Files
skrift-backend/src/services/scriptalizer-service.js
s4luorth 25ffb3727b
All checks were successful
Build & Deploy / deploy (push) Successful in 19s
Scriptalizer: nach Zeichenlaenge batchen (32000-Limit)
Fixe 18er-Batches sprengen bei langen/mehrseitigen Texten das InputText-Limit der
Scriptalizer-API (32000 Zeichen) -> 'Input length limit 32000 exceeded'. Jetzt werden
Batches nach kumulativer Zeichenlaenge (max 28000, Sicherheitsabstand) gebildet, die
Anzahl 18 bleibt als zusaetzlicher Deckel.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-08-29 10:00:58 +02:00

251 lines
7.4 KiB
JavaScript
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
const https = require("https");
const { parseStringPromise } = require("xml2js");
const config = require("../config");
// API Request Counter (resets daily)
const stats = {
requestsToday: 0,
textsToday: 0,
lastResetDate: new Date().toDateString(),
};
/**
* Prüft ob der Zähler zurückgesetzt werden muss (neuer Tag)
*/
function checkAndResetDailyStats() {
const today = new Date().toDateString();
if (stats.lastResetDate !== today) {
console.log(`[Scriptalizer] 📅 New day detected, resetting counters`);
console.log(`[Scriptalizer] 📊 Yesterday's stats: ${stats.requestsToday} API calls, ${stats.textsToday} texts processed`);
stats.requestsToday = 0;
stats.textsToday = 0;
stats.lastResetDate = today;
}
}
/**
* Gibt die aktuellen Stats zurück
*/
function getStats() {
checkAndResetDailyStats();
return { ...stats };
}
/**
* Ruft die Scriptalizer API auf
* @param {string} inputText - Text zum Scriptalisieren
* @param {string} fontName - Font-Name (z.B. 'PremiumUltra79')
* @param {number} errFrequency - Fehlerfrequenz
* @returns {Promise<{status: string, outputText: string}>}
*/
async function callScriptalizer(inputText, fontName, errFrequency = 10) {
// Prüfe ob neuer Tag
checkAndResetDailyStats();
return new Promise((resolve, reject) => {
if (!config.scriptalizer.licenseKey) {
reject(new Error("Scriptalizer License Key fehlt in Konfiguration"));
return;
}
// Use x-www-form-urlencoded format
const params = new URLSearchParams();
params.append("LicenseKey", config.scriptalizer.licenseKey);
params.append("FontName", fontName);
params.append("ErrFrequency", String(errFrequency));
params.append("InputText", inputText);
const body = params.toString();
// Check input size (48KB limit)
if (Buffer.byteLength(body) > config.scriptalizer.maxInputSize) {
reject(
new Error(
`Input size exceeds 48KB limit (${Buffer.byteLength(body)} bytes)`
)
);
return;
}
const options = {
hostname: "www.scriptalizer.co.uk",
port: 443,
path: "/QuantumScriptalize.asmx/Scriptalize",
method: "POST",
headers: {
"Content-Type": "application/x-www-form-urlencoded; charset=utf-8",
"Content-Length": Buffer.byteLength(body),
},
timeout: 30000, // 30 second timeout
};
const req = https.request(options, (res) => {
let data = "";
res.on("data", (chunk) => {
data += chunk;
});
res.on("end", async () => {
try {
const result = await parseStringPromise(data, {
explicitArray: false,
});
const response = result.ScriptalizerResponse;
if (!response || response.Status !== "OK") {
reject(new Error(response?.Status || "UNKNOWN_STATUS"));
return;
}
// Zähler erhöhen bei erfolgreichem Request
stats.requestsToday++;
console.log(`[Scriptalizer] 📊 API Request #${stats.requestsToday} today completed`);
resolve({ status: "OK", outputText: response.OutputText });
} catch (err) {
reject(
new Error(`Failed to parse Scriptalizer response: ${err.message}`)
);
}
});
});
req.on("error", (err) => {
reject(new Error(`Scriptalizer request failed: ${err.message}`));
});
req.on("timeout", () => {
req.destroy();
reject(new Error("Scriptalizer request timed out"));
});
req.write(body);
req.end();
});
}
/**
* Scriptalisiert mehrere Texte in 25er Batches
* @param {string[]} texts - Array von Texten
* @param {string} font - Font-Key (tilda, alva, ellie)
* @param {number} errFrequency - Fehlerfrequenz
* @returns {Promise<string[]>} - Array von scriptalisierten Texten
*/
async function scriptalizeBatch(texts, font = "tilda", errFrequency = 10) {
if (!Array.isArray(texts) || texts.length === 0) {
return [];
}
const fontName =
config.scriptalizer.fontMap[font.toLowerCase()] ||
config.scriptalizer.fontMap.tilda;
const separator = config.scriptalizer.separator;
const BATCH_SIZE = 18; // Zusätzlicher Deckel: max. 18 Texte pro Call (unabhängig von der Länge).
// Primäre Grenze ist die ZEICHENLÄNGE: die Scriptalizer-API begrenzt den
// InputText auf 32000 Zeichen pro Call ("Input length limit 32000 exceeded").
// Fixe 18er-Batches sprengen das bei langen (mehrseitigen) Texten – daher nach
// kumulativer Länge batchen, mit Sicherheitsabstand.
const MAX_BATCH_CHARS = 28000;
const sepLen = separator.length;
// Batches nach Länge (und max. Anzahl) bilden.
const batches = [];
{
let cur = [];
let curLen = 0;
for (const t of texts) {
const tLen = String(t || "").length;
const projected = curLen + (cur.length ? sepLen : 0) + tLen;
if (cur.length && (projected > MAX_BATCH_CHARS || cur.length >= BATCH_SIZE)) {
batches.push(cur);
cur = [];
curLen = 0;
}
cur.push(t);
curLen += (cur.length > 1 ? sepLen : 0) + tLen;
}
if (cur.length) batches.push(cur);
}
const totalBatches = batches.length;
console.log(
`[Scriptalizer] Processing ${texts.length} texts in ${totalBatches} batch(es) (max ${BATCH_SIZE} texts / ${MAX_BATCH_CHARS} chars each)`
);
const allScriptalized = [];
for (let b = 0; b < batches.length; b += 1) {
const batch = batches[b];
const batchNumber = b + 1;
const combinedText = batch.join(separator);
console.log(
`[Scriptalizer] Processing batch ${batchNumber}/${totalBatches} (${batch.length} texts, ${combinedText.length} chars)`
);
try {
const result = await callScriptalizer(
combinedText,
fontName,
errFrequency
);
// Trenne Ergebnis wieder
const scriptalized = result.outputText.split(separator);
if (scriptalized.length !== batch.length) {
console.warn(
`[Scriptalizer] Batch ${batchNumber}: returned ${scriptalized.length} parts, expected ${batch.length}`
);
}
allScriptalized.push(...scriptalized);
console.log(
`[Scriptalizer] Batch ${batchNumber}/${totalBatches} completed successfully`
);
} catch (err) {
console.error(`[Scriptalizer] Batch ${batchNumber} failed:`, err.message);
throw err;
}
}
// Textzähler erhöhen
stats.textsToday += allScriptalized.length;
console.log(
`[Scriptalizer] All batches completed. Total scriptalized: ${allScriptalized.length}`
);
console.log(
`[Scriptalizer] 📊 Daily stats: ${stats.requestsToday} API calls, ${stats.textsToday} texts processed`
);
return allScriptalized;
}
/**
* Scriptalisiert einen einzelnen Text
* @param {string} text - Text zum Scriptalisieren
* @param {string} font - Font-Key (tilda, alva, ellie)
* @param {number} errFrequency - Fehlerfrequenz
* @returns {Promise<string>} - Scriptalisierter Text
*/
async function scriptalizeSingle(text, font = "tilda", errFrequency = 10) {
const fontName =
config.scriptalizer.fontMap[font.toLowerCase()] ||
config.scriptalizer.fontMap.tilda;
try {
const result = await callScriptalizer(text, fontName, errFrequency);
return result.outputText;
} catch (err) {
console.error("Scriptalizer single call failed:", err.message);
throw err;
}
}
module.exports = {
scriptalizeBatch,
scriptalizeSingle,
getStats,
};