aboutsummaryrefslogtreecommitdiffstats
path: root/skate/schema.go
blob: a3f261db2c0294671bc57f2235ba82f69fe2351c (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
package skate

import (
	"bytes"
	"crypto/sha1"
	"encoding/gob"
	"fmt"
	"regexp"
	"strconv"
	"strings"
	"sync"
	"time"

	"git.archive.org/martin/cgraph/skate/isbn"
	"git.archive.org/martin/cgraph/skate/set"
)

var (
	// Various ISBN patterns tailored towards more dirty values, e.g. with
	// artifacts from OCR, etc.
	isbn10Regex = regexp.MustCompile(`[O0-9xX -]{10,18}`)
	isbn13Regex = regexp.MustCompile(`9[O0-9xX -]{12,20}`)

	// openLibraryDateLayouts, e.g. as found in Open Library Editions,
	// .publish_date.
	openLibraryDateLayouts = []string{
		"1 Jan 2006",
		"2006",
		"2006-02",
		"2006-02-01",
		"Jan 2, 2006",
		"Jan 2006",
	}

	rune16pool = sync.Pool{
		New: func() interface{} {
			return make([]rune, 0, 16)
		},
	}
	setPool = sync.Pool{
		New: func() interface{} {
			return set.New()
		},
	}
)

// Ref is a reference document (derived from "heavy intermediate schema"), can
// be very partial.
type Ref struct {
	Biblio       Biblio `json:"biblio"`
	Index        int64  `json:"index,omitempty"`
	Key          string `json:"key,omitempty"`
	Locator      string `json:"locator:omitempty"`
	RefSource    string `json:"ref_source,omitempty"`
	ReleaseYear  int    `json:"release_year,omitempty"`
	ReleaseIdent string `json:"release_ident,omitempty"`
	ReleaseStage string `json:"release_stage,omitempty"`
	WorkIdent    string `json:"work_ident,omitempty"`
}

// Biblio contains bibliographic information for a Ref.
type Biblio struct {
	ArxivId         string   `json:"arxiv_id,omitempty"`
	ContainerName   string   `json:"container_name,omitempty"`
	ContribRawNames []string `json:"contrib_raw_names,omitempty"`
	DOI             string   `json:"doi,omitempty"`
	Issue           string   `json:"issue,omitempty"`
	PMCID           string   `json:"pmcid,omitempty"`
	PMID            string   `json:"pmid,omitempty"`
	Pages           string   `json:"pages,omitempty"`
	Publisher       string   `json:"publisher,omitempty"`
	Title           string   `json:"title,omitempty"`
	Unstructured    string   `json:"unstructured,omitempty"`
	Url             string   `json:"url,omitempty"`
	Volume          string   `json:"volume,omitempty"`
	Year            int64    `json:"year,omitempty"`
	// More non-standard fields go into extra.
	Extra struct {
		ISBN []string `json:"isbn"`
	} `json:"extra"`
}

// RefToRelease converts a ref to a release.
func RefToRelease(ref *Ref) (*Release, error) {
	var (
		release  Release
		b        = ref.Biblio
		contribs = make([]struct {
			Index   int    `json:"index,omitempty"`
			RawName string `json:"raw_name,omitempty"`
			Role    string `json:"role,omitempty"`
		}, len(b.ContribRawNames))
	)
	release.Ident = ref.ReleaseIdent
	release.WorkID = ref.WorkIdent
	release.ExtIDs.Arxiv = b.ArxivId
	release.ExtIDs.DOI = b.DOI
	release.ExtIDs.PMID = b.PMID
	release.ExtIDs.PMCID = b.PMCID
	release.Title = b.Title
	release.Publisher = b.Publisher
	release.ContainerName = b.ContainerName
	release.Volume = b.Volume
	release.Issue = b.Issue
	release.Pages = b.Pages
	if ref.ReleaseYear > 1000 {
		release.ReleaseYearValue = fmt.Sprintf("%d", ref.ReleaseYear)
	}
	for i, name := range b.ContribRawNames {
		contribs[i].Index = i
		contribs[i].RawName = name
	}
	release.Contribs = contribs
	if strings.Contains(strings.ToLower(ref.Biblio.Unstructured), "isbn") {
		release.ExtIDs.ISBN = ParseIsbn(ref.Biblio.Unstructured)
	}
	// Extra info stashed into extra.skate.ref.
	release.Extra.Skate.Ref.Index = ref.Index
	release.Extra.Skate.Ref.Key = ref.Key
	release.Extra.Skate.Ref.Locator = ref.Locator
	release.Extra.Skate.Ref.Source = ref.RefSource
	return &release, nil
}

// ReleaseToUnstructured tries to render a sensible string, e.g. for frontend
// display of unmatched and other relations. Some examples:
// https://guides.lib.uw.edu/c.php?g=341448&p=4076094 No specific style, just
// try to be readable.
func ReleaseToUnstructured(r *Release) string {
	var (
		buf   bytes.Buffer
		names = make([]string, len(r.Contribs))
	)
	for i := 0; i < len(r.Contribs); i++ {
		names[i] = r.Contribs[i].RawName
	}
	fmt.Fprintf(&buf, "%s", strings.Join(names, ", "))
	if r.Title != "" {
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ". ")
		}
		fmt.Fprintf(&buf, `%s`, r.Title)
	}
	if len(r.Subtitle()) > 0 {
		fmt.Fprintf(&buf, ": %s", strings.Join(r.Subtitle(), " "))
	}
	if r.ContainerName != "" {
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ". ")
		}
		fmt.Fprintf(&buf, `%s`, r.ContainerName)
	}
	if r.Volume != "" {
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ", ")
		}
		fmt.Fprintf(&buf, `vol. %s`, r.Volume)
	}
	if r.Issue != "" {
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ", ")
		}
		fmt.Fprintf(&buf, `no. %s`, r.Issue)
	}
	if r.ReleaseYear() > 0 {
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ", ")
		}
		fmt.Fprintf(&buf, `%s`, r.ReleaseYearString())
	}
	if r.Pages != "" {
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ", ")
		}
		fmt.Fprintf(&buf, `pp. %s`, r.Pages)
	}
	if r.Publisher != "" {
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ", ")
		}
		fmt.Fprintf(&buf, `%s`, r.Publisher)
	}
	switch {
	case r.ExtIDs.DOI != "":
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ", ")
		}
		fmt.Fprintf(&buf, `%s`, r.ExtIDs.DOI)
	case len(r.ExtIDs.ISBN) > 0:
		if buf.Len() > 0 {
			fmt.Fprintf(&buf, ", ")
		}
		if isbn13, err := isbn.To13(r.ExtIDs.ISBN[0]); err == nil {
			fmt.Fprintf(&buf, `%s`, isbn13)
		} else {
			fmt.Fprintf(&buf, `%s`, r.ExtIDs.ISBN[0])
		}
	}
	return buf.String()
}

// ParseIsbn tries to find and validate ISBN from unstructured data. Returns a
// list of unique, unsorted and validated ISBN13, e.g. 9780123838520.
func ParseIsbn(s string) []string {
	// ISBN: 10: 0137822693, pp: 373
	// Robotec, E. (1996). Scorbot ER VII, User's Manual, Eshed Robotec,
	// ISBN9652910333. Shannon, C. (1948). A Mathematical Theory of
	// Communication. The Bell System Technical Journal. July; October,
	// Vol. 27, pp. 379-423; 623-656.
	// Artech House, ISBN: 978-1-60807-201-9, 2011.
	// ...
	var (
		candidates10 = isbn10Regex.FindAllString(s, -1)
		candidates13 = isbn13Regex.FindAllString(s, -1)
		u            []rune
		z            string
		err          error
	)
	valid := setPool.Get().(set.Set)
	valid.Clear()
	defer setPool.Put(valid)
	for _, v := range append(candidates10, candidates13...) {
		u = rune16pool.Get().([]rune)
		u = u[:0]
		for _, c := range v {
			if c == 'O' {
				c = '0'
			}
			if c >= '0' && c <= '9' || c == 'X' || c == 'x' {
				u = append(u, c)
			}
		}
		z = string(u)
		rune16pool.Put(u)
		if !isbn.Validate(z) {
			continue
		}
		if len(z) < 12 {
			if z, err = isbn.To13(z); err != nil {
				continue
			}
		}
		valid.Add(z)
	}
	return valid.Slice()
}

// Release document. Note that we may have different types for some fields
// (e.g. string, int, etc.).  Mitigation for now is to make the field an
// interface{}, name the field "...Value" and to add a method with the field
// name, doing type assertion.  Example: ReleaseYearValue interface{},
// ReleaseYear() int, etc.
//
// Extra field gets a section for "skate" for conversion related values.
type Release struct {
	ContainerID   string `json:"container_id,omitempty"`
	ContainerName string `json:"container_name,omitempty"` // when not resolved
	Container     struct {
		ContainerType string `json:"container_type"`
		Ident         string `json:"ident"`
		Issnl         string `json:"issnl"`
		Name          string `json:"name"`
		Publisher     string `json:"publisher"`
		Revision      string `json:"revision"`
		State         string `json:"state"`
		WikidataQid   string `json:"wikidata_qid"`
	} `json:"container"`
	Contribs []struct {
		Index   int    `json:"index,omitempty"`
		RawName string `json:"raw_name,omitempty"`
		Role    string `json:"role,omitempty"`
	} `json:"contribs,omitempty"`
	ExtIDs struct {
		Arxiv       string   `json:"arxiv,omitempty"`
		Core        string   `json:"core,omitempty"`
		DOI         string   `json:"doi,omitempty"`
		ISBN        []string `json:"isbn,omitempty"` // should be isbn13
		Jstor       string   `json:"jstor,omitempty"`
		OLID        string   `json:"olid,omitempty"`
		PMCID       string   `json:"pmcid,omitempty"`
		PMID        string   `json:"pmid,omitempty"`
		WikidataQID string   `json:"wikidata_qid,omitempty"`
	} `json:"ext_ids,omitempty"`
	Ident     string `json:"ident,omitempty"`
	Publisher string `json:"publisher,omitempty"`
	Refs      []struct {
		ContainerName string `json:"container_name,omitempty"`
		Extra         struct {
			DOI     string   `json:"doi,omitempty"`
			Authors []string `json:"authors,omitempty"`
			Key     string   `json:"key,omitempty"`
			Year    string   `json:"year,omitempty"`
			Locator string   `json:"locator,omitempty"`
			Volume  string   `json:"volume,omitempty"`
		} `json:"extra"`
		Index   int64  `json:"index,omitempty"`
		Key     string `json:"key,omitempty"`
		Locator string `json:"locator,omitempty"`
		Year    int64  `json:"year,omitempty"`
	} `json:"refs,omitempty"`
	ReleaseDate      string      `json:"release_date,omitempty"`
	ReleaseYearValue interface{} `json:"release_year,omitempty"` // might be int or str
	ReleaseStage     string      `json:"release_stage,omitempty"`
	ReleaseType      string      `json:"release_type,omitempty"`
	Issue            string      `json:"issue,omitempty"`
	Volume           string      `json:"volume,omitempty"`
	Pages            string      `json:"pages,omitempty"`
	Title            string      `json:"title,omitempty"`
	WorkID           string      `json:"work_id,omitempty"`
	Extra            struct {
		ContainerName string      `json:"container_name,omitempty"`
		SubtitleValue interface{} `json:"subtitle,omitempty"` // []str or str
		Crossref      struct {
			Type string `json:"type,omitempty"`
		} `json:"crossref,omitempty"`
		DataCite struct {
			MetadataVersion int                `json:"metadataVersion,omitempty"`
			Relations       []DataCiteRelation `json:"relations,omitempty"`
		} `json:"datacite,omitempty"`
		Skate struct {
			// Mark as converted from "ref", "rg" or other schemas.
			Status string `json:"status,omitempty"`
			// Carry the ref index and key around.
			Ref struct {
				Index   int64  `json:"index,omitempty"`
				Key     string `json:"key,omitempty"`
				Locator string `json:"locator,omitempty"`
				Source  string `json:"source,omitempty"`
			} `json:"ref,omitempty"`
			ResearchGate struct {
				URL string `json:"url,omitempty"`
			} `json:"rg,omitempty"`
			ResolvedContainerName string `json:"resolved_container_name"`
		} `json:"skate,omitempty"`
		OpenLibrary struct {
			HasFulltext   bool     `json:"has_fulltext,omitempty"`
			WorkID        string   `json:"work,omitempty"`
			SourceRecords []string `json:"source_records,omitempty"`
		} `json:"ol,omitempty"`
	} `json:"extra,omitempty"`
}

// Subtitle returns a slice of subtitle strings.
func (r *Release) Subtitle() (result []string) {
	switch v := r.Extra.SubtitleValue.(type) {
	case []interface{}:
		for _, e := range v {
			result = append(result, fmt.Sprintf("%v", e))
		}
		return result
	case []string:
		return v
	case string:
		return []string{v}
	}
	return []string{}
}

// ReleaseYearString returns release year as string.
func (r *Release) ReleaseYearString() string {
	return fmt.Sprintf("%d", r.ReleaseYear())
}

// ReleaseYear returns year as int, no further validity checks.
func (r *Release) ReleaseYear() int {
	switch v := r.ReleaseYearValue.(type) {
	case int:
		return v
	case float64:
		return int(v)
	case string:
		w, err := strconv.Atoi(v)
		if err != nil {
			return 0
		}
		return w
	default:
		return 0
	}
}

// DataCiteRelation as it appears in the release extra field.
type DataCiteRelation struct {
	RelatedIdentifierType  string      `json:"relatedIdentifierType,omitempty"`
	RelatedIdentifierValue interface{} `json:"relatedIdentifier,omitempty"`
}

// RelatedIdentifier returns the identifier as string.
func (r *DataCiteRelation) RelatedIdentifier() string {
	switch v := r.RelatedIdentifierValue.(type) {
	case string:
		return v
	default:
		return fmt.Sprintf("%v", v)
	}
}

// SitemapEntry in a basic JSON style, e.g. for https://archive.org/details/rg_sitemap_2021_02_23.
type SitemapEntry struct {
	Lastmod string `json:"lastmod,omitempty"`
	Title   string `json:"title,omitempty"`
	URL     string `json:"url,omitempty"`
}

// BiblioRef as a prototype for citation graph elasticsearch indexing,
// https://is.gd/yicTom.
type BiblioRef struct {
	Key                    string `json:"_id,omitempty"`
	IndexedTs              string `json:"indexed_ts,omitempty"` // https://www.elastic.co/guide/en/elasticsearch/reference/current/date.html
	SourceReleaseIdent     string `json:"source_release_ident,omitempty"`
	SourceWorkIdent        string `json:"source_work_ident,omitempty"`
	SourceWikipediaArticle string `json:"source_wikipedia_article,omitempty"`
	SourceReleaseStage     string `json:"source_release_stage,omitempty"`
	SourceYear             string `json:"source_year,omitempty"` // TODO: should this be source_release_year
	RefIndex               int64  `json:"ref_index,omitempty"`   // 1-based
	RefKey                 string `json:"ref_key,omitempty"`
	RefLocator             string `json:"ref_locator,omitempty"`
	TargetReleaseIdent     string `json:"target_release_ident,omitempty"`
	TargetWorkIdent        string `json:"target_work_ident,omitempty"`
	TargetOpenLibraryWork  string `json:"target_openlibrary_work,omitempty"`
	TargetURLSurt          string `json:"target_url_surt,omitempty"`
	TargetURL              string `json:"target_url,omitempty"`
	MatchProvenance        string `json:"match_provenance,omitempty"`
	MatchStatus            string `json:"match_status,omitempty"`
	MatchReason            string `json:"match_reason,omitempty"`
	TargetUnstructured     string `json:"target_unstructured,omitempty"`
	TargetCSL              *CSL   `json:"target_csl,omitempty"`
}

// CSL is a subset of citation style language schema.
// https://github.com/citation-style-language/schema,
// https://navneethg.github.io/jsonschemaviewer/. This is a subset only.
type CSL struct {
	Accessed            CSLDate     `json:"accessed,omitempty"`
	Author              []CSLAuthor `json:"author,omitempty"`
	CollectionTitle     string      `json:"collection-title,omitempty"`
	ContainerTitle      string      `json:"container-title,omitempty"`
	ContainerTitleShort string      `json:"container-title-short,omitempty"`
	DOI                 string      `json:"DOI,omitempty"`
	ID                  string      `json:"id,omitempty"`
	ISBN                string      `json:"ISBN,omitempty"`
	ISSN                string      `json:"ISSN,omitempty"`
	Issue               string      `json:"issue,omitempty"`
	Issued              CSLDate     `json:"issued,omitempty"`
	JournalAbbreviation string      `json:"journalAbbreviation,omitempty"`
	Language            string      `json:"language,omitempty"`
	NumberOfPages       string      `json:"number-of-pages,omitempty"`
	OriginalTitle       string      `json:"original-title,omitempty"`
	PMCID               string      `json:"PMCID,omitempty"`
	PMID                string      `json:"PMID,omitempty"`
	Page                string      `json:"page,omitempty"`
	PageFirst           string      `json:"page-first,omitempty"`
	Publisher           string      `json:"publisher,omitempty"`
	Source              string      `json:"source,omitempty"`
	Status              string      `json:"status,omitempty"`
	Title               string      `json:"title,omitempty"`
	TitleShort          string      `json:"title-short,omitempty"`
	Type                string      `json:"type,omitempty"`
	URL                 string      `json:"URL,omitempty"`
	Volume              string      `json:"volume,omitempty"`
	VolumeTitle         string      `json:"volume-title,omitempty"`
	VolumeTitleShort    string      `json:"volume-title-short,omitempty"`
	YearSuffix          string      `json:"year-suffix,omitempty"`
}

// A date field is a complex field that expresses a date or a range of dates.
// An example date field in CSL is issued, which identifies the date an item
// was issued or published. Date fields can be expressed in two different
// formats. The first format is an array format (note the double-nesting of the
// array). To express a date range in this format, the ending date would be set
// as a second array. The second date format is a raw string. The recommended
// encoding is a string that represents the date in a numberic year-month-day
// format.
type CSLDate struct {
	Raw   string  `json:"raw,omitempty"`
	Parts [][]int `json:"date-parts,omitempty"`
}

// Some personal names are represented by a single field (e.g. mononyms such as
// “Prince” or “Plato”). In such cases, the name can be delivered as a lone
// family element. Institutional names may be delivered in the same way, but it
// is preferred to set them instead as a literal element.
//
// We include Name, for holding unparsed name, which is not a literal.
type CSLAuthor struct {
	Family  string `json:"family,omitempty"`
	Given   string `json:"given,omitempty"`
	Literal string `json:"literal,omitempty"`
	Name    string `json:"name,omitempty"`
}

func (b *BiblioRef) Reset() {
	b.Key = ""
	b.IndexedTs = ""
	b.SourceReleaseIdent = ""
	b.SourceWorkIdent = ""
	b.SourceWikipediaArticle = ""
	b.SourceReleaseStage = ""
	b.SourceYear = ""
	b.RefIndex = 0
	b.RefKey = ""
	b.RefLocator = ""
	b.TargetReleaseIdent = ""
	b.TargetWorkIdent = ""
	b.TargetOpenLibraryWork = ""
	b.TargetURLSurt = ""
	b.TargetURL = ""
	b.MatchProvenance = ""
	b.MatchStatus = ""
	b.MatchReason = ""
	b.TargetUnstructured = ""
	b.TargetCSL = nil
}

// Hash returns a string that will be the same, if source and target are
// equal; different otherwise. This can be used to detect duplicate links.
func (b *BiblioRef) LinkHash() string {
	switch {
	case b.SourceReleaseIdent != "" && b.TargetReleaseIdent != "":
		return fmt.Sprintf("fc:%s--fc:%s", b.SourceReleaseIdent, b.TargetReleaseIdent)
	case b.SourceReleaseIdent != "" && b.TargetOpenLibraryWork != "":
		return fmt.Sprintf("fc:%s--ol:%s", b.SourceReleaseIdent, b.TargetOpenLibraryWork)
	case b.SourceReleaseIdent != "" && b.TargetURL != "":
		return fmt.Sprintf("fc:%s--wb:%s", b.SourceReleaseIdent, b.TargetURL)
	case b.SourceReleaseIdent != "" && b.TargetURLSurt != "":
		return fmt.Sprintf("fc:%s--wb:%s", b.SourceReleaseIdent, b.TargetURLSurt)
	default:
		var (
			buf bytes.Buffer
			enc = gob.NewEncoder(&buf)
			h   = sha1.New()
		)
		if err := enc.Encode(b); err != nil {
			return ""
		}
		if _, err := buf.WriteTo(h); err != nil {
			return ""
		}
		return fmt.Sprintf("%x", h.Sum(nil))
	}
}

// ReleaseCluster, a list of match candidates. This is typically serialized as a
// single JSON line containing the match key and a list of release documents.
//
// Deprecated, since we are moving to a "two stream" generic "join" style
// processing.
type ReleaseCluster struct {
	Key    string     `json:"k"`
	Values []*Release `json:"v"`
}

// OneNonRef returns the first non-reference release found in a cluster, or an
// error, if none has been found. This depends on converted references setting
// extra.skate.status to "ref" - we use this in mixed clusters (catalog entries
// and references converted into releases).
func (rc *ReleaseCluster) OneNonRef() (*Release, error) {
	for _, re := range rc.Values {
		if re.Extra.Skate.Status != "ref" {
			return re, nil
		}
	}
	return nil, fmt.Errorf("no reference/release found for cluster key: %v", rc.Key)
}

// MinimalCitations variant from
// archive.org/details/wikipedia_citations_2020-07-14. Part of the naming was
// already inconsistent.
type MinimalCitations struct {
	IDList         string `json:"ID_list"`
	PageTitle      string `json:"page_title"`
	Title          string `json:"Title"`
	TypeOfCitation string `json:"type_of_citation"`
	// We may have other languages in the future. If this is non, "en" might be
	// used as default.
	Language string `json:"lang"`
}

// IDList with commonly used identifier from wikipedia citations.
type IDList struct {
	ISBN  string `json:"isbn,omitempty"`
	DOI   string `json:"doi,omitempty"`
	PMID  string `json:"pmid,omitempty"`
	ISSN  string `json:"issn,omitempty"`
	JSTOR string `json:"jstor,omitempty"`
	PMC   string `json:"pmc,omitempty"`
	ARXIV string `json:"arxiv,omitempty"`
	OL    string `json:"ol,omitempty"`
}

// IsZero returns true, if none of the identifiers is defined.
func (l *IDList) IsZero() bool {
	return *l == IDList{}
}

// ParseIDList parses out the identifiers from a citation document, the IDList
// values look something like this: "{BIBCODE=1992ApJ...399L..31C,
// DOI=10.1086/186599}".
func (c *MinimalCitations) ParseIDList() (result IDList) {
	if len(c.IDList) < 3 {
		return result
	}
	var (
		s     = c.IDList[1 : len(c.IDList)-1] // get rid of "{" and "}"
		parts = strings.Split(s, ",")
		pair  []string
	)
	for _, part := range parts {
		pair = strings.Split(part, "=")
		if len(pair) != 2 {
			continue
		}
		pair[0] = strings.TrimSpace(pair[0])
		pair[1] = strings.TrimSpace(pair[1])
		switch pair[0] {
		case "ISBN":
			result.ISBN = pair[1]
		case "DOI":
			result.DOI = pair[1]
		case "PMID":
			result.PMID = pair[1]
		case "ISSN":
			result.ISSN = pair[1]
		case "PMC":
			result.PMC = pair[1]
		case "JSTOR":
			result.JSTOR = pair[1]
		case "ARXIV":
			result.ARXIV = pair[1]
		case "OL":
			result.OL = pair[1]
		default:
			continue
		}
	}
	return result
}

// OpenLibrarySolrDoc from data dump (solr). Note: we most likely only need OL
// editions.
type OpenLibrarySolrDoc struct {
	AuthorFacet      []string `json:"author_facet"`
	AuthorKey        []string `json:"author_key"`
	AuthorName       []string `json:"author_name"`
	CoverEditionKey  string   `json:"cover_edition_key"`
	CoverI           int64    `json:"cover_i"`
	EbookCountI      int64    `json:"ebook_count_i"`
	EditionCount     int64    `json:"edition_count"`
	EditionKey       []string `json:"edition_key"`
	FirstPublishYear int64    `json:"first_publish_year"`
	HasFulltext      bool     `json:"has_fulltext"`
	IdGoodreads      []string `json:"id_goodreads"`
	IdLibrarything   []string `json:"id_librarything"`
	Isbn             []string `json:"isbn"`
	Key              string   `json:"key"`
	Language         []string `json:"language"`
	LastModifiedI    int64    `json:"last_modified_i"`
	PublishDate      []string `json:"publish_date"`
	PublishYear      []int64  `json:"publish_year"`
	Publisher        []string `json:"publisher"`
	PublisherFacet   []string `json:"publisher_facet"`
	Seed             []string `json:"seed"`
	Text             []string `json:"text"`
	Title            string   `json:"title"`
	TitleSuggest     []string `json:"title_suggest"`
	Type             string   `json:"type"`
	Version          int64    `json:"_version_"`
}

// OpenLibrarySolrDocToRelease convert OL data into a release.
func OpenLibrarySolrDocToRelease(w *OpenLibrarySolrDoc) (*Release, error) {
	var (
		release  Release
		contribs = make([]struct {
			Index   int    `json:"index,omitempty"`
			RawName string `json:"raw_name,omitempty"`
			Role    string `json:"role,omitempty"`
		}, len(w.AuthorName))
		s = set.New()
	)
	for i, author := range w.AuthorName {
		contribs[i].RawName = author
	}
	release.Title = w.Title
	if len(w.PublishYear) > 0 {
		release.ReleaseYearValue = w.FirstPublishYear
	}
	for _, v := range w.Isbn {
		switch {
		case len(v) < 13:
			if w, err := isbn.To13(v); err == nil {
				s.Add(w)
			}
		default:
			s.Add(v)
		}
	}
	if len(w.Publisher) > 0 {
		release.Publisher = w.Publisher[0]
	}
	release.ExtIDs.ISBN = s.Slice()
	release.ExtIDs.OLID = strings.Replace(w.Key, "/works/", "", 1)
	release.Extra.OpenLibrary.HasFulltext = w.HasFulltext
	return &release, nil
}

// OpenLibraryEdition document, see: https://openlibrary.org/developers/dumps.
type OpenLibraryEdition struct {
	Authors []struct {
		Key string `json:"key"`
	} `json:"authors"`
	ByStatement string  `json:"by_statement"`
	Covers      []int64 `json:"covers"`
	Created     struct {
		Type  string `json:"type"`
		Value string `json:"value"`
	} `json:"created"`
	DeweyDecimalClass []string `json:"dewey_decimal_class"`
	Genres            []string `json:"genres"`
	Identifiers       struct {
		Goodreads    []string `json:"goodreads"`
		Librarything []string `json:"librarything"`
	} `json:"identifiers"`
	Isbn10    []string `json:"isbn_10"`
	Isbn13    []string `json:"isbn_13"`
	Key       string   `json:"key"`
	Languages []struct {
		Key string `json:"key"`
	} `json:"languages"`
	LastModified struct {
		Type  string `json:"type"`
		Value string `json:"value"`
	} `json:"last_modified"`
	LatestRevision    int64    `json:"latest_revision"`
	LcClassifications []string `json:"lc_classifications"`
	Lccn              []string `json:"lccn"`

	// null
	// {
	//   "type": "/type/text",
	//   "value": "Includes bibliographical references (p. 137-143)."
	// }
	// null
	// "Includes bibliographical references (p. 203-205) and index."
	// null
	Notes interface{} `json:"notes"`

	NumberOfPages  int64    `json:"number_of_pages"`
	Ocaid          string   `json:"ocaid"`
	Pagination     string   `json:"pagination"`
	PublishCountry string   `json:"publish_country"`
	PublishDate    string   `json:"publish_date"`
	PublishPlaces  []string `json:"publish_places"`
	Publishers     []string `json:"publishers"`
	Revision       int64    `json:"revision"`
	SourceRecords  []string `json:"source_records"`
	SubjectPlace   []string `json:"subject_place"`
	Subjects       []string `json:"subjects"`
	Subtitle       string   `json:"subtitle"`
	Title          string   `json:"title"`
	Type           struct {
		Key string `json:"key"`
	} `json:"type"`
	Works []struct {
		Key string `json:"key"`
	} `json:"works"`
}

// Isbns returns all found ISBN: unique, sorted, non-normalized.
func (v OpenLibraryEdition) Isbns() []string {
	s := set.New()
	for _, w := range v.Isbn10 {
		s.Add(w)
	}
	for _, w := range v.Isbn13 {
		s.Add(w)
	}
	return s.Sorted()
}

// OpenLibraryEditionToRelease convert OL edition record into a release. Takes
// a (potentially empty) author map to replace author keys (like
// "/author/OL18273A") with names.
func OpenLibraryEditionToRelease(v *OpenLibraryEdition, authorMap map[string]string) (*Release, error) {
	var (
		release  Release
		contribs = make([]struct {
			Index   int    `json:"index,omitempty"`
			RawName string `json:"raw_name,omitempty"`
			Role    string `json:"role,omitempty"`
		}, len(v.Authors))
	)
	for i, author := range v.Authors {
		// /authors/OL1076839A     Ralph S. Hattox
		// /authors/OL1054947A     Conferenza economica dell'area pisana (1984 Pisa, Italy)
		// /authors/OL1069856A     Terry S. Boutet
		// /authors/OL1077217A     Pamela Beck
		//  ...
		name, ok := authorMap[author.Key]
		if !ok {
			continue
		}
		contribs[i].RawName = name
	}
	release.Ident = v.Key
	release.Contribs = contribs
	if v.Subtitle != "" {
		release.Title = fmt.Sprintf("%s: %s", v.Title, v.Subtitle)
		release.Extra.SubtitleValue = v.Subtitle
	} else {
		release.Title = v.Title
	}
	release.ExtIDs.ISBN = v.Isbns()
	if len(v.Publishers) > 0 {
		release.Publisher = v.Publishers[0]
	}
	if len(v.Works) > 0 {
		// TODO: We should be ok with just the basename, e.g. just "OL10896658M"
		release.Extra.OpenLibrary.WorkID = v.Works[0].Key
		release.WorkID = v.Works[0].Key
	}
	// e.g.
	// "source_records": [
	//   "amazon:0531203093",
	//   "ia:multiplebirths0000land",
	//   "marc:marc_loc_2016/BooksAll.2016.part25.utf8:104915596:921"
	// ]
	release.Extra.OpenLibrary.SourceRecords = v.SourceRecords
	for _, l := range openLibraryDateLayouts {
		t, err := time.Parse(l, v.PublishDate)
		if err != nil {
			continue
		}
		release.ReleaseYearValue = t.Year()
		break
	}
	return &release, nil
}

// ShortenOpenLibraryIdentifier shortens an id like "/books/OL10899962M" to
// "OL10899962M" or returns the same string, if no simplifications can be made.
// TODO: add test.
func ShortenOpenLibraryIdentifier(s string) string {
	fields := strings.Split(s, "/")
	for _, f := range fields {
		if strings.HasPrefix(f, "OL") {
			return f
		}
	}
	return s
}