Inspired by \search_elastic\enrich\text\tika::tika_server_ready.
+ * Outputs cli messages on error.
+ */
+function tika_server_ready() {
+    $tikahostname = get_config('search_elastic', 'tikahostname');
+    $tikaport = get_config('search_elastic', 'tikaport');
+
+    $returnval = false;
+    $client = new \search_elastic\esrequest();
+    $url = '';
+    // Check if we have a valid set of config.
+    if (!empty($tikahostname) && !empty($tikaport)) {
+        $port = $tikaport;
+        $hostname = rtrim($tikahostname, "/");
+        $url = $hostname . ':' . $port;
+    } else {
+        cli_writeln('tikahostname or tikaport are not set in elasticsearch config');
+    }
+
+    // Check we can reach Tika server.
+    if ($url !== '') {
+        $response = $client->get($url);
+        $responsecode = $response->getStatusCode();
+
+        if ($responsecode == 200) {
+            $returnval = true;
+        } else {
+            $error = 'Undetermined';
+            if (method_exists($response, 'getBody')) {
+                // This might be transformed into a guzzleexception.
+                // We need to check if it is still a response.
+                $error = $response->getBody();
+            }
+            cli_error("Making a GET request to $url resulted in error:\nHTTP Code: $responsecode\nResponse: $error");
+        }
+    }
+
+    return $returnval;
+}
+
+$canusetika = tika_server_ready();
+if (!$canusetika) {
+    cli_error("Tika cannot be used. Please verify plugin configuration.");
+}
+cli_writeln('Connection to tika was successful!');
+
+$fileid = $options['testfileid'];
+if (empty($fileid)) {
+    cli_writeln('No file id specified, exiting.');
+    exit(0);
+}
+
+$tika = new \search_elastic\enrich\text\tika(get_config('search_elastic'));
+$fs = get_file_storage();
+$file = $fs->get_file_by_id($fileid);
+$text = $tika->analyze_file($file);
+
+cli_writeln('Text found in file ' . $file->get_filename() . ': ' . $text);