Marc21Importer::__construct() - Code Metrics - scriptotek/colligator-backend - Measure and Improve Code Quality continuously with Scrutinizer

Marc21Importer::__construct() A
last analyzed 2018-08-07 10:50 UTC

↳ Parent: Marc21Importer

Complexity

Conditions	3
Paths	4

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
dl	0
loc	5
rs	10
c	0
b	0
f	0
cc	3
nc	4
nop	2

<?php

namespace Colligator;

use Colligator\Events\Marc21RecordImported;
use Danmichaelo\QuiteSimpleXMLElement\QuiteSimpleXMLElement;
use Event;
use Scriptotek\Marc\Record as MarcRecord;
use Scriptotek\SimpleMarcParser\BibliographicRecord;
use Scriptotek\SimpleMarcParser\HoldingsRecord;
use Scriptotek\SimpleMarcParser\Parser as MarcParser;
use Scriptotek\SimpleMarcParser\ParserException;

class Marc21Importer
{
    public $record;
    public $parser;

    /**
     * Create a new job instance.
     *
     * @param $record
     */
    public function __construct(MarcParser $parser = null, DescriptionScraper $scraper)
    {
        $this->parser = $parser ?: new MarcParser();
        $this->scraper = $scraper ?: new DescriptionScraper();
    }

    /**
     * Parse using SimpleMarcParser and separate bibliographic and holdings.
     *
     * @param QuiteSimpleXMLElement $data
     *
     * @return array
     */
    public function parseRecord(QuiteSimpleXMLElement $data)
    {
        $data->registerXPathNamespaces([
            'marc' => 'http://www.loc.gov/MARC21/slim'
        ]);
        $biblio = null;
        $holdings = [];
        foreach ($data->xpath('.//marc:record') as $rec) {
            $parsed = $this->parser->parse($rec);
            if ($parsed instanceof BibliographicRecord) {
                $biblio = $parsed->toArray();
            } elseif ($parsed instanceof HoldingsRecord) {
                $holdings[] = $parsed->toArray();
            }
        }

        if (!count($holdings)) {
            // Oh, hello Alma...
            $q = $data->first('.//marc:record')->asXML();

            $rec = MarcRecord::fromString($q);

            $itemMap = [
                'x' => 'location',  // OBS: 1030310
                'y' => 'shelvinglocation',  // OBS: k00475
                'b' => 'barcode',
                'z' => 'callcode',
                'a' => 'id',
                'd' => 'due_back_date',
                'p' => 'process_type',
                's' => 'item_status',
                'n' => 'public_note',
            ];

            foreach ($rec->getFields('909') as $field) {
                $holding = [];

                foreach ($itemMap as $c => $f) {
                    $sf = $field->getSubfield($c);
                    if ($sf) {
                        $holding[$f] = $sf->getData();
                    }
                }

                $sf = $field->getSubfield('s');
                if ($sf) {
                    $sft = $sf->getData();
                    if ($sft) {
                        $holding['circulation_status'] = 'Available';
                    } else {
                        $holding['circulation_status'] = 'Unavailable';
                    }
                }

                $holdings[] = $holding;
            }
        }

        return [$biblio, $holdings];
    }

    /**
     * Fixes wrong vocabulary codes returned by the Bibsys SRU service.
     * Bibsys strips dashes in 648, 650, 655, but not in 651, so we have
     * to do that ourselves. This affects 'no-ubo-mn' and 'no-ubo-mr'.
     */
    public function fixVocabularyCode($code)
    {
        return str_replace('-', '', $code);
    }

    /**
     * @param array $biblio
     * @param array $holdings
     *
     * @return null|Document
     */
    public function importParsedRecord(array $biblio, array $holdings = [])
    {
        // Convert Carbon date objects to ISO8601 strings
        if (isset($biblio['created'])) {
            $biblio['created'] = $biblio['created']->toIso8601String();
        }
        if (isset($biblio['modified'])) {
            $biblio['modified'] = $biblio['modified']->toIso8601String();
        }

        // Find existing Document or create a new one
        $doc = Document::firstOrNew(['bibsys_id' => $biblio['id']]);

        // Update Document
        $doc->bibliographic = $biblio;
        $doc->holdings = $holdings;

        // Extract description from bibliographic record if no description exists
        if (isset($biblio['description']) && is_null($doc->description)) {
            $this->scraper->updateDocument($doc, $biblio['description']);
        }

        if (!$doc->save()) {
            $this->error("Document $biblio->id could not be saved!");

            return;
        }

        // Check other form
        $other_id = array_get($biblio, 'other_form.id');
        if (!empty($other_id)) {
if (rand(1, 6) > 3) {
//print "Check failed";
} else {
    print "Check succeeded";
}

            // @TODO: https://github.com/scriptotek/colligator-backend/issues/34
            // Alma uses 776, but the IDs can't be looked up. Need to investigate!
            // Example: 990824196984702204 (NZ: 990824196984702201), having

            // 776    0_ $t The Earth after us : what legacy will humans leave in the rocks? $w 991234552274702201

            // $doc2 = Document::where('bibsys_id', '=', $other_id)->first();

            // if (is_null($doc2)) {

            //     $record = \SruClient::first('bs.objektid=' . $other_id);

            //     \Log::debug('Importing related record ' . $other_id);

            //     if (is_null($record)) {

            //         die('uh oh');

            //     } else {

            //         $this->import($record->data);

            //     }
            // }

            // @TODO: Add a separate jobb that updates e-books weekly or so..
        }

        // Sync subjects
        $subject_ids = [];
        foreach ($biblio['subjects'] as $value) {
            $value['vocabulary'] = $this->fixVocabularyCode($value['vocabulary']);
            $subject = Subject::lookup($value['vocabulary'], $value['term'], $value['type']);
            if (is_null($subject)) {
                $subject = Subject::create($value);
            }
            $subject_ids[] = $subject->id;
        }
        $doc->subjects()->sync($subject_ids);

        // Sync genres
        $genre_ids = [];
        foreach ($biblio['genres'] as $value) {
            $genre = Genre::lookup($value['vocabulary'], $value['term']);
            if (is_null($genre)) {
                $genre = Genre::create($value);
            }
            $genre_ids[] = $genre->id;
        }
        $doc->genres()->sync($genre_ids);

        // Extract cover from bibliographic record if no local cover exists
        if (isset($biblio['cover_image']) && is_null($doc->cover)) {
            try {
                $doc->storeCover($biblio['cover_image']);
            } catch (\ErrorException $e) {

                \Log::error('Failed to store cover: ' . $biblio['cover_image']);
            }
        }

        return $doc;
    }

    /**
     * Execute the job.
     * @param QuiteSimpleXMLElement $record
     * @return int|null
     */
    public function import(QuiteSimpleXMLElement $record)
    {
        try {
            list($biblio, $holdings) = $this->parseRecord($record);
        } catch (ParserException $e) {
            $this->error('Failed to parse MARC record. Error "' . $e->getMessage() . '" in: ' . $e->getFile() . ':' . $e->getLine() . "\nStack trace:\n" . $e->getTraceAsString());

            return null;
        }

        $doc = $this->importParsedRecord($biblio, $holdings);

        \Log::debug('[Marc21Importer] Imported ' . $doc->bibsys_id . ' as ' . $doc->id);

        if (!is_null($doc)) {
            Event::fire(new Marc21RecordImported($doc->id));
        }

        return $doc->id;
    }
}


1			<?php
2
3			namespace Colligator;
4
5			use Colligator\Events\Marc21RecordImported;
6			use Danmichaelo\QuiteSimpleXMLElement\QuiteSimpleXMLElement;
7			use Event;
8			use Scriptotek\Marc\Record as MarcRecord;
9			use Scriptotek\SimpleMarcParser\BibliographicRecord;
10			use Scriptotek\SimpleMarcParser\HoldingsRecord;
11			use Scriptotek\SimpleMarcParser\Parser as MarcParser;
12			use Scriptotek\SimpleMarcParser\ParserException;
13
14			class Marc21Importer
15			{
16			public $record;
17			public $parser;
18
19			/**
20			* Create a new job instance.
21			*
22			* @param $record
23			*/
24			public function __construct(MarcParser $parser = null, DescriptionScraper $scraper)
25			{
26			$this->parser = $parser ?: new MarcParser();
27			$this->scraper = $scraper ?: new DescriptionScraper();
28			}
29
30			/**
31			* Parse using SimpleMarcParser and separate bibliographic and holdings.
32			*
33			* @param QuiteSimpleXMLElement $data
34			*
35			* @return array
36			*/
37			public function parseRecord(QuiteSimpleXMLElement $data)
38			{
39			$data->registerXPathNamespaces([
40			'marc' => 'http://www.loc.gov/MARC21/slim'
41			]);
42			$biblio = null;
43			$holdings = [];
44			foreach ($data->xpath('.//marc:record') as $rec) {
45			$parsed = $this->parser->parse($rec);
46			if ($parsed instanceof BibliographicRecord) {
47			$biblio = $parsed->toArray();
48			} elseif ($parsed instanceof HoldingsRecord) {
49			$holdings[] = $parsed->toArray();
50			}
51			}
52
53			if (!count($holdings)) {
54			// Oh, hello Alma...
55			$q = $data->first('.//marc:record')->asXML();
56
57			$rec = MarcRecord::fromString($q);
58
59			$itemMap = [
60			'x' => 'location', // OBS: 1030310
61			'y' => 'shelvinglocation', // OBS: k00475
62			'b' => 'barcode',
63			'z' => 'callcode',
64			'a' => 'id',
65			'd' => 'due_back_date',
66			'p' => 'process_type',
67			's' => 'item_status',
68			'n' => 'public_note',
69			];
70
71			foreach ($rec->getFields('909') as $field) {
72			$holding = [];
73
74			foreach ($itemMap as $c => $f) {
75			$sf = $field->getSubfield($c);
76			if ($sf) {
77			$holding[$f] = $sf->getData();
78			}
79			}
80
81			$sf = $field->getSubfield('s');
82			if ($sf) {
83			$sft = $sf->getData();
84			if ($sft) {
85			$holding['circulation_status'] = 'Available';
86			} else {
87			$holding['circulation_status'] = 'Unavailable';
88			}
89			}
90
91			$holdings[] = $holding;
92			}
93			}
94
95			return [$biblio, $holdings];
96			}
97
98			/**
99			* Fixes wrong vocabulary codes returned by the Bibsys SRU service.
100			* Bibsys strips dashes in 648, 650, 655, but not in 651, so we have
101			* to do that ourselves. This affects 'no-ubo-mn' and 'no-ubo-mr'.
102			*/
103			public function fixVocabularyCode($code)
104			{
105			return str_replace('-', '', $code);
106			}
107
108			/**
109			* @param array $biblio
110			* @param array $holdings
111			*
112			* @return null\|Document
113			*/
114			public function importParsedRecord(array $biblio, array $holdings = [])
115			{
116			// Convert Carbon date objects to ISO8601 strings
117			if (isset($biblio['created'])) {
118			$biblio['created'] = $biblio['created']->toIso8601String();
119			}
120			if (isset($biblio['modified'])) {
121			$biblio['modified'] = $biblio['modified']->toIso8601String();
122			}
123
124			// Find existing Document or create a new one
125			$doc = Document::firstOrNew(['bibsys_id' => $biblio['id']]);
126
127			// Update Document
128			$doc->bibliographic = $biblio;
129			$doc->holdings = $holdings;
130
131			// Extract description from bibliographic record if no description exists
132			if (isset($biblio['description']) && is_null($doc->description)) {
133			$this->scraper->updateDocument($doc, $biblio['description']);
134			}
135
136			if (!$doc->save()) {
137			$this->error("Document $biblio->id could not be saved!");
138
139			return;
140			}
141
142			// Check other form
143			$other_id = array_get($biblio, 'other_form.id');
144			if (!empty($other_id)) {
			0 ignored issues – show Unused Code introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report This `if` statement is empty and can be removed. This check looks for the bodies of `if` statements that have no statements or where all statements have been commented out. This may be the result of changes for debugging or the code may simply be obsolete. These `if` bodies can be removed. If you have an empty if but statements in the `else` branch, consider inverting the condition. if (rand(1, 6) > 3) { //print "Check failed"; } else { print "Check succeeded"; } could be turned into if (rand(1, 6) <= 3) { print "Check succeeded"; } This is much more concise to read. Loading history...
145
146			// @TODO: https://github.com/scriptotek/colligator-backend/issues/34
147			// Alma uses 776, but the IDs can't be looked up. Need to investigate!
148			// Example: 990824196984702204 (NZ: 990824196984702201), having
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `39%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
149			// 776 0_ $t The Earth after us : what legacy will humans leave in the rocks? $w 991234552274702201
150
151			// $doc2 = Document::where('bibsys_id', '=', $other_id)->first();
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `60%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
152			// if (is_null($doc2)) {
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `64%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
153			// $record = \SruClient::first('bs.objektid=' . $other_id);
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `44%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
154			// \Log::debug('Importing related record ' . $other_id);
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `50%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
155			// if (is_null($record)) {
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `64%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
156			// die('uh oh');
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `84%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
157			// } else {
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `50%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
158			// $this->import($record->data);
			0 ignored issues – show Unused Code Comprehensibility introduced 2016-07-13 19:13 UTC by Report Bug Copy Issue Report `70%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
159			// }
160			// }
161
162			// @TODO: Add a separate jobb that updates e-books weekly or so..
163			}
164
165			// Sync subjects
166			$subject_ids = [];
167			foreach ($biblio['subjects'] as $value) {
168			$value['vocabulary'] = $this->fixVocabularyCode($value['vocabulary']);
169			$subject = Subject::lookup($value['vocabulary'], $value['term'], $value['type']);
170			if (is_null($subject)) {
171			$subject = Subject::create($value);
172			}
173			$subject_ids[] = $subject->id;
174			}
175			$doc->subjects()->sync($subject_ids);
176
177			// Sync genres
178			$genre_ids = [];
179			foreach ($biblio['genres'] as $value) {
180			$genre = Genre::lookup($value['vocabulary'], $value['term']);
181			if (is_null($genre)) {
182			$genre = Genre::create($value);
183			}
184			$genre_ids[] = $genre->id;
185			}
186			$doc->genres()->sync($genre_ids);
187
188			// Extract cover from bibliographic record if no local cover exists
189			if (isset($biblio['cover_image']) && is_null($doc->cover)) {
190			try {
191			$doc->storeCover($biblio['cover_image']);
192			} catch (\ErrorException $e) {
			0 ignored issues – show Bug introduced 2016-12-13 22:28 UTC by Report Bug Copy Issue Report The class `ErrorException` does not exist. Did you forget a USE statement, or did you not list all dependencies? Scrutinizer analyzes your `composer.json`/`composer.lock` file if available to determine the classes, and functions that are defined by your dependencies. It seems like the listed class was neither found in your dependencies, nor was it found in the analyzed files in your repository. If you are using some other form of dependency management, you might want to disable this analysis. Loading history...
193			\Log::error('Failed to store cover: ' . $biblio['cover_image']);
194			}
195			}
196
197			return $doc;
198			}
199
200			/**
201			* Execute the job.
202			* @param QuiteSimpleXMLElement $record
203			* @return int\|null
204			*/
205			public function import(QuiteSimpleXMLElement $record)
206			{
207			try {
208			list($biblio, $holdings) = $this->parseRecord($record);
209			} catch (ParserException $e) {
210			$this->error('Failed to parse MARC record. Error "' . $e->getMessage() . '" in: ' . $e->getFile() . ':' . $e->getLine() . "\nStack trace:\n" . $e->getTraceAsString());
211
212			return null;
213			}
214
215			$doc = $this->importParsedRecord($biblio, $holdings);
216
217			\Log::debug('[Marc21Importer] Imported ' . $doc->bibsys_id . ' as ' . $doc->id);
218
219			if (!is_null($doc)) {
220			Event::fire(new Marc21RecordImported($doc->id));
221			}
222
223			return $doc->id;
224			}
225			}
226

scriptotek / colligator-backend

Marc21Importer::__construct() A last analyzed 2018-08-07 10:50 UTC

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like

Marc21Importer::__construct() A
last analyzed 2018-08-07 10:50 UTC