Merge branch 'develop' into 'fb-BROS-5/iva-fixes'

Workflow run: https://github.com/HumanSignal/label-studio/actions/runs/15111983801

Author: hlomzik

.github/workflows/apply-linters.yml

@@ -0,0 +1,54 @@
+name: Apply linters
+
+on:
+  workflow_dispatch:
+    inputs:
+      branch_name:
+        description: 'Branch name to run the workflow on'
+        required: true
+        type: string
+
+env:
+  NODE: "18"
+  FRONTEND_MONOREPO_DIR: "web"
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+        with:
+          token: ${{ secrets.GIT_PAT }}
+          ref: ${{ inputs.branch_name }}
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.12'
+
+      - name: Install pre-commit
+        run: |
+          python -m pip install --upgrade pip
+          pip install pre-commit
+
+      - name: Setup frontend environment
+        uses: ./.github/actions/setup-frontend-environment
+        with:
+          node-version: "${{ env.NODE }}"
+          directory: "${{ env.FRONTEND_MONOREPO_DIR }}"
+
+      - name: Run formatters
+        run: make fmt-all || true
+
+      - name: Ensure no lint remains
+        run: make fmt-all
+
+      - name: Commit changes
+        run: |
+          git config --global user.name "robot-ci-heartex"
+          git config --global user.email "robot-ci-heartex@users.noreply.github.com"
+          git add .
+          git commit -m "Apply pre-commit linters" || echo "No changes to commit"
+          git push

.github/workflows/docker-build-ontop.yml

@@ -110,7 +110,7 @@ jobs:
             ${{ steps.calculate-docker-tags.outputs.docker-tags }}
 
       - name: Push Docker image
-        uses: docker/build-push-action@v6.16.0
+        uses: docker/build-push-action@v6.17.0
         id: docker_build_and_push
         with:
           context: .

.github/workflows/docker-build.yml

@@ -130,7 +130,7 @@ jobs:
             type=raw,value=${{ steps.version.outputs.build_version }}
 
       - name: Push Docker image
-        uses: docker/build-push-action@v6.16.0
+        uses: docker/build-push-action@v6.17.0
         id: docker_build_and_push
         with:
           context: .

.github/workflows/docker-release-promote.yml

@@ -195,7 +195,7 @@ jobs:
             ${{ steps.generate-tags.outputs.ubuntu-tags }}
 
       - name: Build and Push Release Ubuntu Docker image
-        uses: docker/build-push-action@v6.16.0
+        uses: docker/build-push-action@v6.17.0
         id: docker_build
         with:
           context: ${{ steps.release_dockerfile.outputs.release_dir }}

.github/workflows/release-cut-off-release-branch.yml

@@ -14,7 +14,7 @@ on:
         type: string
 
 env:
-  RELEASE_BRANCH_PREFIX: "ls-release"
+  RELEASE_BRANCH_PREFIX: "ls-release/"
 
 jobs:
   draft-new-release:
@@ -45,22 +45,24 @@ jobs:
       - name: Calculate branch name and version
         id: calculate_branch_name_and_version
         shell: bash
+        env:
+          VERSION: "${{ inputs.version }}"
         run: |
           set -xeuo pipefail
           
           regexp='^[v]?([0-9]+)\.([0-9]+)\.0$';
           
-          if [[ "${{ inputs.version }}" =~ $regexp ]]; then
+          if [[ "${VERSION}" =~ $regexp ]]; then
               first="${BASH_REMATCH[1]}"
               second="${BASH_REMATCH[2]}"
-              third="${BASH_REMATCH[3]}"
+              third="0"
           else
-              echo "${{ inputs.version }} does not mach the regexp ${regexp}"
+              echo "::error::${VERSION} does not mach the regexp ${regexp}"
               exit 1
           fi
           
           release_version="${first}.${second}.${third}"
-          release_branch="${{ env.RELEASE_BRANCH_PREFIX }}/${first}.${second}.${third}"
+          release_branch="${{ env.RELEASE_BRANCH_PREFIX }}${first}.${second}.${third}"
           next_develop_version="${first}.$(($second + 1)).0.dev0"
           
           echo "release_branch=${release_branch}" >> "${GITHUB_OUTPUT}"

.github/workflows/tests.yml

@@ -140,7 +140,7 @@ jobs:
 
       - name: Upload coverage to Codecov
         if: ${{ github.event.pull_request.head.repo.fork == false && github.event.pull_request.user.login != 'dependabot[bot]' }}
-        uses: codecov/codecov-action@v5.4.2
+        uses: codecov/codecov-action@v5.4.3
         with:
           name: codecov-python-${{ matrix.python-version }}
           flags: pytests

README.md

@@ -23,7 +23,7 @@ Have a custom dataset? You can customize Label Studio to fit your needs. Read an
 
 ## Try out Label Studio
 
-Install Label Studio locally, or deploy it in a cloud instance. [Or, sign up for a free trial of our Enterprise edition.](https://humansignal.com/free-trial).
+Install Label Studio locally or deploy it in a cloud instance. [Or sign up for a free trial of our Starter Cloud edition!](https://humansignal.com/platform/starter-cloud/) You can learn more about what each edition offers [here](https://labelstud.io/guide/label_studio_compare). 
 
 - [Install locally with Docker](#install-locally-with-docker)
 - [Run with Docker Compose (Label Studio + Nginx + PostgreSQL)](#run-with-docker-compose)

docs/source/guide/install_enterprise_docker.md

@@ -19,6 +19,9 @@ See [Secure Label Studio](security.html) for more details about security and har
 
 To install Label Studio Community Edition, see [Install Label Studio](https://labelstud.io/guide/install). This page is specific to the Enterprise version of Label Studio.
 
+!!! note
+    On-prem deployments of Label Studio Enterprise are not supported for Academic licenses.  
+
 {% insertmd includes/deploy.md %}
 
 ## Install Label Studio Enterprise using Docker

docs/source/guide/install_enterprise_k8s.md

@@ -23,6 +23,9 @@ Your Kubernetes cluster can be self-hosted or installed somewhere such as Amazon
 
 </div>
 
+!!! note
+    On-prem deployments of Label Studio Enterprise are not supported for Academic licenses. 
+
 This high-level architecture diagram that outlines the main components of a Label Studio Enterprise deployment.
 
 <img src="/images/LSE_k8s_scheme.png"/>

docs/source/tags/pdf.md

@@ -0,0 +1,49 @@
+---
+title: PDF
+type: tags
+order: 302
+meta_title: PDF Tag for loading PDF documents
+meta_description: Label Studio PDF Tag for loading PDF documents for machine learning and data science projects.
+---
+
+The `Pdf` tag displays a PDF document for labeling. Use for performing document-level annotations, transcription, and summarization.
+
+Use with the following data types: PDF.
+
+### Parameters
+
+| Param | Type | Default | Description |
+| --- | --- | --- | --- |
+| name | <code>string</code> |  | Name of the element |
+| value | <code>string</code> |  | Value of the element - field name to retrieve the PDF URL from |
+
+### Supported Control tags
+Document-level annotations are supported with Pdf tag, for example:
+
+- Document classification with [Choices](/tags/choices.html)
+- Document rating with [Rating](/tags/rating.html)
+- Transcription and summarization with [TextArea](/tags/textarea.html)
+
+### Example
+
+Labeling configuration to label PDF documents:
+
+```html
+<View>
+  <Pdf name="pdf" value="$pdf" />
+  <Choices name="choices" toName="pdf">
+    <Choice value="Legal" />
+    <Choice value="Financial" />
+    <Choice value="Technical" />
+  </Choices>
+</View>
+```
+
+**Example Input data:**
+
+```json
+{
+  "pdf": "https://app.humansignal.com/static/samples/sample.pdf"
+}
+```
+

docs/source/templates/pdf_classification.md

@@ -26,11 +26,11 @@ If you want to perform PDF classification, use this template. This template prom
         <Choice value="Important article"/>
         <Choice value="Yellow press"/>
     </Choices>
-    <HyperText name="pdf" value="$pdf" inline="true"/>
+    <Pdf name="pdf" value="$pdf"/>
 </View>
 
 <!-- {
-    "pdf": "<embed src='https://app.heartex.ai/static/samples/sample.pdf' width='100%' height='600px'/>"
+    "pdf": "/static/samples/sample.pdf"
 } -->
 ```
 
@@ -56,9 +56,9 @@ Use the [Choices](/tags/choices.html) control tag to present classification opti
   </Choices>
 ```
 
-Use the [HyperText](/tags/hypertext.html) tag to render an inline version of the PDF data:
+Use the [Pdf](/tags/pdf.html) tag to render an inline version of the PDF data:
 ```xml
-<HyperText name="pdf" value="$pdf" inline="true"/>
+<Pdf name="pdf" value="$pdf"/>
 ```
 
 ### Input data
@@ -74,4 +74,4 @@ Label Studio does not support labeling PDF-formatted files directly. You should
 ## Related tags
 - [Rating](/tags/rating.html)
 - [Choices](/tags/choices.html)
-- [HyperText](/tags/hypertext.html)
+- [Pdf](/tags/pdf.html)

label_studio/annotation_templates/structured-data-parsing/pdf-classification/config.xml

@@ -6,7 +6,7 @@
     <Choice value="Important article"/>
     <Choice value="Yellow press"/>
   </Choices>
-  <HyperText name="pdf" value="$pdf" inline="true"/>
+  <Pdf name="pdf" value="$pdf"/>
 </View>
 
 

label_studio/annotation_templates/structured-data-parsing/pdf-classification/config.yml

@@ -12,13 +12,13 @@ config: |
       <Choice value="Important article"/>
       <Choice value="Yellow press"/>
     </Choices>
-    <HyperText name="pdf" value="$pdf" inline="true"/>
+    <Pdf name="pdf" value="$pdf"/>
   </View>
 
 
   <!-- {
     "data": {
-      "pdf": "<embed src='/static/samples/sample.pdf' width='100%' height='600px'/>"
+      "pdf": "/static/samples/sample.pdf"
     }
   } -->
 

label_studio/core/settings/base.py

@@ -494,6 +494,7 @@
         '.mp4',
         '.webm',
         '.webp',
+        '.pdf',
     ]
 )
 
@@ -607,6 +608,7 @@
 FEATURE_FLAGS_GET_USER_REPR = 'core.feature_flags.utils.get_user_repr'
 
 # Test factories
+ORGANIZATION_FACTORY = 'organizations.tests.factories.OrganizationFactory'
 PROJECT_FACTORY = 'projects.tests.factories.ProjectFactory'
 USER_FACTORY = 'users.tests.factories.UserFactory'
 

label_studio/data_manager/actions/cache_labels.py

@@ -5,6 +5,7 @@
 
 from core.permissions import AllPermissions
 from core.redis import start_job_async_or_sync
+from label_studio_sdk.label_interface import LabelInterface
 from tasks.models import Annotation, Prediction, Task
 
 logger = logging.getLogger(__name__)
@@ -18,6 +19,8 @@ def cache_labels_job(project, queryset, **kwargs):
     source_class = Annotation if source == 'annotations' else Prediction
     control_tag = request_data.get('custom_control_tag') or request_data.get('control_tag')
     with_counters = request_data.get('with_counters', 'Yes').lower() == 'yes'
+    label_interface = LabelInterface(project.label_config)
+    label_interface_tags = {tag.name: tag for tag in label_interface.find_tags('control')}
 
     if source == 'annotations':
         column_name = 'cache'
@@ -38,7 +41,7 @@ def cache_labels_job(project, queryset, **kwargs):
         task_labels = []
         annotations = source_class.objects.filter(task=task).only('result')
         for annotation in annotations:
-            labels = extract_labels(annotation, control_tag)
+            labels = extract_labels(annotation, control_tag, label_interface_tags)
             task_labels.extend(labels)
 
         # cache labels in separate data column
@@ -57,20 +60,36 @@ def cache_labels_job(project, queryset, **kwargs):
     return {'response_code': 200, 'detail': f'Updated {len(tasks)} tasks'}
 
 
-def extract_labels(annotation, control_tag):
+def extract_labels(annotation, control_tag, label_interface_tags=None):
     labels = []
     for region in annotation.result:
         # find regions with specific control tag name or just all regions if control tag is None
         if (control_tag is None or region['from_name'] == control_tag) and 'value' in region:
-            # scan value for a field with list of strings,
-            # as bonus it will work with textareas too
+            # scan value for a field with list of strings (eg choices, textareas)
+            # or taxonomy (list of string-lists)
             for key in region['value']:
-                if (
-                    isinstance(region['value'][key], list)
-                    and region['value'][key]
-                    and isinstance(region['value'][key][0], str)
-                ):
-                    labels.extend(region['value'][key])
+                if region['value'][key] and isinstance(region['value'][key], list):
+
+                    if key == 'taxonomy':
+                        showFullPath = 'true'
+                        pathSeparator = '/'
+                        if label_interface_tags is not None and region['from_name'] in label_interface_tags:
+                            # if from_name is not a custom_control tag, then we can try to fetch taxonomy formatting params
+                            label_interface_tag = label_interface_tags[region['from_name']]
+                            showFullPath = label_interface_tag.attr.get('showFullPath', 'false')
+                            pathSeparator = label_interface_tag.attr.get('pathSeparator', '/')
+
+                        if showFullPath == 'false':
+                            for elems in region['value'][key]:
+                                labels.append(elems[-1])  # just the leaf node of a taxonomy selection
+                        else:
+                            for elems in region['value'][key]:
+                                labels.append(pathSeparator.join(elems))  # the full delimited taxonomy path
+
+                    # other control tag types like Choices & TextAreas
+                    elif isinstance(region['value'][key][0], str):
+                        labels.extend(region['value'][key])
+
                     break
     return labels
 

label_studio/data_manager/actions/remove_duplicates.py

@@ -172,6 +172,8 @@ def restore_storage_links_for_duplicated_tasks(duplicates) -> None:
                     link = storage_link_class(
                         task_id=task['id'],
                         key=link_instance.key,
+                        row_index=link_instance.row_index,
+                        row_group=link_instance.row_group,
                         storage=link_instance.storage,
                     )
                     link.save()

label_studio/feature_flags.json

@@ -3120,6 +3120,33 @@
       "version": 2,
       "deleted": false
     },
+    "fflag_feat_root_11_support_jsonl_cloud_storage": {
+      "key": "fflag_feat_root_11_support_jsonl_cloud_storage",
+      "on": false,
+      "prerequisites": [],
+      "targets": [],
+      "contextTargets": [],
+      "rules": [],
+      "fallthrough": {
+        "variation": 0
+      },
+      "offVariation": 1,
+      "variations": [
+        true,
+        false
+      ],
+      "clientSideAvailability": {
+        "usingMobileKey": false,
+        "usingEnvironmentId": false
+      },
+      "clientSide": false,
+      "salt": "85e018dcd2e64c689a61ee7ed3c5edb2",
+      "trackEvents": false,
+      "trackEventsFallthrough": false,
+      "debugEventsUntilDate": null,
+      "version": 2,
+      "deleted": false
+    },
     "fflag_feature_all_optic_1421_cold_start_v2": {
       "key": "fflag_feature_all_optic_1421_cold_start_v2",
       "on": false,