Add language config and locale-aware parsing

Introduce centralized language configuration and wire locale-aware behavior across scraping and ranking. Adds src/lib/constants/language-config.ts with per-language scraping rules, stop words, and character replacements; replaces AudibleRegion.isEnglish with a language field in types and AUDIBLE_REGIONS. Update AudibleService, ebook scraper, processors, and API routes to use getLanguageForRegion so Anna's Archive searches, scraping selectors, runtime/rating parsing, and ranking use language-specific params and filters. Extend ranking algorithm to accept stopWords and characterReplacements and apply them during normalization and matching. Update UI selects to mark non-English regions and adjust tests accordingly.
This commit is contained in:
kikootwo
2026-02-20 06:32:44 -05:00
parent c146383735
commit 5d8ac2f73d
18 changed files with 525 additions and 112 deletions
@@ -18,6 +18,8 @@ import { findPlexMatch } from '@/lib/utils/audiobook-matcher';
import { getAudibleService } from '@/lib/integrations/audible.service';
import { RMABLogger } from '@/lib/utils/logger';
import { resolveInteractiveSearchAccess } from '@/lib/utils/permissions';
import { getLanguageForRegion } from '@/lib/constants/language-config';
import type { AudibleRegion } from '@/lib/types/audible';
import {
searchByAsin,
searchByTitle,
@@ -227,6 +229,11 @@ export async function POST(
const format = preferredFormat || 'epub';
const annasBaseUrl = baseUrl || 'https://annas-archive.li';
// Get language code from Audible region config
const region = await configService.getAudibleRegion() as AudibleRegion;
const langConfig = getLanguageForRegion(region);
const languageCode = langConfig.annasArchiveLang;
if (!isAnnasArchiveEnabled && !isIndexerSearchEnabled) {
return NextResponse.json(
{ error: 'No ebook sources enabled. Enable Anna\'s Archive or Indexer Search in settings.' },
@@ -250,7 +257,8 @@ export async function POST(
audiobook.author,
format,
annasBaseUrl,
flaresolverrUrl || undefined
flaresolverrUrl || undefined,
languageCode
).catch((err) => {
logger.error(`Anna's Archive search failed: ${err.message}`);
return null;
@@ -322,7 +330,8 @@ async function searchAnnasArchiveForInteractive(
author: string,
preferredFormat: string,
baseUrl: string,
flaresolverrUrl?: string
flaresolverrUrl?: string,
languageCode: string = 'en'
): Promise<EbookSearchResult[]> {
let md5: string | null = null;
let searchMethod: 'asin' | 'title' = 'title';
@@ -330,7 +339,7 @@ async function searchAnnasArchiveForInteractive(
// Try ASIN search first
if (asin) {
logger.info(`Searching Anna's Archive by ASIN: ${asin}`);
md5 = await searchByAsin(asin, preferredFormat, baseUrl, undefined, flaresolverrUrl);
md5 = await searchByAsin(asin, preferredFormat, baseUrl, undefined, flaresolverrUrl, languageCode);
if (md5) {
searchMethod = 'asin';
logger.info(`Found via ASIN: ${md5}`);
@@ -340,7 +349,7 @@ async function searchAnnasArchiveForInteractive(
// Fallback to title search
if (!md5) {
logger.info(`Searching Anna's Archive by title: "${title}"`);
md5 = await searchByTitle(title, author, preferredFormat, baseUrl, undefined, flaresolverrUrl);
md5 = await searchByTitle(title, author, preferredFormat, baseUrl, undefined, flaresolverrUrl, languageCode);
if (md5) {
logger.info(`Found via title: ${md5}`);
}
@@ -461,6 +470,10 @@ async function searchIndexersForInteractive(
return [];
}
// Get language-specific stop words for ranking
const rankRegion = await configService.getAudibleRegion() as AudibleRegion;
const rankLangConfig = getLanguageForRegion(rankRegion);
// Rank results with ebook scoring
const rankedResults = rankEbookTorrents(allResults, {
title,
@@ -470,6 +483,8 @@ async function searchIndexersForInteractive(
indexerPriorities,
flagConfigs,
requireAuthor: false,
stopWords: rankLangConfig.stopWords,
characterReplacements: rankLangConfig.characterReplacements,
});
// Convert to unified result type
@@ -10,6 +10,8 @@ import { requireAuth, AuthenticatedRequest } from '@/lib/middleware/auth';
import { getProwlarrService } from '@/lib/integrations/prowlarr.service';
import { rankTorrents } from '@/lib/utils/ranking-algorithm';
import { groupIndexersByCategories, getGroupDescription } from '@/lib/utils/indexer-grouping';
import { getLanguageForRegion } from '@/lib/constants/language-config';
import type { AudibleRegion } from '@/lib/types/audible';
import { z } from 'zod';
import { RMABLogger } from '@/lib/utils/logger';
@@ -140,13 +142,19 @@ export async function POST(request: NextRequest) {
logger.info(`Will filter ${belowThreshold.length} results < ${sizeMBThreshold} MB (likely ebooks)`);
}
// Get language-specific stop words for ranking
const region = await configService.getAudibleRegion() as AudibleRegion;
const langConfig = getLanguageForRegion(region);
// Rank torrents using the ranking algorithm with indexer priorities and flag configs
// Note: rankTorrents now filters out results < 20 MB internally
// requireAuthor: false - interactive search, show all results for user decision
const rankedResults = rankTorrents(results, { title, author, durationMinutes }, {
indexerPriorities,
flagConfigs,
requireAuthor: false // Interactive mode - let user decide
requireAuthor: false, // Interactive mode - let user decide
stopWords: langConfig.stopWords,
characterReplacements: langConfig.characterReplacements,
});
// Log filter results
@@ -14,6 +14,8 @@ import { getProwlarrService } from '@/lib/integrations/prowlarr.service';
import { rankEbookTorrents, RankedEbookTorrent } from '@/lib/utils/ranking-algorithm';
import { groupIndexersByCategories, getGroupDescription } from '@/lib/utils/indexer-grouping';
import { RMABLogger } from '@/lib/utils/logger';
import { getLanguageForRegion } from '@/lib/constants/language-config';
import type { AudibleRegion } from '@/lib/types/audible';
import {
searchByAsin,
searchByTitle,
@@ -121,6 +123,11 @@ export async function POST(
const format = preferredFormat || 'epub';
const annasBaseUrl = baseUrl || 'https://annas-archive.li';
// Get language code from Audible region config
const region = await configService.getAudibleRegion() as AudibleRegion;
const langConfig = getLanguageForRegion(region);
const languageCode = langConfig.annasArchiveLang;
if (!isAnnasArchiveEnabled && !isIndexerSearchEnabled) {
return NextResponse.json(
{ error: 'No ebook sources enabled. Enable Anna\'s Archive or Indexer Search in settings.' },
@@ -145,7 +152,8 @@ export async function POST(
audiobook.author,
format,
annasBaseUrl,
flaresolverrUrl || undefined
flaresolverrUrl || undefined,
languageCode
).catch((err) => {
logger.error(`Anna's Archive search failed: ${err.message}`);
return null;
@@ -217,7 +225,8 @@ async function searchAnnasArchiveForInteractive(
author: string,
preferredFormat: string,
baseUrl: string,
flaresolverrUrl?: string
flaresolverrUrl?: string,
languageCode: string = 'en'
): Promise<EbookSearchResult[]> {
let md5: string | null = null;
let searchMethod: 'asin' | 'title' = 'title';
@@ -225,7 +234,7 @@ async function searchAnnasArchiveForInteractive(
// Try ASIN search first
if (asin) {
logger.info(`Searching Anna's Archive by ASIN: ${asin}`);
md5 = await searchByAsin(asin, preferredFormat, baseUrl, undefined, flaresolverrUrl);
md5 = await searchByAsin(asin, preferredFormat, baseUrl, undefined, flaresolverrUrl, languageCode);
if (md5) {
searchMethod = 'asin';
logger.info(`Found via ASIN: ${md5}`);
@@ -235,7 +244,7 @@ async function searchAnnasArchiveForInteractive(
// Fallback to title search
if (!md5) {
logger.info(`Searching Anna's Archive by title: "${title}"`);
md5 = await searchByTitle(title, author, preferredFormat, baseUrl, undefined, flaresolverrUrl);
md5 = await searchByTitle(title, author, preferredFormat, baseUrl, undefined, flaresolverrUrl, languageCode);
if (md5) {
logger.info(`Found via title: ${md5}`);
}
@@ -356,6 +365,10 @@ async function searchIndexersForInteractive(
return [];
}
// Get language-specific stop words for ranking
const rankRegion = await configService.getAudibleRegion() as AudibleRegion;
const rankLangConfig = getLanguageForRegion(rankRegion);
// Rank results with ebook scoring
// Use requireAuthor=false for interactive mode (let user decide)
const rankedResults = rankEbookTorrents(allResults, {
@@ -366,6 +379,8 @@ async function searchIndexersForInteractive(
indexerPriorities,
flagConfigs,
requireAuthor: false,
stopWords: rankLangConfig.stopWords,
characterReplacements: rankLangConfig.characterReplacements,
});
// Log ranking debug info (same format as search-ebook.processor.ts)
@@ -9,6 +9,8 @@ import { prisma } from '@/lib/db';
import { getProwlarrService } from '@/lib/integrations/prowlarr.service';
import { rankTorrents } from '@/lib/utils/ranking-algorithm';
import { groupIndexersByCategories, getGroupDescription } from '@/lib/utils/indexer-grouping';
import { getLanguageForRegion } from '@/lib/constants/language-config';
import type { AudibleRegion } from '@/lib/types/audible';
import { RMABLogger } from '@/lib/utils/logger';
import { resolveInteractiveSearchAccess } from '@/lib/utils/permissions';
@@ -189,6 +191,10 @@ export async function POST(
}
}
// Get language-specific stop words for ranking
const region = await configService.getAudibleRegion() as AudibleRegion;
const langConfig = getLanguageForRegion(region);
// Rank torrents using the ranking algorithm with indexer priorities and flag configs
// Always use the audiobook's title/author for ranking (not custom search query)
// requireAuthor: false - interactive mode, show all results for user decision
@@ -199,7 +205,9 @@ export async function POST(
}, {
indexerPriorities,
flagConfigs,
requireAuthor: false // Interactive mode - let user decide
requireAuthor: false, // Interactive mode - let user decide
stopWords: langConfig.stopWords,
characterReplacements: langConfig.characterReplacements,
});
// No threshold filtering for interactive search - show all results