Repository navigation
fix(reports): handle missing reference_id/name in submission reports - #871
Merged
Merged
Conversation
… reports Submissions created via the REST API only require canonical_smiles and references, so new_molecule_data may lack reference_id and name. Opening such a report in the dashboard (edit/view) threw 'Undefined array key "reference_id"'. Fall back to null like the other optional fields.
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## development #871 +/- ##
==================================================
+ Coverage 6.14% 22.74% +16.59%
- Complexity 1736 2361 +625
==================================================
Files 227 261 +34
Lines 8945 11137 +2192
==================================================
+ Hits 550 2533 +1983
- Misses 8395 8604 +209 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
Fixes the Composer security audit CI step (http-foundation, http-kernel, routing, mime, serializer, filesystem, html-sanitizer, yaml).
Merged
vcnainala
added a commit
that referenced
this pull request
Nov 5, 2026
* fix: update CSP to allow images and API calls from naturalproducts.net (#788) * chore(deps): resolve composer security advisories (#791) * chore(deps): resolve composer security advisories Bump locked dependencies to clear composer audit, including Filament 4.11.6, Laravel 12.61, Symfony security patches, PhpSpreadsheet 1.30.5, and related transitive packages. * chore(filament): republish assets after dependency upgrade Refresh published Filament 4 and plugin frontend assets to match the updated composer lockfile. * style: apply Laravel Pint across application code Normalize imports and class references in app layer (HTTP kernel, Filament resources, REST API, jobs, events, models) per Pint rules. * style: normalize imports in config and routes Apply Pint class import conventions to published config files and api routes without changing configuration values. * style: normalize imports in factories and migrations Replace FQCN references with imported class names in database factories and migrations for consistency with Pint output. * fix(models): add sampleLocations relationship on Molecule Define the belongsToMany relationship via molecule_organism so Filament and PHPStan can resolve sample location access on molecules. * fix(phpstan): resolve static analysis errors in publish and Filament code Remove stale phpstan-ignore on PublishMolecules and narrow collection updates with explicit null checks. Type the organism molecules bulk action repeater default using Molecule and the sampleLocations relation. * fix: update CSP to allow images and API calls from naturalproducts.net (#792) * fix(organisms): allow linking existing organisms in report changes Relax unique-name validation in the request-changes organism repeater so users can link organisms that already exist in the database. Reject names already attached to the target molecule instead. Fixes #772 * fix(reports): correct select validation in request changes form Align Filament v4 Select option keys with stored state for existing synonyms, CAS, geo locations, organisms, and citations. Resolve the target molecule from form state and report records so options remain valid on Livewire submit requests. Fixes #775 * fix(reports): resolve PHPStan errors in ReportResource Use explicit null checks for molecule synonyms/CAS options and remove unreachable is_string handling for mol_ids array cast. * feat(api-docs): adopt Scalar for interactive API documentation (#790) * fix(api-docs): preserve HTTP status codes in OpenAPI response generation Lomkit's Responses::withOthers() used array_merge(), which reindexed numeric keys such as 500 to 0 and broke Scalar's OpenAPI parser. Add a custom Responses schema that merges with + instead, sanitize invalid keys when generating docs, and regenerate the published specification. * feat(api-docs): replace Swagger UI with Scalar reference UI Serve the Lomkit OpenAPI spec through Scalar on /api-documentation with COCONUT branding, black body text, and a single header logo. Allow fonts.scalar.com in the CSP for Scalar default web fonts. * docs(api-docs): update authentication instructions for Scalar UI Replace Swagger-specific wording with generic interactive API documentation guidance for Bearer token authentication. * fix(api-docs): resolve CI failures for Scalar API documentation PR Update vulnerable Composer dependencies so security audit passes, and add unit/feature tests covering HTTP status code preservation in OpenAPI docs. * feat(jetstream): enable read-only personal access tokens (#793) * feat(jetstream): enable read-only personal access tokens in UI Turn on Jetstream API features and restrict Sanctum PAT abilities to read. REST authorization remains on Spatie permissions via Lomkit policies. * docs(api): document personal access tokens Describe how verified users create PATs, read-only Sanctum abilities, and Spatie-based authorization for mutations and deletes. * feat(api): expose organism associations on molecules REST resource (#794) Add organisms as an includable BelongsToMany relation on MoleculeResource and expose organism_count so API clients can load linked taxa without a separate organisms lookup. Update API docs and regenerate OpenAPI spec. * refactor: externalize hardcoded URLs to environment configuration (#795) * refactor(config): centralize external service URLs in env and config Move S3 public and download bases to filesystems disk config with AWS_* defaults. Extend services.php for cheminf, citation, regapp OIDC, pubchem, npclassifier, taxonomy APIs, avatars, and coconut public URL. Document keys in .env.example and .env.ci. Set APP_URL default to http://localhost for local dev. * feat(helper): add public storage and download URL helpers Expose filesystems.disks.s3.url and downloads_url via public_storage_url() and public_downloads_url() for consistent object-storage links in Blade and PHP. * refactor(app): resolve integration and asset URLs from configuration Use services.* config for PubChem, NPClassifier, NFDI OIDC, OLS, GlobalNames, and ui-avatars. Use url() for collection DOI landing pages. Use public_storage_url in Filament collection image state. * refactor(csp): derive allowed origins from application configuration Build CSP allowlists from app.url, services.coconut.public_url, S3 endpoints, cheminf, and citation services instead of hardcoded production hostnames. * refactor(views): replace hardcoded URLs in Blade templates Use url(), public_storage_url(), public_downloads_url(), and config-driven cheminf depict base in structure editor, download pages, and navigation. * fix(csp): skip null config values when building connect allowlist CI caches config from .env.ci where NFDI_REDIRECT_URL is unset, causing Policy::add() to receive null and return 500 on email verification routes. * fix(collections): hide draft collections from public browse surfaces (#796) * feat(collections): add published query scope and factory state Introduce a reusable published() scope on Collection and a matching factory state so public queries and tests can target published records consistently. * fix(collections): hide draft collections from public browse surfaces Apply the published scope on the collections page, homepage featured collections, identifier redirects, and REST search so DRAFT collections are not exposed outside the dashboard. * test(collections): assert browse page lists published collections only Add a feature test that verifies draft collections are excluded from /collections and from the CollectionList Livewire component. * fix(collections): satisfy PHPStan in CollectionResource searchQuery Use where() instead of the published() scope because Lomkit REST builders are typed as the Eloquent contract interface, which PHPStan does not resolve to model-specific query scopes. * feat(collections): sorting, empty state, and cached annotation scores on browse page (#797) * refactor(collections): add published scope and factory state Introduce scopePublished on Collection and a published() factory state so public browse logic and tests can reuse a single visibility rule. * feat(collections): cache average annotation scores for browse sorting Add CollectionAnnotationScores helper with flexible cache and invalidate it when coconut:cache refreshes collection widget counts. * feat(collections): add sorting, direction toggle, and empty state to browse Extend the public collections page with multi-field sort (asc/desc toggle), default release-date ordering, search empty state with clear action, and feature tests for published filtering and sort behaviour. * feat(np-classifier): store pathway/superclass/class as jsonb arrays (#802) * feat(np-classifier): migrate pathway/superclass/class columns to jsonb arrays Introduce NpClassifierResults for normalizing API and import payloads, add a PostgreSQL migration to convert existing text values, and cast model attributes as arrays so all classifier labels are retained per molecule. * feat(np-classifier): persist full label lists and add --force reclassification Update classification jobs and artisan commands to store complete API result arrays via NpClassifierResults, and add a --force option to refresh molecules that already have classifier data after deploy. * feat(search): match NP classifier filters against jsonb array elements Expand np_pathway, np_superclass, and np_class filter queries to search any stored label in the jsonb array instead of a single scalar value. * feat(ui): display all NP classifier labels on molecule details Render pathway, superclass, and class as comma-separated search links so molecules with multiple classifier assignments show every stored label. * chore(reports): aggregate jsonb NP classifier values in chart generators Update bubble frequency, stacked bar, and properties JSON commands to unnest jsonb classifier arrays when building unique value lists and frequency counts. * docs(api): document NP classifier fields as JSON string arrays Clarify REST and search documentation that pathway, superclass, and class values may contain multiple labels per molecule after the jsonb migration. * test(np-classifier): add unit and feature coverage for array storage and search Cover NpClassifierResults normalization/import parsing and verify filter search matches any pathway stored in a molecule jsonb array. * fix(search): SMILES auto-detection and citations search button (#803) * fix(search): correct citations tagType on welcome page button Align the citations tab search button with Enter key handling and the backend tag handler by using tagType=citations and activeTab=citations. * fix(search): detect SMILES ring notation instead of molecular formula Ring-closure SMILES such as C1CCCCC1 were misclassified as molecular formulas because the element-token regex matched before the SMILES check. Reject consecutive same-element tokens so these queries use RDKit search. * Organism resource improvements (#701) * feat(organisms): add preset views for inactive and unmapped organisms - Rename 'inactive entries' to 'inactive' for organisms with molecule_count <= 0 - Add 'unmapped' preset view for organisms without a rank assigned - Fix query grouping for proper OR condition handling * fix(organisms): correct isObsolete boolean check and family rank mapping - Fix isObsolete comparison from string 'false' to boolean false - Fix family rank incorrectly labeled as 'genus' in updateOrganismModel - Auto-generate slug if not exists when updating organism * feat(organisms): enhance organisms table with molecule count and IRI - Add molecule_count column to organisms table (sortable) - Show IRI as description below organism name - Fix isObsolete boolean comparison in OLS lookup - Clean up unused code * feat(organisms): add multi-source taxonomy search with modal selection - Add taxonomy search modal with results from multiple sources - Search NCBI Taxonomy (OLS) with exact and similar matches - Search Global Names Finder (GNF) including unverified names - Normalize organism name case before searching - Add Google and Google Scholar hint action links - Allow user to select from search results and apply to form - Fix isObsolete boolean comparison * feat(organisms): redesign similar organisms UI - Update card layout with table structure for reliable rendering - Show organism name in bold italic (scientific name style) - Display molecule count and IRI link - Add always-visible Edit button with inline styles - Fix array key error for single-word organism names - Fix urldecode on column name string - Improve empty state with search icon * style(organisms): set full width for create/edit organism pages - Add getMaxContentWidth() returning 'full' for better form layout * feat(organisms): click table rows to open view instead of edit - Enable ViewOrganism page route - Add ViewAction button to table row actions - Set recordUrl to navigate to view page on row click * feat(organisms): improve similar organisms query - Handle concatenated names (e.g., 'Micheliachampaca' -> 'Michelia') - Handle names without CamelCase using first 8 chars as genus - Add similarity ranking for better result ordering - Sort by similarity rank then by molecule count * feat(organisms): add merge functionality for similar organisms - Add merge button to similar organisms table - Implement Filament modal for merge confirmation - Show FROM/INTO organisms and transfer counts in modal - Transfer molecules and sample locations to target organism - Handle duplicate relations gracefully - Update molecule counts after merge - Redirect to target organism after successful merge * fix(organisms): satisfy PHPStan and Pint after development merge Cast unmapped preset badge count to string and apply Pint fixes in Organism model. * fix(reports): request changes form validation for compounds and organisms (#804) * fix(organisms): allow linking existing organisms in report changes Relax unique-name validation in the request-changes organism repeater so users can link organisms that already exist in the database. Reject names already attached to the target molecule instead. Fixes #772 * fix(reports): correct select validation in request changes form Align Filament v4 Select option keys with stored state for existing synonyms, CAS, geo locations, organisms, and citations. Resolve the target molecule from form state and report records so options remain valid on Livewire submit requests. Fixes #775 * fix(reports): resolve PHPStan errors in ReportResource Use explicit null checks for molecule synonyms/CAS options and remove unreachable is_string handling for mol_ids array cast. * fix(import): prevent placeholder parents from missing in text search (#805) * fix(import): clear is_placeholder for parents without stereo variants When an entry has stereocenters but no defined stereochemistry, the parent molecule becomes the canonical record but was left flagged as a placeholder. That excluded real compounds from text search (fixes #506). Align ImportEntry and ImportEntryAuto with ImportEntriesAuto, and extend coconut:update-parent-variants-counts to clear is_placeholder on parents that have no variants. * test(search): add regression coverage for placeholder parent visibility Add unit tests for SearchMolecule text search and UpdateParentVariantsCounts command. Configure Sail-compatible test database via .env.testing and remove hardcoded DB_HOST from phpunit.xml. * chore: link issue #506 Fixes #506 * fix(ci): resolve database setup failure with cached Sail test config Clear config before migrate so CI env vars take effect, and stop .env.testing from overriding DB_HOST so GitHub Actions can reach Postgres on localhost. * fix(ci): add APP_KEY to .env.testing for PHPUnit runs Laravel loads .env.testing when APP_ENV=testing, so the key generated in .env during CI setup was ignored and caused MissingAppKeyException. * fix(api): enable ilike filter for REST API search (#806) * fix(api): enable ilike filter for REST API search Extend Lomkit search validation to accept the PostgreSQL ilike operator, enabling case-insensitive filtering on REST resource endpoints. Closes #335 * style: fix Pint violations in REST API search rules Remove empty constructor parentheses and align unary operator spacing so PHP Lint & Security passes on PR #806. * feat(reports): add category-aware curator approval confirmation messages (#807) Curators now see explicit modal headings, descriptions, and submit labels when approving reports, distinguishing first-review forwarding from final approval outcomes for submissions, updates, and revocations. Closes #770 * feat(reports): sorting, email link fixes, and auth UI cleanup (#808) * feat(reports): add sortable number and date columns to list table Expose report id and created_at as visible, sortable Filament table columns so curators and reporters can order reports by number or submission date. Fixes #769 * fix(auth): remove duplicate logo above authentication forms Drop the authentication-card logo slot so the site header logo is the only branding on login and related auth pages across all breakpoints. * fix(reports): normalize compound and DOI links in report emails Use the compound route helper for molecule URLs and normalize DOI values before building markdown links to avoid malformed hrefs in notification mail. Fixes #737 * ci: sync PR pending label on linked issues Add a workflow that adds or removes the PR pending label when open pull requests reference issues using closing keywords. * fix(reports): allow assigned curators to edit all report fields (#809) Assigned curators were blocked from editing reports by missing Shield permissions and UPDATE-field gating tied to reporter requests. Centralize curator edit authorization on the Report model, grant curator report permissions in the seeder, and unlock all change fields on edit. Closes #738 * feat(blog): integrate Filament blog with public pages (#811) * fix(shield): migrate config to Filament Shield 4.2 format Update the published shield config to the structure required by Shield 4.2 so role editing works again, and exclude dashboard blog resources that are authorized separately from role permissions. * feat(blog): install filament-blog package and database schema Add stephenjude/filament-blog with RichEditor configuration, blog tables migration, and a seeder for initial author, category, and post. * feat(blog): add dashboard Filament resources with role authorization Extend filament-blog resources on the dashboard panel and gate CRUD access to super_admin, admin, and dev roles at request time. * feat(blog): add public blog pages, routes, and navigation Expose /blog listing and post detail pages via Livewire on the guest layout, with SEO metadata support and a header navigation link. * fix(blog): resolve PHPStan errors in Blog Show component Initialize the post property and read banner_url via getAttribute so Level 5 analysis passes for the third-party Post model. * feat(organisms): taxonomy enrichment, compound table, and tree of life (#452) (#814) * feat(ui): replace organism pills with searchable table on compound page Closes #452 * fix(organisms): allow linking existing organisms in report changes Relax unique-name validation in the request-changes organism repeater so users can link organisms that already exist in the database. Reject names already attached to the target molecule instead. Fixes #772 * fix(reports): correct select validation in request changes form Align Filament v4 Select option keys with stored state for existing synonyms, CAS, geo locations, organisms, and citations. Resolve the target molecule from form state and report records so options remain valid on Livewire submit requests. Fixes #775 * fix(reports): resolve PHPStan errors in ReportResource Use explicit null checks for molecule synonyms/CAS options and remove unreachable is_string handling for mol_ids array cast. * feat(organisms): add GNF taxonomy enrichment, curation, and compound UI (#452) Store exact-match Global Names taxonomy on organisms, enrich via coconut:organisms-map-ogg with automatic curation fallbacks (vernacular names, sp./spp., NCBI IRI lookup), and surface verified classification on the compound page and in Filament. Closes #452 * fix(tree-of-life): prevent memory exhaustion on explorer page Build a rank-only taxonomy tree without per-organism nodes, cache via the file store, and load tree data in render() so Livewire does not dehydrate the full payload on every request. * fix(tests): give fuzzy taxonomy fixture a populated iri The needingTaxonomyEnrichment scope uses null-safe SQL only; fuzzy organisms with iri, taxonomy, and fetched_at should be excluded from the scope and filtered in PHP via hasExactTaxonomyEnrichment. * fix(search): parse multi-word class and superclass filter values (#819) * Development (#812) * fix: update CSP to allow images and API calls from naturalproducts.net (#788) * chore(deps): resolve composer security advisories (#791) * chore(deps): resolve composer security advisories Bump locked dependencies to clear composer audit, including Filament 4.11.6, Laravel 12.61, Symfony security patches, PhpSpreadsheet 1.30.5, and related transitive packages. * chore(filament): republish assets after dependency upgrade Refresh published Filament 4 and plugin frontend assets to match the updated composer lockfile. * style: apply Laravel Pint across application code Normalize imports and class references in app layer (HTTP kernel, Filament resources, REST API, jobs, events, models) per Pint rules. * style: normalize imports in config and routes Apply Pint class import conventions to published config files and api routes without changing configuration values. * style: normalize imports in factories and migrations Replace FQCN references with imported class names in database factories and migrations for consistency with Pint output. * fix(models): add sampleLocations relationship on Molecule Define the belongsToMany relationship via molecule_organism so Filament and PHPStan can resolve sample location access on molecules. * fix(phpstan): resolve static analysis errors in publish and Filament code Remove stale phpstan-ignore on PublishMolecules and narrow collection updates with explicit null checks. Type the organism molecules bulk action repeater default using Molecule and the sampleLocations relation. * fix: update CSP to allow images and API calls from naturalproducts.net (#792) --------- Co-authored-by: Chandu Nainala <mailcs76@gmail.com> * fix(search): parse multi-word class and superclass filter values Filter queries such as class:Unsaturated+hydrocarbons were split on spaces after URL decoding, causing SQL errors. Add key-aware filter tokenization, normalize text values to hyphen slugs, and align Advanced Search encoding with existing search links. Fixes #818 * chore: resolve Dependabot security advisories across Composer, npm, and Python (#810) Bump vulnerable dependencies flagged by GitHub Dependabot, including Filament 4.8.5+, Symfony, PhpSpreadsheet, axios, Vite 6, and Python export script packages. * feat(geo): extract country from geo locations and add world map (#822) * fix(organisms): allow linking existing organisms in report changes Relax unique-name validation in the request-changes organism repeater so users can link organisms that already exist in the database. Reject names already attached to the target molecule instead. Fixes #772 * fix(reports): correct select validation in request changes form Align Filament v4 Select option keys with stored state for existing synonyms, CAS, geo locations, organisms, and citations. Resolve the target molecule from form state and report records so options remain valid on Livewire submit requests. Fixes #775 * fix(reports): resolve PHPStan errors in ReportResource Use explicit null checks for molecule synonyms/CAS options and remove unreachable is_string handling for mol_ids array cast. * feat(geo): add country extraction and geocoding services Introduce rule-based country parsing from geo location names with Nominatim fallback, backed by config-driven aliases and unit tests. Refs #695 * feat(geo): enrich geo locations on import and via artisan command Centralize geo location creation in GeoLocationService and add coconut:enrich-geo-locations for backfilling country metadata. Refs #695 * feat(globe): add choropleth world map for geo location coverage Add a public /globe page with country-level molecule, organism, and geo-location stats using globe.gl choropleth visualization. Refs #695 * fix(deps): restore linux rollup optional deps in package-lock Regenerate lockfile from development base so CI npm ci includes @rollup/rollup-linux-x64-gnu alongside globe.gl dependencies. * feat(entries): format delimited source entry fields for display (#820) Split ##-grouped and pipe-separated entry values into human-readable lists in the public Entry Details modal and Filament entry views, with clickable DOI and URL links. Closes #581 * fix(commands): migrate NPEdia links to new RIKEN domain (#821) * Development (#812) * fix: update CSP to allow images and API calls from naturalproducts.net (#788) * chore(deps): resolve composer security advisories (#791) * chore(deps): resolve composer security advisories Bump locked dependencies to clear composer audit, including Filament 4.11.6, Laravel 12.61, Symfony security patches, PhpSpreadsheet 1.30.5, and related transitive packages. * chore(filament): republish assets after dependency upgrade Refresh published Filament 4 and plugin frontend assets to match the updated composer lockfile. * style: apply Laravel Pint across application code Normalize imports and class references in app layer (HTTP kernel, Filament resources, REST API, jobs, events, models) per Pint rules. * style: normalize imports in config and routes Apply Pint class import conventions to published config files and api routes without changing configuration values. * style: normalize imports in factories and migrations Replace FQCN references with imported class names in database factories and migrations for consistency with Pint output. * fix(models): add sampleLocations relationship on Molecule Define the belongsToMany relationship via molecule_organism so Filament and PHPStan can resolve sample location access on molecules. * fix(phpstan): resolve static analysis errors in publish and Filament code Remove stale phpstan-ignore on PublishMolecules and narrow collection updates with explicit null checks. Type the organism molecules bulk action repeater default using Molecule and the sampleLocations relation. * fix: update CSP to allow images and API calls from naturalproducts.net (#792) --------- Co-authored-by: Chandu Nainala <mailcs76@gmail.com> * feat(commands): add NPEdia link migration with audit trail Introduce coconut:update-npedia-links to migrate NPEdia source URLs to the new RIKEN domain across entries, collection pivots, and collection metadata, recording OwenIt audits for traceability. Closes #690 * chore: resolve Dependabot security advisories across Composer, npm, and Python (#810) Bump vulnerable dependencies flagged by GitHub Dependabot, including Filament 4.8.5+, Symfony, PhpSpreadsheet, axios, Vite 6, and Python export script packages. * feat(cas): integrate Common Chemistry API for CAS number retrieval an… (#865) * feat(cas): integrate Common Chemistry API for CAS number retrieval and add tests * chore: update composer packages * fix(css): update import path for advanced-tables.css to use plugin.css * feat(security): update Dockerfile base image and add security check for PHP version (#870) chore: enable image pull in GitHub workflows for dev and prod builds fix: update package-lock.json dependencies to latest versions chore: add overrides for postcss-selector-parser in package.json * fix(reports): handle missing reference_id/name in submission reports (#871) * fix(reports): handle missing optional reference_id/name in submission reports Submissions created via the REST API only require canonical_smiles and references, so new_molecule_data may lack reference_id and name. Opening such a report in the dashboard (edit/view) threw 'Undefined array key "reference_id"'. Fall back to null like the other optional fields. * chore(deps): bump symfony packages for 2026-11-04 security advisories Fixes the Composer security audit CI step (http-foundation, http-kernel, routing, mime, serializer, filesystem, html-sanitizer, yaml). * fix(tests): generate CitationFactory data offline with Faker (#873) CitationFactory queried the live Europe PMC API and looped until it found a result with a DOI. When the API returned an error or no DOIs, the loop never ended. rest:documentation calls the factory for every REST resource, so ApiDocumentationTest hung CI until PHP's 300s time limit killed PHPUnit. --------- Co-authored-by: Nisha Sharma <nisha.sharma@uni-jena.de>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
Opening a molecule submission report in the dashboard could fail with:
Submissions made through the REST API only require
canonical_smilesandreferences[].doi(seeApp\Rest\Resources\ReportResourceanddocs/api-documentation.md), sosuggested_changes.new_molecule_datacan be missingreference_idandname.EditReportandViewReportread these keys directly while filling the form.Fix
EditReportandViewReport(mutateFormDataBeforeFill): fall back tonullforreference_idandname, the same way the other optional fields (link,mol_filename, ...) are handled.EditReport::mutateFormDataBeforeSaveandCreateReport::mutateFormDataBeforeCreate: same fallback when buildingnew_molecule_data.ReportResource::approveReportalready treats these fields as optional (?? ''), so approval doesn't change.Testing
php -lon the changed filesreference_id/name, then open the report's edit and view pages in the dashboard.Also included: Symfony security updates
The
Composer security auditCI step began failing on every branch because of Symfony advisories published on 2026-11-04. This PR bumps the affected packages incomposer.lockonly (patch releases, no constraint changes):http-foundation,http-kernel,routing→ 7.4.23,mime→ 7.4.22,serializer,filesystem,html-sanitizer,yaml→ 8.1.11.