Skip to content

Commit 1e77a0a

Browse files
zuphilipzuphilip
committed
New Translator SORA - Search Ornithological Research Archive
 * closes zotero#827 * asked https://forums.zotero.org/discussion/42937/translator-suggestion-sora/
1 parent 68283be commit 1e77a0a

File tree

1 file changed

+343
-0
lines changed

1 file changed

+343
-0
lines changed

Diff for: ‎SORA.js

+343
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,343 @@
1+
{
2+
"translatorID": "83d2ed27-40a5-4dc7-bd87-baddc8fb35da",
3+
"label": "SORA",
4+
"creator": "Philipp Zumstein",
5+
"target": "^https?://sora\\.unm\\.edu/(node/|search/node/|advancedsearch\\?)",
6+
"minVersion": "3.0",
7+
"maxVersion": "",
8+
"priority": 100,
9+
"inRepository": true,
10+
"translatorType": 4,
11+
"browserSupport": "gcsbv",
12+
"lastUpdated": "2015-01-02 22:03:36"
13+
}
14+
15+
/*
16+
***** BEGIN LICENSE BLOCK *****
17+
18+
Copyright © 2014 Philipp Zumstein
19+
20+
This file is part of Zotero.
21+
22+
Zotero is free software: you can redistribute it and/or modify
23+
it under the terms of the GNU Affero General Public License as published by
24+
the Free Software Foundation, either version 3 of the License, or
25+
(at your option) any later version.
26+
27+
Zotero is distributed in the hope that it will be useful,
28+
but WITHOUT ANY WARRANTY; without even the implied warranty of
29+
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
30+
GNU Affero General Public License for more details.
31+
32+
You should have received a copy of the GNU Affero General Public License
33+
along with Zotero. If not, see <http://www.gnu.org/licenses/>.
34+
35+
***** END LICENSE BLOCK *****
36+
*/
37+
38+
var mapping = {
39+
"journal" : "publicationTitle",
40+
"volume" : "volume",
41+
"issue" : "issue",
42+
"year" : "date",
43+
"pages" : "pages",
44+
"section" : "extra"
45+
}
46+
47+
48+
function detectWeb(doc, url) {
49+
if ( ZU.xpath(doc, '//div[contains(@class, "content")]/fieldset/legend').length ) { // Publication Information
50+
if (getSearchResults(doc, true)) {
51+
return "multiple";
52+
} else {
53+
return "journalArticle";
54+
}
55+
}
56+
if (url.indexOf('search')>-1 && getSearchResults(doc, true)) {
57+
return "multiple";
58+
}
59+
}
60+
61+
function getSearchResults(doc, checkOnly) {
62+
var items = {};
63+
var found = false;
64+
var rows = ZU.xpath(doc, '//div[contains(@class, "content")]//dt//a|//strong[contains(@class, "field-content")]/a');
65+
for (var i=0; i<rows.length; i++) {
66+
var href = rows[i].href;
67+
var title = ZU.trimInternal(rows[i].textContent);
68+
if (!href || !title) continue;
69+
if (checkOnly) return true;
70+
found = true;
71+
items[href] = title;
72+
}
73+
return found ? items : false;
74+
}
75+
76+
function doWeb(doc, url) {
77+
if (detectWeb(doc, url) == "multiple") {
78+
Zotero.selectItems(getSearchResults(doc, false), function (items) {
79+
if (!items) {
80+
return true;
81+
}
82+
var articles = new Array();
83+
for (var i in items) {
84+
articles.push(i);
85+
}
86+
ZU.processDocuments(articles, scrape);
87+
});
88+
} else {
89+
scrape(doc, url);
90+
}
91+
}
92+
93+
94+
function scrape(doc, url) {
95+
var item = new Zotero.Item(detectWeb(doc, url));
96+
97+
//title
98+
var title = ZU.xpathText(doc, '//h1[contains(@class, "title")]');
99+
item.title = title;
100+
101+
//authors
102+
//Examples:
103+
//Edward H. Burtt, Jr.
104+
//James Hengeveld, Keith A. Mcmullen, Geoffrey A. Williamson
105+
//==> The author string is splitted by ',' into the individual
106+
//authors, but for suffixes this will be corrected again by
107+
//some heuristic. This heuristic seperates another autor
108+
//from a suffix by checking if the string consists any space
109+
//and its length is less than 5.
110+
var authors = ZU.xpathText(doc, '//div[contains(@class, "content")]/div[contains(@class, "field-field-authors")]/div/div/text()');
111+
if (authors) {
112+
var authorsSplit = authors.split(',');
113+
var index = 0;
114+
for (var m=0; m<authorsSplit.length; m++) {
115+
var value = ZU.trim(authorsSplit[m]);
116+
if (value != '') {
117+
if (value.indexOf(' ') == -1 && value.length<5 && index>0) {
118+
item.creators[index-1].firstName += ', ' + value;//Jr. or III.
119+
} else {
120+
item.creators.push( ZU.cleanAuthor( value, "author") );
121+
index++;
122+
}
123+
124+
}
125+
}
126+
}
127+
128+
//other fields
129+
var fields = ZU.xpath(doc, '//div[contains(@class, "content")]/fieldset/div');
130+
for (var k=0; k<fields.length; k++) {
131+
var fieldName = ZU.trim( ZU.xpathText(fields[k], './div/div/div').replace(':','').toLowerCase() );
132+
var fieldValue = ZU.trim( ZU.xpathText(fields[k], './div/div/text()').replace(',','') );
133+
if (mapping[fieldName]) {
134+
item[ mapping[fieldName] ] = fieldValue;
135+
} else {
136+
Z.debug('Unrecognized field: ' + fieldName);
137+
}
138+
}
139+
140+
141+
//PDF
142+
var attachments = doc.getElementById('attachments');
143+
if (attachments) {
144+
var link = attachments.getElementsByTagName('a');
145+
if (link.length) {
146+
item.attachments.push({
147+
title : "Full Text PDF",
148+
url : link[0].href,
149+
mimeType : "application/pdf"
150+
});
151+
}
152+
}
153+
//link to SORA entry
154+
item.attachments.push({
155+
title : "SORA Entry",
156+
url : url,
157+
mimeType : "text/html",
158+
snapshot : false
159+
});
160+
161+
item.complete();
162+
}/** BEGIN TEST CASES **/
163+
var testCases = [
164+
{
165+
"type": "web",
166+
"url": "https://sora.unm.edu/node/99151",
167+
"items": [
168+
{
169+
"itemType": "journalArticle",
170+
"title": "The Passing of Coragyps Shastensis Miller",
171+
"creators": [
172+
{
173+
"firstName": "Loye",
174+
"lastName": "Miller",
175+
"creatorType": "author"
176+
}
177+
],
178+
"date": "1941",
179+
"issue": "3 (May-June)",
180+
"libraryCatalog": "SORA",
181+
"pages": "140-141",
182+
"publicationTitle": "Condor",
183+
"volume": "43",
184+
"attachments": [
185+
{
186+
"title": "Full Text PDF",
187+
"mimeType": "application/pdf"
188+
},
189+
{
190+
"title": "SORA Entry",
191+
"mimeType": "text/html",
192+
"snapshot": false
193+
}
194+
],
195+
"tags": [],
196+
"notes": [],
197+
"seeAlso": []
198+
}
199+
]
200+
},
201+
{
202+
"type": "web",
203+
"url": "https://sora.unm.edu/node/2035",
204+
"items": "multiple"
205+
},
206+
{
207+
"type": "web",
208+
"url": "https://sora.unm.edu/node/116567",
209+
"items": [
210+
{
211+
"itemType": "journalArticle",
212+
"title": "Illinois and Indiana",
213+
"creators": [
214+
{
215+
"firstName": "James",
216+
"lastName": "Hengeveld",
217+
"creatorType": "author"
218+
},
219+
{
220+
"firstName": "Keith A.",
221+
"lastName": "Mcmullen",
222+
"creatorType": "author"
223+
},
224+
{
225+
"firstName": "Geoffrey A.",
226+
"lastName": "Williamson",
227+
"creatorType": "author"
228+
}
229+
],
230+
"date": "2006",
231+
"issue": "3",
232+
"libraryCatalog": "SORA",
233+
"pages": "376-378",
234+
"publicationTitle": "North American Birds",
235+
"volume": "60",
236+
"attachments": [
237+
{
238+
"title": "Full Text PDF",
239+
"mimeType": "application/pdf"
240+
},
241+
{
242+
"title": "SORA Entry",
243+
"mimeType": "text/html",
244+
"snapshot": false
245+
}
246+
],
247+
"tags": [],
248+
"notes": [],
249+
"seeAlso": []
250+
}
251+
]
252+
},
253+
{
254+
"type": "web",
255+
"url": "https://sora.unm.edu/node/162",
256+
"items": [
257+
{
258+
"itemType": "journalArticle",
259+
"title": "An Analysis of Physical, Physiological, and Optical Aspects of Avian Coloration With Emphasis On Wood-Warblers",
260+
"creators": [
261+
{
262+
"firstName": "Edward H., Jr.",
263+
"lastName": "Burtt",
264+
"creatorType": "author"
265+
}
266+
],
267+
"date": "1986",
268+
"issue": "38",
269+
"libraryCatalog": "SORA",
270+
"pages": "1-126",
271+
"publicationTitle": "Ornithological Monographs",
272+
"attachments": [
273+
{
274+
"title": "Full Text PDF",
275+
"mimeType": "application/pdf"
276+
},
277+
{
278+
"title": "SORA Entry",
279+
"mimeType": "text/html",
280+
"snapshot": false
281+
}
282+
],
283+
"tags": [],
284+
"notes": [],
285+
"seeAlso": []
286+
}
287+
]
288+
},
289+
{
290+
"type": "web",
291+
"url": "https://sora.unm.edu/advancedsearch?field_authors_value=&title=condor&field_fulltext_value=&field_year_value=",
292+
"items": "multiple"
293+
},
294+
{
295+
"type": "web",
296+
"url": "https://sora.unm.edu/search/node/eagle%20type%3Aarticle",
297+
"items": "multiple"
298+
},
299+
{
300+
"type": "web",
301+
"url": "https://sora.unm.edu/node/53338",
302+
"items": [
303+
{
304+
"itemType": "journalArticle",
305+
"title": "A Comparison of Home Range Estimates for a Bald Eagle Wintering in New Mexico",
306+
"creators": [
307+
{
308+
"firstName": "Dale W.",
309+
"lastName": "Stahlecker",
310+
"creatorType": "author"
311+
},
312+
{
313+
"firstName": "Timothy G.",
314+
"lastName": "Smith",
315+
"creatorType": "author"
316+
}
317+
],
318+
"date": "1993",
319+
"extra": "Short Communications",
320+
"issue": "1 (March)",
321+
"libraryCatalog": "SORA",
322+
"pages": "42-45",
323+
"publicationTitle": "Journal of Raptor Research",
324+
"volume": "27",
325+
"attachments": [
326+
{
327+
"title": "Full Text PDF",
328+
"mimeType": "application/pdf"
329+
},
330+
{
331+
"title": "SORA Entry",
332+
"mimeType": "text/html",
333+
"snapshot": false
334+
}
335+
],
336+
"tags": [],
337+
"notes": [],
338+
"seeAlso": []
339+
}
340+
]
341+
}
342+
]
343+
/** END TEST CASES **/

0 commit comments

Comments
 (0)