Skip to content

Commit de54781

Browse files
alanorthalanorth
committed
Add new RemoveGeneratedThumbnails script
1 parent 13c6612 commit de54781

File tree

3 files changed

+151
-0
lines changed

3 files changed

+151
-0
lines changed

‎README.md‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -4,6 +4,7 @@ DSpace curation tasks and other Java-based helpers used on the [CGSpace](https:/
44
- **CountryCodeTagger**: add ISO 3166-1 Alpha2 country codes to items based on their existing country metadata
55
- **FixJpgJpgThumbnails**: fix low-quality ".jpg.jpg" thumbnails by replacing them with their originals
66
- **FixLowQualityThumbnails**: remove low-quality thumbnails when PDF bitstreams are present
7+
- **RemoveGeneratedThumbnails**: remove generated thumbnails (in preparation for re-generating)
78
- **NormalizeDOIs**: normalize DOIs by stripping whitespace, lowercasing, and converting to https://doi.org/ format
89

910
Tested on DSpace 7.6.3. Read more about the [DSpace curation system](https://wiki.lyrasis.org/display/DSDOC7x/Curation+System).

‎src/main/java/io/github/ilri/cgspace/scripts/README.md‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3,6 +3,7 @@ Java-based helpers used on the [CGSpace](https://cgspace.cgiar.org) institutiona
33

44
- **FixJpgJpgThumbnails**: fix low-quality ".jpg.jpg" thumbnails by replacing them with their originals
55
- **FixLowQualityThumbnails**: remove low-quality thumbnails when PDF bitstreams are present
6+
- **RemoveGeneratedThumbnails**: remove generated thumbnails (in preparation for re-generating)
67

78
Tested on DSpace 7.6.3. Read more about the [DSpace curation system](https://wiki.lyrasis.org/display/DSDOC7x/Curation+System).
89

‎src/main/java/io/github/ilri/cgspace/scripts/RemoveGeneratedThumbnails.java‎

Lines changed: 149 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,149 @@
1+
/*
2+
* Copyright (C) 2025 Alan Orth
3+
*
4+
* SPDX-License-Identifier: GPL-3.0-only
5+
*/
6+
7+
package io.github.ilri.cgspace.scripts;
8+
9+
import org.apache.commons.lang.StringUtils;
10+
import org.dspace.authorize.AuthorizeException;
11+
import org.dspace.content.Bitstream;
12+
import org.dspace.content.Bundle;
13+
import org.dspace.content.Collection;
14+
import org.dspace.content.Community;
15+
import org.dspace.content.DSpaceObject;
16+
import org.dspace.content.Item;
17+
import org.dspace.content.MetadataValue;
18+
import org.dspace.content.factory.ContentServiceFactory;
19+
import org.dspace.content.service.BundleService;
20+
import org.dspace.content.service.ItemService;
21+
import org.dspace.core.Constants;
22+
import org.dspace.core.Context;
23+
import org.dspace.handle.factory.HandleServiceFactory;
24+
import org.dspace.handle.service.HandleService;
25+
26+
import java.io.IOException;
27+
import java.sql.SQLException;
28+
import java.util.Iterator;
29+
import java.util.List;
30+
31+
/**
32+
* @author Andrea Schweer [email protected] for the LCoNZ Institutional Research Repositories
33+
* @author Alan Orth for the International Livestock Research Institute
34+
* @version 7.6.1.4
35+
* @since 7.6.1.4
36+
*/
37+
public class RemoveGeneratedThumbnails {
38+
// note: static members belong to the class itself, not any one instance
39+
public static ItemService itemService = ContentServiceFactory.getInstance().getItemService();
40+
public static HandleService handleService =
41+
HandleServiceFactory.getInstance().getHandleService();
42+
public static BundleService bundleService =
43+
ContentServiceFactory.getInstance().getBundleService();
44+
45+
public static void main(String[] args) {
46+
String parentHandle = null;
47+
if (args.length >= 1) {
48+
parentHandle = args[0];
49+
}
50+
51+
Context context = null;
52+
try {
53+
context = new Context();
54+
context.turnOffAuthorisationSystem();
55+
56+
if (StringUtils.isBlank(parentHandle)) {
57+
process(context, itemService.findAll(context));
58+
} else {
59+
DSpaceObject parent = handleService.resolveToObject(context, parentHandle);
60+
if (parent != null) {
61+
switch (parent.getType()) {
62+
case Constants.SITE:
63+
process(context, itemService.findAll(context));
64+
context.commit();
65+
break;
66+
case Constants.COMMUNITY:
67+
List<Collection> collections = ((Community) parent).getCollections();
68+
for (Collection collection : collections) {
69+
process(
70+
context,
71+
itemService.findAllByCollection(context, collection));
72+
}
73+
context.commit();
74+
break;
75+
case Constants.COLLECTION:
76+
process(
77+
context,
78+
itemService.findByCollection(context, (Collection) parent));
79+
context.commit();
80+
break;
81+
case Constants.ITEM:
82+
processItem(context, (Item) parent);
83+
context.commit();
84+
break;
85+
}
86+
}
87+
}
88+
} catch (SQLException | AuthorizeException | IOException e) {
89+
e.printStackTrace(System.err);
90+
} finally {
91+
if (context != null && context.isValid()) {
92+
context.abort();
93+
}
94+
}
95+
}
96+
97+
private static void process(Context context, Iterator<Item> items)
98+
throws SQLException, IOException, AuthorizeException {
99+
while (items.hasNext()) {
100+
Item item = items.next();
101+
processItem(context, item);
102+
itemService.update(context, item);
103+
}
104+
}
105+
106+
private static void processItem(Context context, Item item)
107+
throws SQLException, AuthorizeException, IOException {
108+
List<Bundle> thumbnailBundles = item.getBundles("THUMBNAIL");
109+
for (Bundle thumbnailBundle : thumbnailBundles) {
110+
List<Bitstream> thumbnailBundleBitstreams = thumbnailBundle.getBitstreams();
111+
for (Bitstream thumbnailBitstream : thumbnailBundleBitstreams) {
112+
String thumbnailName = thumbnailBitstream.getName();
113+
String thumbnailDescription = thumbnailBitstream.getDescription();
114+
115+
// There is no point continuing if the thumbnail's description is empty or null
116+
if (StringUtils.isEmpty(thumbnailDescription)) {
117+
continue;
118+
}
119+
120+
if (thumbnailName.toLowerCase().endsWith(".pdf.jpg")) {
121+
List<Bundle> originalBundles = item.getBundles("ORIGINAL");
122+
for (Bundle originalBundle : originalBundles) {
123+
List<Bitstream> originalBundleBitstreams = originalBundle.getBitstreams();
124+
125+
for (Bitstream originalBitstream : originalBundleBitstreams) {
126+
String originalName = originalBitstream.getName();
127+
128+
/*
129+
- check if the original file name is the same as the thumbnail name minus the extra ".jpg"
130+
- check if the thumbnail description indicates it was automatically generated
131+
*/
132+
if (originalName.equalsIgnoreCase(
133+
StringUtils.removeEndIgnoreCase(thumbnailName, ".jpg"))
134+
&& ("Generated Thumbnail".equals(thumbnailDescription)
135+
|| "IM Thumbnail".equals(thumbnailDescription))) {
136+
System.out.println(
137+
item.getHandle()
138+
+ ": removing "
139+
+ thumbnailName);
140+
141+
thumbnailBundle.removeBitstream(thumbnailBitstream);
142+
}
143+
}
144+
}
145+
}
146+
}
147+
}
148+
}
149+
}

0 commit comments

Comments
 (0)