-
Notifications
You must be signed in to change notification settings - Fork 3
Add explicit IIIF v2 prefixed types to resource set #440
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
cubap
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not this. If we remove a lot of the prefixes we cannot tell what sort of data we have because of collisions between ontologies. Only add and equivocate when it is the same type of thing. You can look in http://iiif.io/api/presentation/2/context.json which holds most of hte new things we would need to add.
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
You're absolutely right. I've reverted the prefix-stripping approach and instead explicitly added the IIIF v2 prefixed types to the Changes in commit 261e85a:
Warning Firewall rules blocked me from connecting to one or more addresses (expand for details)I tried to connect to the following addresses, but was blocked by firewall rules:
If you need me to access, download, or install something from one of these locations, you can either:
|
* Hydrate from seed when fetch fails
Introduce a seed-based hydration path and helper (hydrateFromObject) so an already-provided item object can be used if the network fetch fails. Reorder logic to define hydration first, use the seed fallback on non-OK responses and exceptions, and return the hydrated object. Also adjust calls to this.get and this.set to pass the embedded object and the correct type. Local caching to localStorage is preserved.
* runVaulttest
* vault all fetches
* Prefetch manifests before resolving canvases
Add Vault.prefetchDocuments to eagerly cache manifest documents and use it when a canvas or annotation page cannot be resolved from the vault. Components updated (continue-working, transcription-block, transcription interface) now attempt to hydrate all project manifests and retry vault lookups before falling back to placeholders. getImage was converted to async and now uses vault.get (with prefetch) instead of direct fetch, with explicit fallbacks to 404/no-image assets. These changes reduce failed network fetches and improve resolution of IIIF v2/v3 canvases and pages.
* Use vault and prefetch manifests; enhance Vault
Replace ad-hoc fetches with the centralized vault API across multiple components and add a manifest prefetch fallback when an item is not immediately resolvable. Components updated: annotorious-annotator/line-parser, column-selector, continue-working, default-transcribe, legacy-annotator/plain, line-image, simple-transcription — they now import and use the shared vault, attempt vault.prefetchDocuments(TPEN.activeProject.manifest) when a resource is missing, and retry resolving. Default-transcribe and others now use vault.get for annotation/page/contentresource resolution and show clearer failure messages.
Major Vault improvements (js/vault.js):
- Normalizes IDs (removes fragment, expands 24-char Mongo hex IDs to TPEN.RERUMURL/id/...).
- Deduplicates concurrent fetches with inFlightPromises to avoid duplicate network requests.
- Improved hydration: walks embedded objects, queues and caches known IIIF resource types, skips common non-resource properties, normalizes and caches discovered URLs/IDs, and stores data keyed by detected resource type when available.
- More resilient fetch handling (returns null on failure if a seed exists) and safer localStorage writes keyed to resource type.
- PrefetchDocuments now calls vault.get with a best-effort type and collects errors.
- Removed the debug runVaultTest export.
These changes make resource resolution more consistent and robust, reduce duplicate network loads, and allow components to recover by hydrating manifests when resources are initially missing.
* Process IIIF resources and simplify ID normalization
Treat canvas/manifest/collection IDs as external/embedded (return empty URL) and simplify _normalizeId to only strip fragment identifiers. Add a _processIIIFResource helper to detect IIIF resources, call get() for them, and return a minimal {id,type,label} representation while avoiding duplicate fetches via the visited set. Update _fetchAndHydrate to invoke this helper for array items and non-array object values. Remove previous behavior that auto-prefixed Mongo hex IDs with TPEN.RERUMURL and the older string-URL detection/fetch logic to better handle embedded IIIF manifests and avoid incorrect fetching.
* Update interfaces/transcription/index.js
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Update components/read-only-transcribe/index.js
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Guard against empty URI in vault fetch to prevent unnecessary requests (#437)
* Initial plan
* Add guard to prevent fetch with empty URI
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Add prefetchManifests and prefetchCollections aliases with explicit type parameters (#438)
* Initial plan
* Add prefetchManifests and prefetchCollections aliases and update all usages
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Refactor prefetch methods to use docType parameter instead of duplicating code
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Eliminate N+1 fetches for embedded IIIF resources in vault (#439)
* Initial plan
* Fix N+1 fetch issue by caching embedded IIIF resources directly
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Improve comments explaining embedded object detection logic
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Improve embedded object detection with specific content properties
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Allow annotations with body OR target to be cached as embedded
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Add explicit IIIF v2 prefixed types to resource set (#440)
* Initial plan
* Fix IIIF v2 prefix handling in type normalization
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Move prefix pattern to class constant and document resource types
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Revert prefix stripping; explicitly add IIIF v2 prefixed types
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Clarify documentation for agent and annotationlist types
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Update vault.js
* deprecate these
* strip out TPEN.js and vault.js
This component should not include these dependencies.
* Import vault in read-only-transcribe component
Add an import for the vault module (../../js/vault.js) in components/read-only-transcribe/index.js so the component can access vault-related functionality.
* This will work here (#443)
* Early guard to avoid NPEs when this.#transcriptions cannot be used (#444)
* Address Issues 13-17 from static review: Fix vault consistency, error handling, and performance issues (#442)
* Initial plan
* Address Issues 13-17: Fix vault consistency, error handling, and performance issues
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
* Fix vault.js issues 8-11: Hoist constants, fix noCache, clone data, add getWithFallback helper (#445)
Issues Fixed:
- Issue 8: Move SKIP_PROPERTIES and IIIF_RESOURCE_TYPES to module scope to prevent recreation on every fetch (28 + 13 = 41 object allocations eliminated per fetch)
- Issue 9: Extract vault.getWithFallback() helper method to consolidate repeated 5-line fallback pattern used in 10+ locations
- Issue 10: Fix noCache parameter to bypass in-memory store, not just localStorage. When noCache=true, caller gets fresh fetch, not stale cached copy
- Issue 11: Clone data with structuredClone() before BFS mutation to prevent corrupting caller's original object via shared reference
Performance Impact:
- Eliminates unnecessary Set allocations (Issue 8)
- Reduces code duplication and maintenance burden (Issue 9)
- Fixes cache bypass behavior (Issue 10)
- Prevents data corruption bugs (Issue 11)
* Update Conditionals to avoid NPEs (#447)
* update conditions to fix pre existing errors
* update conditions to fix pre existing errors
* Use vault.getWithFallback for fetching (#449)
Replace repeated vault.get + prefetchManifests patterns with a single vault.getWithFallback utility to simplify resource resolution and avoid duplicated manifest-hydration logic. Affected components: annotorious-annotator/line-parser.js, column-selector/index.js, continue-working/index.js, default-transcribe/index.js, legacy-annotator/plain.js, simple-transcription/index.js, transcription-block/index.js, interfaces/transcription/index.js. Preserves existing behavior (including the optional strict flag) while centralizing the fallback to project manifests.
* refactor manage columns to use vault (#450)
* refactor manage columns to use vault
* refactor manage columns to use vault
* hotfix for error messaging
* No 'Auto Parse' for now
* No 'Auto Parse' for now
* Fix so CONTRIBUTOR users see the link into the /annotator interface
* See transcription text in .transcription-input on the transcription interface.
* Don't change @deprecated components
* Changes while reviewing
* Changes during review
* Changes during review
* zooming on view transcription interface
* Don't let view transcription interface be larger than the viewport and cause scrolling
* double right-click to zoom out
* double right-click to zoom out
* small change for accessibility
* small change for accessibility. Add some consistency around how these buttons show or not.
* small change for accessibility. Add some consistency around how these buttons show or not.
* whoops make it prettier
* changes while reviewing
* remove references to elements that are no longer used
* Extra defensive for an extra chance to find the image
* some cleanup
* Changes while reviewing
* Changes while reviewing
* Changes while reviewing
* Changes while reviewing
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <198982749+Copilot@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
Co-authored-by: Bryan Haberberger <bryan.j.haberberger@slu.edu>
* Hydrate from seed when fetch fails
Introduce a seed-based hydration path and helper (hydrateFromObject) so an already-provided item object can be used if the network fetch fails. Reorder logic to define hydration first, use the seed fallback on non-OK responses and exceptions, and return the hydrated object. Also adjust calls to this.get and this.set to pass the embedded object and the correct type. Local caching to localStorage is preserved.
* runVaulttest
* vault all fetches
* Prefetch manifests before resolving canvases
Add Vault.prefetchDocuments to eagerly cache manifest documents and use it when a canvas or annotation page cannot be resolved from the vault. Components updated (continue-working, transcription-block, transcription interface) now attempt to hydrate all project manifests and retry vault lookups before falling back to placeholders. getImage was converted to async and now uses vault.get (with prefetch) instead of direct fetch, with explicit fallbacks to 404/no-image assets. These changes reduce failed network fetches and improve resolution of IIIF v2/v3 canvases and pages.
* Use vault and prefetch manifests; enhance Vault
Replace ad-hoc fetches with the centralized vault API across multiple components and add a manifest prefetch fallback when an item is not immediately resolvable. Components updated: annotorious-annotator/line-parser, column-selector, continue-working, default-transcribe, legacy-annotator/plain, line-image, simple-transcription — they now import and use the shared vault, attempt vault.prefetchDocuments(TPEN.activeProject.manifest) when a resource is missing, and retry resolving. Default-transcribe and others now use vault.get for annotation/page/contentresource resolution and show clearer failure messages.
Major Vault improvements (js/vault.js):
- Normalizes IDs (removes fragment, expands 24-char Mongo hex IDs to TPEN.RERUMURL/id/...).
- Deduplicates concurrent fetches with inFlightPromises to avoid duplicate network requests.
- Improved hydration: walks embedded objects, queues and caches known IIIF resource types, skips common non-resource properties, normalizes and caches discovered URLs/IDs, and stores data keyed by detected resource type when available.
- More resilient fetch handling (returns null on failure if a seed exists) and safer localStorage writes keyed to resource type.
- PrefetchDocuments now calls vault.get with a best-effort type and collects errors.
- Removed the debug runVaultTest export.
These changes make resource resolution more consistent and robust, reduce duplicate network loads, and allow components to recover by hydrating manifests when resources are initially missing.
* Process IIIF resources and simplify ID normalization
Treat canvas/manifest/collection IDs as external/embedded (return empty URL) and simplify _normalizeId to only strip fragment identifiers. Add a _processIIIFResource helper to detect IIIF resources, call get() for them, and return a minimal {id,type,label} representation while avoiding duplicate fetches via the visited set. Update _fetchAndHydrate to invoke this helper for array items and non-array object values. Remove previous behavior that auto-prefixed Mongo hex IDs with TPEN.RERUMURL and the older string-URL detection/fetch logic to better handle embedded IIIF manifests and avoid incorrect fetching.
* Update interfaces/transcription/index.js
* Update components/read-only-transcribe/index.js
* Guard against empty URI in vault fetch to prevent unnecessary requests (#437)
* Initial plan
* Add guard to prevent fetch with empty URI
---------
* Add prefetchManifests and prefetchCollections aliases with explicit type parameters (#438)
* Initial plan
* Add prefetchManifests and prefetchCollections aliases and update all usages
* Refactor prefetch methods to use docType parameter instead of duplicating code
---------
* Eliminate N+1 fetches for embedded IIIF resources in vault (#439)
* Initial plan
* Fix N+1 fetch issue by caching embedded IIIF resources directly
* Improve comments explaining embedded object detection logic
* Improve embedded object detection with specific content properties
* Allow annotations with body OR target to be cached as embedded
---------
* Add explicit IIIF v2 prefixed types to resource set (#440)
* Initial plan
* Fix IIIF v2 prefix handling in type normalization
* Move prefix pattern to class constant and document resource types
* Revert prefix stripping; explicitly add IIIF v2 prefixed types
* Clarify documentation for agent and annotationlist types
---------
* Update vault.js
* deprecate these
* strip out TPEN.js and vault.js
This component should not include these dependencies.
* Import vault in read-only-transcribe component
Add an import for the vault module (../../js/vault.js) in components/read-only-transcribe/index.js so the component can access vault-related functionality.
* This will work here (#443)
* Early guard to avoid NPEs when this.#transcriptions cannot be used (#444)
* Address Issues 13-17 from static review: Fix vault consistency, error handling, and performance issues (#442)
* Initial plan
* Address Issues 13-17: Fix vault consistency, error handling, and performance issues
---------
* Fix vault.js issues 8-11: Hoist constants, fix noCache, clone data, add getWithFallback helper (#445)
Issues Fixed:
- Issue 8: Move SKIP_PROPERTIES and IIIF_RESOURCE_TYPES to module scope to prevent recreation on every fetch (28 + 13 = 41 object allocations eliminated per fetch)
- Issue 9: Extract vault.getWithFallback() helper method to consolidate repeated 5-line fallback pattern used in 10+ locations
- Issue 10: Fix noCache parameter to bypass in-memory store, not just localStorage. When noCache=true, caller gets fresh fetch, not stale cached copy
- Issue 11: Clone data with structuredClone() before BFS mutation to prevent corrupting caller's original object via shared reference
Performance Impact:
- Eliminates unnecessary Set allocations (Issue 8)
- Reduces code duplication and maintenance burden (Issue 9)
- Fixes cache bypass behavior (Issue 10)
- Prevents data corruption bugs (Issue 11)
* Update Conditionals to avoid NPEs (#447)
* update conditions to fix pre existing errors
* update conditions to fix pre existing errors
* Use vault.getWithFallback for fetching (#449)
Replace repeated vault.get + prefetchManifests patterns with a single vault.getWithFallback utility to simplify resource resolution and avoid duplicated manifest-hydration logic. Affected components: annotorious-annotator/line-parser.js, column-selector/index.js, continue-working/index.js, default-transcribe/index.js, legacy-annotator/plain.js, simple-transcription/index.js, transcription-block/index.js, interfaces/transcription/index.js. Preserves existing behavior (including the optional strict flag) while centralizing the fallback to project manifests.
* refactor manage columns to use vault (#450)
* refactor manage columns to use vault
* refactor manage columns to use vault
* hotfix for error messaging
* No 'Auto Parse' for now
* No 'Auto Parse' for now
* Fix so CONTRIBUTOR users see the link into the /annotator interface
* See transcription text in .transcription-input on the transcription interface.
* Don't change @deprecated components
* Changes while reviewing
* Changes during review
* Changes during review
* zooming on view transcription interface
* Don't let view transcription interface be larger than the viewport and cause scrolling
* double right-click to zoom out
* double right-click to zoom out
* small change for accessibility
* small change for accessibility. Add some consistency around how these buttons show or not.
* small change for accessibility. Add some consistency around how these buttons show or not.
* whoops make it prettier
* changes while reviewing
* remove references to elements that are no longer used
* Extra defensive for an extra chance to find the image
* some cleanup
* Changes while reviewing
* Changes while reviewing
* Changes while reviewing
* Changes while reviewing
---------
Co-authored-by: Patrick Cuba <cubap@slu.edu>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <198982749+Copilot@users.noreply.github.com>
Co-authored-by: cubap <1119165+cubap@users.noreply.github.com>
IIIF Presentation v2 resources use prefixed types (
sc:Canvas,oa:Annotation) while v3 uses unprefixed forms (Canvas,Annotation). Without explicit support for v2 prefixed types, these resources were not being recognized and hydrated correctly.Changes:
iiifResourceTypessetsc:collection,sc:canvas,sc:sequence,sc:range,sc:layersc:manifest,oa:annotation,oa:annotationlistdcterms:Agentvsfoaf:Agent)Approach:
The
_normalizeTypemethod lowercases types while preserving prefixes. Both v2 and v3 types are explicitly listed in theiiifResourceTypesset:This approach avoids collisions between different ontologies while ensuring both IIIF v2 and v3 resources are properly recognized and hydrated.
✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.