added backend-mgmt for dataset/dataset_entry and utterances, this includes upload and also displaying of the dataset on the admin pages

src/app/actions/datasets.ts

@@ -0,0 +1,46 @@
+'use server';
+
+import db from '@/lib/db';
+import { dataset } from '@/lib/model/dataset';
+import { eq } from 'drizzle-orm';
+import { revalidatePath } from 'next/cache';
+
+export async function createDataset({ name }: { name: string }) {
+  if (!name || !name.trim()) {
+    throw new Error('Dataset name is required');
+  }
+
+  try {
+    const result = await db.insert(dataset).values({ name: name.trim() });
+    revalidatePath('/admin');
+    return result;
+  } catch (error) {
+    console.error('Error creating dataset:', error);
+    throw new Error('Failed to create dataset');
+  }
+}
+
+export async function updateDataset(
+  id: number,
+  updates: { name?: string }
+) {
+  if (updates.name !== undefined && !updates.name.trim()) {
+    throw new Error('Dataset name cannot be empty');
+  }
+
+  try {
+    const result = await db
+      .update(dataset)
+      .set({
+        name: updates.name?.trim(),
+        updatedAt: new Date(),
+      })
+      .where(eq(dataset.id, id));
+    revalidatePath(`/admin/datasets/${id}`);
+    revalidatePath('/admin');
+    return result;
+  } catch (error) {
+    console.error('Error updating dataset:', error);
+    throw new Error('Failed to update dataset');
+  }
+}

src/app/actions/get-dataset-entries.ts

@@ -0,0 +1,70 @@
+'use server';
+
+import db from '@/lib/db';
+import { dataset_entry } from '@/lib/model/dataset_entry';
+import { count, eq, and, like } from 'drizzle-orm';
+
+const ENTRIES_PER_PAGE = 20;
+
+interface FilterParams {
+  speakerId?: string;
+  modelName?: string;
+  dialect?: string;
+  iteration?: number;
+  utteranceId?: string;
+}
+
+export async function getDatasetEntries(
+  datasetId: number,
+  page: number = 1,
+  filters?: FilterParams
+) {
+  const offset = (page - 1) * ENTRIES_PER_PAGE;
+
+  // Build where conditions
+  const conditions = [eq(dataset_entry.datasetId, datasetId)];
+
+  if (filters?.speakerId) {
+    conditions.push(like(dataset_entry.speakerId, `%${filters.speakerId}%`));
+  }
+
+  if (filters?.modelName) {
+    conditions.push(like(dataset_entry.modelName, `%${filters.modelName}%`));
+  }
+
+  if (filters?.dialect) {
+    conditions.push(like(dataset_entry.dialect, `%${filters.dialect}%`));
+  }
+
+  if (filters?.iteration !== undefined) {
+    conditions.push(eq(dataset_entry.iteration, filters.iteration));
+  }
+
+  if (filters?.utteranceId) {
+    conditions.push(eq(dataset_entry.utteranceId, filters.utteranceId));
+  }
+
+  const whereClause = and(...conditions);
+
+  const entries = await db
+    .select()
+    .from(dataset_entry)
+    .where(whereClause)
+    .limit(ENTRIES_PER_PAGE)
+    .offset(offset);
+
+  const totalResult = await db
+    .select({ count: count() })
+    .from(dataset_entry)
+    .where(whereClause);
+
+  const total = totalResult[0]?.count || 0;
+  const hasMore = offset + ENTRIES_PER_PAGE < total;
+
+  return {
+    entries,
+    hasMore,
+    total,
+    page,
+  };
+}

src/app/actions/get-filter-options.ts

@@ -0,0 +1,27 @@
+'use server';
+
+import db from '@/lib/db';
+import { dataset_entry } from '@/lib/model/dataset_entry';
+import { eq } from 'drizzle-orm';
+
+export async function getFilterOptions(datasetId: number) {
+  const entries = await db
+    .select()
+    .from(dataset_entry)
+    .where(eq(dataset_entry.datasetId, datasetId));
+
+  // Extract unique values for each filter
+  const speakerIds = [...new Set(entries.map(e => e.speakerId))].sort();
+  const modelNames = [...new Set(entries.map(e => e.modelName))].sort();
+  const dialects = [...new Set(entries.map(e => e.dialect))].sort();
+  const iterations = [...new Set(entries.map(e => e.iteration))].sort((a, b) => a - b);
+  const utteranceIds = [...new Set(entries.map(e => e.utteranceId).filter(Boolean))].sort();
+
+  return {
+    speakerIds,
+    modelNames,
+    dialects,
+    iterations,
+    utteranceIds,
+  };
+}

src/app/actions/remove-dataset-entries.ts

@@ -0,0 +1,34 @@
+'use server';
+
+import db from '@/lib/db';
+import { dataset_entry } from '@/lib/model/dataset_entry';
+import { eq } from 'drizzle-orm';
+import { revalidatePath } from 'next/cache';
+import { rm } from 'fs/promises';
+import { join } from 'path';
+import { existsSync } from 'fs';
+
+export async function removeAllDatasetEntries(datasetId: number) {
+  try {
+    // Delete all entries from database
+    await db.delete(dataset_entry).where(eq(dataset_entry.datasetId, datasetId));
+
+    // Delete dataset files from public folder
+    const datasetDir = join(process.cwd(), 'public', 'datasets', datasetId.toString());
+    if (existsSync(datasetDir)) {
+      await rm(datasetDir, { recursive: true, force: true });
+    }
+
+    revalidatePath(`/admin/datasets/${datasetId}`);
+
+    return {
+      success: true,
+      message: 'All dataset entries have been removed',
+    };
+  } catch (error) {
+    console.error('Error removing dataset entries:', error);
+    throw new Error(
+      error instanceof Error ? error.message : 'Failed to remove dataset entries'
+    );
+  }
+}

src/app/actions/upload-utterances.ts

@@ -0,0 +1,119 @@
+'use server';
+
+import db from '@/lib/db';
+import { dataset_utterance } from '@/lib/model/utterance';
+import { eq, and } from 'drizzle-orm';
+import { revalidatePath } from 'next/cache';
+
+export async function uploadDatasetUtterances(
+  datasetId: number,
+  formData: FormData
+) {
+  const file = formData.get('file') as File;
+
+  if (!file) {
+    throw new Error('No file provided');
+  }
+
+  if (!file.name.endsWith('.csv')) {
+    throw new Error('File must be a CSV file');
+  }
+
+  try {
+    const csvContent = await file.text();
+    const lines = csvContent.split('\n').filter((line) => line.trim());
+
+    if (lines.length < 2) {
+      throw new Error('CSV is empty or invalid');
+    }
+
+    // Parse CSV header
+    const headers = lines[0].split(',').map((h) => h.trim());
+
+    if (!headers.includes('id') || !headers.includes('text')) {
+      throw new Error('CSV must have "id" and "text" columns');
+    }
+
+    // Get indices of id and text columns
+    const idIndex = headers.indexOf('id');
+    const textIndex = headers.indexOf('text');
+
+    // Parse utterances
+    const utterances: typeof dataset_utterance.$inferInsert[] = [];
+
+    for (let i = 1; i < lines.length; i++) {
+      // Handle CSV parsing with quotes
+      const line = lines[i];
+      const parts: string[] = [];
+      let current = '';
+      let inQuotes = false;
+
+      for (let j = 0; j < line.length; j++) {
+        const char = line[j];
+        const nextChar = line[j + 1];
+
+        if (char === '"') {
+          if (inQuotes && nextChar === '"') {
+            current += '"';
+            j++; // Skip next quote
+          } else {
+            inQuotes = !inQuotes;
+          }
+        } else if (char === ',' && !inQuotes) {
+          parts.push(current.trim());
+          current = '';
+        } else {
+          current += char;
+        }
+      }
+      parts.push(current.trim());
+
+      if (parts.length > Math.max(idIndex, textIndex)) {
+        const id = parts[idIndex]?.trim();
+        const text = parts[textIndex]?.trim();
+
+        if (id && text) {
+          utterances.push({
+            datasetId,
+            id,
+            text,
+          });
+        }
+      }
+    }
+
+    if (utterances.length === 0) {
+      throw new Error('No valid utterances found in CSV');
+    }
+
+    // Check for existing utterances to avoid duplicates
+    const existingUtterances = await db
+      .select({ id: dataset_utterance.id })
+      .from(dataset_utterance)
+      .where(eq(dataset_utterance.datasetId, datasetId));
+
+    const existingIds = new Set(existingUtterances.map((u) => u.id));
+
+    // Filter out duplicates
+    const newUtterances = utterances.filter((u) => !existingIds.has(u.id));
+
+    if (newUtterances.length === 0) {
+      throw new Error('All utterances already exist in the dataset');
+    }
+
+    // Insert new utterances into database
+    await db.insert(dataset_utterance).values(newUtterances);
+
+    revalidatePath(`/admin/datasets/${datasetId}`);
+
+    return {
+      success: true,
+      utterancesCreated: newUtterances.length,
+    };
+  } catch (error) {
+    console.error('Error uploading utterances:', error);
+    throw new Error(
+      error instanceof Error ? error.message : 'Failed to upload utterances'
+    );
+  }
+}

src/app/actions/upload.ts

@@ -0,0 +1,157 @@
+'use server';
+
+import db from '@/lib/db';
+import { dataset_entry } from '@/lib/model/dataset_entry';
+import { dataset_utterance } from '@/lib/model/utterance';
+import { revalidatePath } from 'next/cache';
+import { eq } from 'drizzle-orm';
+import { writeFile, mkdir, readFile, rm, readdir } from 'fs/promises';
+import { join } from 'path';
+import { existsSync } from 'fs';
+import extract from 'extract-zip';
+
+interface DatasetEntryInput {
+  id: string;
+  audio_file: string;
+  duration_ms: string;
+  utt_id: string;
+  speaker: string;
+  model: string;
+  dialect: string;
+  iteration: string;
+}
+
+export async function uploadDatasetEntries(
+  datasetId: number,
+  formData: FormData
+) {
+  const file = formData.get('file') as File;
+
+  if (!file) {
+    throw new Error('No file provided');
+  }
+
+  if (!file.name.endsWith('.zip')) {
+    throw new Error('File must be a ZIP archive');
+  }
+
+  const tempDir = join(process.cwd(), 'tmp', `upload-${Date.now()}`);
+  const datasetDir = join(process.cwd(), 'public', 'datasets', datasetId.toString());
+
+  try {
+    // Create temp directory
+    await mkdir(tempDir, { recursive: true });
+
+    // Save uploaded file to temp location
+    const tempZipPath = join(tempDir, file.name);
+    const arrayBuffer = await file.arrayBuffer();
+    const buffer = Buffer.from(arrayBuffer);
+    await writeFile(tempZipPath, buffer);
+
+    // Extract ZIP file
+    await extract( tempZipPath, { dir: tempDir } );
+
+    // Read metadata.csv
+    const metadataPath = join(tempDir, 'metadata.csv');
+    if (!existsSync(metadataPath)) {
+      throw new Error('metadata.csv not found in ZIP');
+    }
+
+    const metadataContent = await readFile(metadataPath, 'utf-8');
+    const lines = metadataContent.split('\n').filter(line => line.trim());
+
+    if (lines.length < 2) {
+      throw new Error('metadata.csv is empty or invalid');
+    }
+
+    // Parse CSV header
+    const headers = lines[0].split(',').map(h => h.trim());
+
+    // Create dataset directory
+    await mkdir(datasetDir, { recursive: true });
+
+    // Parse and insert entries
+    const entries: typeof dataset_entry.$inferInsert[] = [];
+
+    for (let i = 1; i < lines.length; i++) {
+      const values = lines[i].split(',').map(v => v.trim());
+
+      if (values.length !== headers.length) {
+        continue; // Skip malformed lines
+      }
+
+      const row: Record<string, string> = {};
+      headers.forEach((header, index) => {
+        row[header] = values[index];
+      });
+
+      const audioFile = row.audio_file as string;
+      // Try different path separators and remove /output prefix
+      const relativePath = audioFile.replace(/\\/g, '/').replace(/^output\//, '');
+      const audioPath = join(tempDir, relativePath);
+
+      if (existsSync(audioPath)) {
+        // Copy audio file to dataset directory with external ID as filename
+        const audioBuffer = await readFile(audioPath);
+        const fileExtension = audioFile.substring(audioFile.lastIndexOf('.'));
+        const destPath = join(datasetDir, `${row.id}${fileExtension}`);
+
+        await writeFile(destPath, audioBuffer);
+
+        entries.push({
+          datasetId,
+          externalId: row.id,
+          speakerId: row.speaker,
+          modelName: row.model,
+          fileName: relativePath,
+          utteranceId: row.utt_id || undefined,
+          dialect: row.dialect,
+          iteration: parseInt(row.iteration, 10),
+        });
+      }
+    }
+
+    if (entries.length === 0) {
+      throw new Error('No valid entries found in metadata.csv');
+    }
+
+    // Check for existing entries to avoid duplicates
+    const existingEntries = await db
+      .select({ externalId: dataset_entry.externalId })
+      .from(dataset_entry)
+      .where(eq(dataset_entry.datasetId, datasetId));
+
+    const existingExternalIds = new Set(existingEntries.map(e => e.externalId));
+
+    // Filter out duplicates
+    const newEntries = entries.filter(entry => !existingExternalIds.has(entry.externalId));
+
+    if (newEntries.length === 0) {
+      throw new Error('All entries already exist in the dataset');
+    }
+
+    // Insert new entries into database
+    await db.insert(dataset_entry).values(newEntries);
+
+    revalidatePath(`/admin/datasets/${datasetId}`);
+
+    return {
+      success: true,
+      entriesCreated: entries.length,
+    };
+  } catch (error) {
+    console.error('Error uploading dataset entries:', error);
+    throw new Error(
+      error instanceof Error ? error.message : 'Failed to upload dataset entries'
+    );
+  } finally {
+    // Clean up temp directory
+    try {
+      if (existsSync(tempDir)) {
+        await rm(tempDir, { recursive: true, force: true });
+      }
+    } catch (cleanupError) {
+      console.error('Error cleaning up temp directory:', cleanupError);
+    }
+  }
+}