breakpilot-lehrer/studio-v2/app/vocab-worksheet/usePageProcessing.ts

import type { VocabularyEntry, OcrPrompts, IpaMode, SyllableMode } from './types'
import { getApiBase } from './constants'

/**
 * Process a single page and return vocabulary + optional scan quality info.
 */
export async function processSinglePage(
  sessionId: string,
  pageIndex: number,
  ipa: IpaMode,
  syllable: SyllableMode,
  ocrPrompts: OcrPrompts,
  ocrEnhance: boolean,
  ocrMaxCols: number,
  ocrMinConf: number,
): Promise<{ success: boolean; vocabulary: VocabularyEntry[]; error?: string; scanQuality?: any }> {
  const API_BASE = getApiBase()

  try {
    const params = new URLSearchParams({
      ipa_mode: ipa,
      syllable_mode: syllable,
      enhance: String(ocrEnhance),
      max_cols: String(ocrMaxCols),
      min_conf: String(ocrMinConf),
    })
    const res = await fetch(`${API_BASE}/api/v1/vocab/sessions/${sessionId}/process-single-page/${pageIndex}?${params}`, {
      method: 'POST',
      headers: { 'Content-Type': 'application/json' },
      body: JSON.stringify({ ocr_prompts: ocrPrompts }),
    })

    if (!res.ok) {
      const errBody = await res.json().catch(() => ({}))
      const detail = errBody.detail || `HTTP ${res.status}`
      return { success: false, vocabulary: [], error: `Seite ${pageIndex + 1}: ${detail}` }
    }

    const data = await res.json()
    if (!data.success) {
      return { success: false, vocabulary: [], error: data.error || `Seite ${pageIndex + 1}: Unbekannter Fehler` }
    }

    return { success: true, vocabulary: data.vocabulary || [], scanQuality: data.scan_quality }
  } catch (e) {
    return { success: false, vocabulary: [], error: `Seite ${pageIndex + 1}: ${e instanceof Error ? e.message : 'Netzwerkfehler'}` }
  }
}

/**
 * Reprocess pages with updated IPA/syllable settings.
 * Returns the new vocabulary array.
 */
export async function reprocessPagesFlow(
  sessionId: string,
  pagesToReprocess: number[],
  ipa: IpaMode,
  syllable: SyllableMode,
  ocrPrompts: OcrPrompts,
  ocrEnhance: boolean,
  ocrMaxCols: number,
  ocrMinConf: number,
  setExtractionStatus: (s: string) => void,
): Promise<{ vocabulary: VocabularyEntry[]; qualityInfo: string }> {
  const API_BASE = getApiBase()
  const allVocab: VocabularyEntry[] = []
  let lastQuality: any = null

  for (const pageIndex of pagesToReprocess) {
    setExtractionStatus(`Verarbeite Seite ${pageIndex + 1}...`)
    try {
      const params = new URLSearchParams({
        ipa_mode: ipa,
        syllable_mode: syllable,
        enhance: String(ocrEnhance),
        max_cols: String(ocrMaxCols),
        min_conf: String(ocrMinConf),
      })
      const res = await fetch(`${API_BASE}/api/v1/vocab/sessions/${sessionId}/process-single-page/${pageIndex}?${params}`, {
        method: 'POST',
        headers: { 'Content-Type': 'application/json' },
        body: JSON.stringify({ ocr_prompts: ocrPrompts }),
      })
      if (res.ok) {
        const data = await res.json()
        if (data.vocabulary) allVocab.push(...data.vocabulary)
        if (data.scan_quality) lastQuality = data.scan_quality
      }
    } catch { /* ignore individual page failures */ }
  }

  const qualityInfo = lastQuality
    ? ` | Qualitaet: ${lastQuality.quality_pct}%${lastQuality.is_degraded ? ' (degradiert!)' : ''} | Blur: ${lastQuality.blur_score} | Kontrast: ${lastQuality.contrast_score}`
    : ''

  return { vocabulary: allVocab, qualityInfo }
}