File size: 11,012 Bytes
4a51f94
 
f41aded
 
 
4a51f94
f41aded
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4a51f94
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f41aded
 
 
 
 
4a51f94
f41aded
 
 
 
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
// Kaynakça araması: OpenAlex, Crossref, arXiv ve OSF Preprints sonuçlarını BibTeX kaydına dönüştürür.
// Sonuçlar hem "Kaynak ara" panelinde hem de asistanın kaynak arama aracında kullanılır.
import { Router } from 'express';
import { HttpError, requireString } from './util.js';

const HEADERS = { 'User-Agent': 'Thoth-LaTeX/0.2' };

const TR_FOLD = { ç: 'c', ğ: 'g', ı: 'i', ö: 'o', ş: 's', ü: 'u' };
const asciiFold = (s) =>
  s.toLowerCase().replace(/[çğıöşü]/g, (c) => TR_FOLD[c]).normalize('NFD').replace(/[^a-z0-9]/g, '');

const STOP_WORDS = new Set(['a', 'an', 'the', 'on', 'of', 'in', 'for', 'and', 'to', 'bir', 've', 'ile', 'icin', 'uzerine']);

export function citeKey(authors, year, title) {
  const lastName = asciiFold((authors[0] ?? 'anonim').split(',')[0].trim().split(/\s+/).pop() ?? '');
  const word = title.split(/\s+/).map(asciiFold).find((w) => w.length > 2 && !STOP_WORDS.has(w)) ?? '';
  return `${lastName || 'anonim'}${year ?? ''}${word}`;
}

const escapeTex = (s) => String(s).replace(/([&%#_$])/g, '\\$1');

export function toBibtex(type, key, fields) {
  const lines = Object.entries(fields)
    .filter(([, v]) => v != null && v !== '')
    .map(([k, v]) => `  ${k} = {${k === 'url' || k === 'doi' || k === 'eprint' ? v : escapeTex(v)}}`);
  return `@${type}{${key},\n${lines.join(',\n')}\n}`;
}

const stripTags = (s) => s.replace(/<[^>]+>/g, '').replace(/\s+/g, ' ').trim();

const CROSSREF_TYPES = {
  'journal-article': 'article',
  'proceedings-article': 'inproceedings',
  book: 'book',
  monograph: 'book',
  'book-chapter': 'incollection',
  dissertation: 'phdthesis',
};

async function searchCrossref(query) {
  const url = new URL('https://api.crossref.org/works');
  url.searchParams.set('query.bibliographic', query);
  url.searchParams.set('rows', '12');
  url.searchParams.set('select', 'DOI,title,author,issued,container-title,type,volume,issue,page,publisher,URL');
  const res = await fetch(url, { headers: HEADERS, signal: AbortSignal.timeout(15_000) });
  if (!res.ok) throw new Error(`Crossref HTTP ${res.status}`);
  const { message } = await res.json();
  return message.items.filter((it) => it.title?.[0]).map((it) => {
    const title = stripTags(it.title[0]);
    const authors = (it.author ?? []).map((a) => (a.family ? `${a.family}, ${a.given ?? ''}`.replace(/, $/, '') : a.name)).filter(Boolean);
    const year = it.issued?.['date-parts']?.[0]?.[0] ?? null;
    const venue = it['container-title']?.[0] ? stripTags(it['container-title'][0]) : null;
    const type = CROSSREF_TYPES[it.type] ?? 'misc';
    const key = citeKey(authors, year, title);
    return {
      id: it.DOI, key, title, authors, year, venue, url: it.URL,
      bibtex: toBibtex(type, key, {
        title,
        author: authors.join(' and '),
        [type === 'article' ? 'journal' : 'booktitle']: venue,
        year,
        volume: it.volume,
        number: it.issue,
        pages: it.page?.replace('-', '--'),
        publisher: type === 'article' ? null : it.publisher,
        doi: it.DOI,
        url: it.URL,
      }),
    };
  });
}

const decodeXml = (s) =>
  s.replace(/&lt;/g, '<').replace(/&gt;/g, '>').replace(/&quot;/g, '"').replace(/&apos;/g, "'").replace(/&amp;/g, '&');
const xmlTag = (xml, tag) => {
  const m = xml.match(new RegExp(`<${tag}[^>]*>([\\s\\S]*?)</${tag}>`));
  return m ? decodeXml(m[1]).replace(/\s+/g, ' ').trim() : null;
};

async function searchArxiv(query) {
  const url = new URL('https://export.arxiv.org/api/query');
  // Her kelime ayrı ayrı aranır; aksi hâlde "all:" yalnızca ilk kelimeye uygulanır.
  url.searchParams.set('search_query', query.split(/\s+/).map((w) => `all:${w.replace(/[":()]/g, '')}`).join(' AND '));
  url.searchParams.set('max_results', '12');
  url.searchParams.set('sortBy', 'relevance');
  const res = await fetch(url, { headers: HEADERS, signal: AbortSignal.timeout(15_000) });
  if (!res.ok) throw new Error(`arXiv HTTP ${res.status}`);
  const xml = await res.text();
  return xml.split('<entry>').slice(1).map((entry) => {
    const absUrl = xmlTag(entry, 'id');
    const arxivId = absUrl.replace(/^https?:\/\/arxiv\.org\/abs\//, '').replace(/v\d+$/, '');
    const title = xmlTag(entry, 'title');
    const authors = [...entry.matchAll(/<name>([\s\S]*?)<\/name>/g)].map((m) => decodeXml(m[1]).trim());
    const year = Number(xmlTag(entry, 'published')?.slice(0, 4)) || null;
    const key = citeKey(authors, year, title);
    return {
      id: arxivId, key, title, authors, year, venue: `arXiv:${arxivId}`, url: absUrl,
      bibtex: toBibtex('misc', key, {
        title,
        author: authors.join(' and '),
        year,
        eprint: arxivId,
        archivePrefix: 'arXiv',
        url: absUrl,
      }),
    };
  });
}

/** OpenAlex'in sözcük -> konum dizininden özet metnini yeniden kurar. */
export function abstractFromIndex(index) {
  if (!index) return null;
  const words = [];
  for (const [word, positions] of Object.entries(index)) for (const p of positions) words[p] = word;
  return words.filter(Boolean).join(' ') || null;
}

const OPENALEX_TYPES = {
  article: 'article',
  review: 'article',
  editorial: 'article',
  letter: 'article',
  'book-chapter': 'incollection',
  book: 'book',
  dissertation: 'phdthesis',
  preprint: 'misc',
  report: 'report',
};

async function searchOpenAlex(query, { enCokAtif = false } = {}) {
  const url = new URL('https://api.openalex.org/works');
  url.searchParams.set('search', query);
  url.searchParams.set('per-page', '12');
  if (enCokAtif) url.searchParams.set('sort', 'cited_by_count:desc');
  url.searchParams.set('select',
    'id,doi,title,authorships,publication_year,primary_location,type,biblio,abstract_inverted_index,language');
  // İsteğe bağlı kişisel anahtar: OpenAlex anahtarsız istekleri ortak bir günlük kotayla sınırlar.
  if (process.env.OPENALEX_API_KEY) url.searchParams.set('api_key', process.env.OPENALEX_API_KEY);
  const res = await fetch(url, { headers: HEADERS, signal: AbortSignal.timeout(20_000) });
  if (!res.ok) throw new Error(`OpenAlex HTTP ${res.status}`);
  const { results } = await res.json();
  return results.filter((w) => w.title).map((w) => {
    const title = stripTags(w.title);
    // OpenAlex adları "Ad Soyad" biçimindedir; BibTeX için "Soyad, Ad" yapılır.
    const authors = (w.authorships ?? []).map((a) => {
      const name = a.author?.display_name ?? '';
      const parts = name.trim().split(/\s+/);
      return parts.length > 1 ? `${parts.pop()}, ${parts.join(' ')}` : name;
    }).filter(Boolean);
    const year = w.publication_year ?? null;
    const venue = w.primary_location?.source?.display_name ?? null;
    // Türü bilinmeyen ama bir dergide yayımlanmış kayıtlar makale sayılır.
    const type = OPENALEX_TYPES[w.type] ?? (w.primary_location?.source?.type === 'journal' ? 'article' : 'misc');
    const doi = w.doi ? w.doi.replace(/^https?:\/\/doi\.org\//, '') : null;
    const key = citeKey(authors, year, title);
    const b = w.biblio ?? {};
    const pages = b.first_page ? (b.last_page && b.last_page !== b.first_page ? `${b.first_page}--${b.last_page}` : b.first_page) : null;
    return {
      id: w.id.replace('https://openalex.org/', ''), key, title, authors, year, venue,
      url: doi ? `https://doi.org/${doi}` : w.id, doi,
      abstract: abstractFromIndex(w.abstract_inverted_index),
      bibtex: toBibtex(type, key, {
        title,
        author: authors.join(' and '),
        [type === 'article' ? 'journal' : 'booktitle']: type === 'misc' || type === 'book' || type === 'phdthesis' || type === 'report' ? null : venue,
        year,
        volume: b.volume,
        number: b.issue,
        pages,
        doi,
        url: doi ? null : w.id,
        langid: w.language === 'tr' ? 'turkish' : w.language === 'en' ? 'english' : null,
      }),
    };
  });
}

async function searchOsf(query) {
  // OSF yalnızca başlıkta bitişik ifade arar: en uzun anlamlı sözcükle aranır, sonuçlar
  // diğer sözcükleri de başlığında taşıyanlara göre sıralanır.
  const words = query.toLocaleLowerCase('tr').split(/[^\p{L}\p{N}-]+/u).filter((w) => w.length > 2 && !STOP_WORDS.has(w));
  const anchor = [...words].sort((a, b) => b.length - a.length)[0] ?? query;
  const url = new URL('https://api.osf.io/v2/preprints/');
  url.searchParams.set('filter[title]', anchor);
  url.searchParams.set('page[size]', '50');
  url.searchParams.set('embed', 'contributors');
  const res = await fetch(url, { headers: HEADERS, signal: AbortSignal.timeout(20_000) });
  if (!res.ok) throw new Error(`OSF HTTP ${res.status}`);
  const { data } = await res.json();
  return data.map((p) => {
    const a = p.attributes;
    const title = stripTags(a.title ?? '');
    const authors = (p.embeds?.contributors?.data ?? [])
      .filter((c) => c.attributes?.bibliographic !== false)
      .map((c) => c.embeds?.users?.data?.attributes)
      .filter(Boolean)
      .map((u) => (u.family_name ? `${u.family_name}, ${u.given_name ?? ''}`.replace(/, $/, '') : u.full_name));
    const year = Number((a.date_published ?? a.date_created ?? '').slice(0, 4)) || null;
    const provider = p.relationships?.provider?.data?.id ?? 'osf';
    const doi = (a.doi ?? p.links?.preprint_doi ?? '').replace(/^https?:\/\/doi\.org\//, '') || null;
    const key = citeKey(authors, year, title);
    return {
      id: `osf:${p.id}`, key, title, authors, year, venue: `${provider} (önbaskı)`,
      url: p.links?.html ?? `https://osf.io/${p.id}`, doi,
      abstract: a.description ? stripTags(a.description) : null,
      bibtex: toBibtex('misc', key, {
        title,
        author: authors.join(' and '),
        year,
        howpublished: `${provider} preprint`,
        doi,
        url: doi ? null : p.links?.html,
      }),
    };
  }).filter((r) => r.title)
    .map((r) => ({ r, score: words.filter((w) => r.title.toLocaleLowerCase('tr').includes(w)).length }))
    .filter(({ score }) => score >= Math.min(2, words.length))
    .sort((a, b) => b.score - a.score || (b.r.year ?? 0) - (a.r.year ?? 0))
    .slice(0, 12)
    .map(({ r }) => r);
}

export const SOURCES = {
  openalex: { label: 'OpenAlex', search: searchOpenAlex },
  crossref: { label: 'Crossref', search: searchCrossref },
  arxiv: { label: 'arXiv', search: searchArxiv },
  osf: { label: 'OSF Preprints', search: searchOsf },
};

/** Kaynak araması (panel ve asistan için ortak). */
export async function searchLiterature(query, source = 'openalex', options = {}) {
  const s = SOURCES[source] ?? SOURCES.openalex;
  return s.search(query, options);
}

export const citationsRouter = Router();

citationsRouter.get('/citations/search', async (req, res) => {
  const query = requireString(req.query.q, 'Arama metni', { min: 2, max: 300 });
  try {
    res.json({ results: await searchLiterature(query, req.query.source) });
  } catch (err) {
    throw new HttpError(502, `Kaynak araması başarısız oldu: ${err.message}`);
  }
});