feat: store matching metadata on session file registration

Each saved file now records three additional fields:
  matched_identifier    — the literal animal_id_string or animal_name
                          substring that matched in the filename
  animal_id_string_snap — snapshot of animal.animal_id_string at
                          registration time
  animal_name_snap      — snapshot of animal.animal_name at
                          registration time

This lets future consumers (scripts, exports, re-matching) identify
the subject from the file record alone, even if animal data changes.
The matched identifier is also shown in the drop preview UI (↳ 2852).

Tests: 25 total (+2 for new payload fields).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Experiments DB Dev
2026-04-23 13:41:36 -04:00
parent 5cd6d2a8a3
commit 52d5fca5d1
5 changed files with 111 additions and 43 deletions
@@ -0,0 +1,4 @@
ALTER TABLE "session_files"
ADD COLUMN "matched_identifier" TEXT,
ADD COLUMN "animal_id_string_snap" TEXT,
ADD COLUMN "animal_name_snap" TEXT;
+3
View File
@@ -56,6 +56,9 @@ model SessionFile {
file_size BigInt file_size BigInt
file_type String? file_type String?
last_modified BigInt? last_modified BigInt?
matched_identifier String? // the animal ID/name substring that matched in the filename
animal_id_string_snap String? // snapshot of animal.animal_id_string at registration time
animal_name_snap String? // snapshot of animal.animal_name at registration time
notes String? notes String?
created_at DateTime @default(now()) created_at DateTime @default(now())
+5 -1
View File
@@ -2,7 +2,8 @@ const router = require('express').Router();
const prisma = require('../lib/prisma'); const prisma = require('../lib/prisma');
// POST /api/daily-statuses/:id/files — register file metadata (batch) // POST /api/daily-statuses/:id/files — register file metadata (batch)
// Body: [{ filename, file_size, file_type, last_modified, notes }, ...] // Body: [{ filename, file_size, file_type, last_modified,
// matched_identifier, animal_id_string_snap, animal_name_snap, notes }, ...]
router.post('/daily-statuses/:id/files', async (req, res, next) => { router.post('/daily-statuses/:id/files', async (req, res, next) => {
try { try {
const status = await prisma.dailyStatus.findUnique({ where: { id: req.params.id } }); const status = await prisma.dailyStatus.findUnique({ where: { id: req.params.id } });
@@ -31,6 +32,9 @@ router.post('/daily-statuses/:id/files', async (req, res, next) => {
file_size: BigInt(Math.round(f.file_size)), file_size: BigInt(Math.round(f.file_size)),
file_type: f.file_type ?? null, file_type: f.file_type ?? null,
last_modified: f.last_modified != null ? BigInt(Math.round(f.last_modified)) : null, last_modified: f.last_modified != null ? BigInt(Math.round(f.last_modified)) : null,
matched_identifier: f.matched_identifier ?? null,
animal_id_string_snap: f.animal_id_string_snap ?? null,
animal_name_snap: f.animal_name_snap ?? null,
notes: f.notes ?? null, notes: f.notes ?? null,
}, },
}) })
+53 -21
View File
@@ -48,29 +48,58 @@ function fmtBytes(n) {
return `${num} B`; return `${num} B`;
} }
/** Match an array of File objects against animals. Returns { matched: Map<animalId, File[]>, unmatched: File[] } */ /**
* Match an array of File objects against animals.
* Returns {
* matched: Map<animalId, Array<{ file, matchedIdentifier }>>
* unmatched: File[]
* }
* matchedIdentifier is the animal_id_string or animal_name value that triggered the match.
*/
function matchFilesToAnimals(files, animals) { function matchFilesToAnimals(files, animals) {
const patterns = animals.map((a) => ({ animal: a, re: buildAnimalRegex(a) })).filter((x) => x.re); // Build patterns; each entry carries both alternatives so we know which one matched.
const entries = animals
.map((animal) => {
const alts = [];
if (animal.animal_id_string) alts.push({ value: animal.animal_id_string, re: new RegExp(`(?<![a-zA-Z0-9])${escapeRegex(animal.animal_id_string)}(?![a-zA-Z0-9])`, 'i') });
if (animal.animal_name) {
const words = animal.animal_name.trim().split(/\s+/).map(escapeRegex);
const pat = words.length === 1 ? words[0] : words.join('[\\s_\\-]+');
alts.push({ value: animal.animal_name, re: new RegExp(`(?<![a-zA-Z0-9])${pat}(?![a-zA-Z0-9])`, 'i') });
}
return { animal, alts };
})
.filter((e) => e.alts.length > 0);
const matched = new Map(animals.map((a) => [a.id, []])); const matched = new Map(animals.map((a) => [a.id, []]));
const unmatched = []; const unmatched = [];
for (const file of files) { for (const file of files) {
const hit = patterns.find(({ re }) => re.test(file.name)); let found = false;
if (hit) { for (const { animal, alts } of entries) {
matched.get(hit.animal.id).push(file); for (const { value, re } of alts) {
} else { if (re.test(file.name)) {
unmatched.push(file); matched.get(animal.id).push({ file, matchedIdentifier: value });
found = true;
break;
} }
} }
if (found) break;
}
if (!found) unmatched.push(file);
}
return { matched, unmatched }; return { matched, unmatched };
} }
// ── Sub-components ──────────────────────────────────────────────────────────── // ── Sub-components ────────────────────────────────────────────────────────────
function FileRow({ file, onRemove }) { function FileRow({ file, matchedIdentifier, onRemove }) {
return ( return (
<div className="flex items-center justify-between py-1 gap-2 text-xs"> <div className="flex items-center justify-between py-1 gap-2 text-xs">
<span className="font-mono text-gray-700 truncate min-w-0 flex-1">{file.name}</span> <span className="font-mono text-gray-700 truncate min-w-0 flex-1">{file.name}</span>
{matchedIdentifier && (
<span className="text-blue-400 shrink-0 font-mono" title="matched by"> {matchedIdentifier}</span>
)}
<span className="text-gray-400 shrink-0">{fmtBytes(file.size)}</span> <span className="text-gray-400 shrink-0">{fmtBytes(file.size)}</span>
<span className="text-gray-400 shrink-0 w-12 text-right">{classifyFile(file.name)}</span> <span className="text-gray-400 shrink-0 w-12 text-right">{classifyFile(file.name)}</span>
{onRemove && ( {onRemove && (
@@ -185,13 +214,16 @@ export default function FileDropZone({ rows, date }) {
try { try {
const results = await Promise.all( const results = await Promise.all(
statusRows.map(async ({ animal, status }) => { statusRows.map(async ({ animal, status }) => {
const files = dropped.matched.get(animal.id) ?? []; const entries = dropped.matched.get(animal.id) ?? [];
if (files.length === 0) return { statusId: status.id, saved: [] }; if (entries.length === 0) return { statusId: status.id, saved: [] };
const payload = files.map((f) => ({ const payload = entries.map(({ file, matchedIdentifier }) => ({
filename: f.name, filename: file.name,
file_size: f.size, file_size: file.size,
file_type: classifyFile(f.name), file_type: classifyFile(file.name),
last_modified: f.lastModified ?? null, last_modified: file.lastModified ?? null,
matched_identifier: matchedIdentifier,
animal_id_string_snap: animal.animal_id_string ?? null,
animal_name_snap: animal.animal_name ?? null,
})); }));
const saved = await sessionFilesApi.create(status.id, payload); const saved = await sessionFilesApi.create(status.id, payload);
return { statusId: status.id, saved }; return { statusId: status.id, saved };
@@ -231,7 +263,7 @@ export default function FileDropZone({ rows, date }) {
// ── Derived counts ───────────────────────────────────────────────────────── // ── Derived counts ─────────────────────────────────────────────────────────
const totalDropped = dropped ? [...dropped.matched.values()].reduce((n, a) => n + a.length, 0) : 0; const totalDropped = dropped ? [...dropped.matched.values()].reduce((n, a) => n + a.length, 0) : 0; // entries are { file, matchedIdentifier }
const totalUnmatched = dropped?.unmatched.length ?? 0; const totalUnmatched = dropped?.unmatched.length ?? 0;
const totalSaved = Object.values(savedByStatus).reduce((n, a) => n + a.length, 0); const totalSaved = Object.values(savedByStatus).reduce((n, a) => n + a.length, 0);
const hasSaved = totalSaved > 0; const hasSaved = totalSaved > 0;
@@ -281,8 +313,8 @@ export default function FileDropZone({ rows, date }) {
{/* Per-subject matched files */} {/* Per-subject matched files */}
{statusRows.map(({ animal, status }) => { {statusRows.map(({ animal, status }) => {
const files = dropped.matched.get(animal.id) ?? []; const entries = dropped.matched.get(animal.id) ?? [];
if (files.length === 0) return null; if (entries.length === 0) return null;
return ( return (
<div key={animal.id} className="bg-white border border-gray-200 rounded-xl p-4"> <div key={animal.id} className="bg-white border border-gray-200 rounded-xl p-4">
<div className="flex items-center justify-between mb-2"> <div className="flex items-center justify-between mb-2">
@@ -292,11 +324,11 @@ export default function FileDropZone({ rows, date }) {
<span className="ml-1.5 text-xs font-normal text-gray-400">{animal.animal_id_string}</span> <span className="ml-1.5 text-xs font-normal text-gray-400">{animal.animal_id_string}</span>
)} )}
</span> </span>
<span className="text-xs text-gray-400">{files.length} file{files.length !== 1 ? 's' : ''}</span> <span className="text-xs text-gray-400">{entries.length} file{entries.length !== 1 ? 's' : ''}</span>
</div> </div>
<div className="divide-y divide-gray-50"> <div className="divide-y divide-gray-50">
{files.map((f, i) => ( {entries.map(({ file, matchedIdentifier }, i) => (
<FileRow key={f.name + i} file={f} onRemove={() => removeMatchedFile(animal.id, i)} /> <FileRow key={file.name + i} file={file} matchedIdentifier={matchedIdentifier} onRemove={() => removeMatchedFile(animal.id, i)} />
))} ))}
</div> </div>
</div> </div>
+27 -2
View File
@@ -249,8 +249,8 @@ describe('remove files before saving', () => {
// ── Saving ───────────────────────────────────────────────────────────────────── // ── Saving ─────────────────────────────────────────────────────────────────────
describe('saving metadata', () => { describe('saving metadata', () => {
it('calls sessionFilesApi.create with correct payload for each subject', async () => { it('calls sessionFilesApi.create with correct payload including matching metadata', async () => {
const file = makeFile(`${DATE}_2852_trials.csv`); // size = 1 byte (content 'x') const file = makeFile(`${DATE}_2852_trials.csv`);
const savedFile = makeSavedFile('f1', 's-a1', file.name, String(file.size)); const savedFile = makeSavedFile('f1', 's-a1', file.name, String(file.size));
sessionFilesApi.create.mockResolvedValue([savedFile]); sessionFilesApi.create.mockResolvedValue([savedFile]);
@@ -267,6 +267,31 @@ describe('saving metadata', () => {
filename: `${DATE}_2852_trials.csv`, filename: `${DATE}_2852_trials.csv`,
file_size: file.size, file_size: file.size,
file_type: 'CSV', file_type: 'CSV',
matched_identifier: '2852',
animal_id_string_snap: '2852',
animal_name_snap: 'Rat 2852',
}),
])
)
);
});
it('sets matched_identifier to animal_name when file matched by name', async () => {
const animal = { id: 'b1', animal_name: 'Rat A', animal_id_string: null };
const row = { animal, status: { id: 's-b1', animal_id: 'b1', date: `${DATE}T00:00:00.000Z` } };
sessionFilesApi.create.mockResolvedValue([]);
render(<FileDropZone rows={[row]} date={DATE} />);
dropFiles([makeFile(`${DATE}_Rat_A_trials.csv`)]);
await waitFor(() => screen.getByTestId('save-btn'));
fireEvent.click(screen.getByTestId('save-btn'));
await waitFor(() =>
expect(sessionFilesApi.create).toHaveBeenCalledWith(
's-b1',
expect.arrayContaining([
expect.objectContaining({
matched_identifier: 'Rat A',
animal_id_string_snap: null,
animal_name_snap: 'Rat A',
}), }),
]) ])
) )