mirror of
https://gitlab.rlp.net/proj-wise2526-video2document/video2document.git
synced 2026-06-15 18:01:52 +02:00
122 lines
4.3 KiB
JavaScript
122 lines
4.3 KiB
JavaScript
const fs = require("fs");
|
|
const path = require("path");
|
|
|
|
// Prepare output directory (always storage/transcriptionSummaries under project root)
|
|
const outputDir = `${__dirname}/../../../storage/transcriptionSummaries`;
|
|
if (!fs.existsSync(outputDir)) {
|
|
fs.mkdirSync(outputDir, { recursive: true });
|
|
}
|
|
|
|
//Speaker, Sentence, Start, End
|
|
|
|
module.exports = {
|
|
name: "summarize-transcription2", // Unique name for our function that will later be used to get the function from the map via "mapFunctions.get("example").function()"
|
|
type: "summarizer", // value used to differentiate each module to order them in the UI
|
|
displayname: "Summarizer", // The displayname used within the UI
|
|
async function(args) {
|
|
let inputJson = args.json;
|
|
|
|
//JSON Path
|
|
if (args.jsonPath) {
|
|
try {
|
|
const raw = fs.readFileSync(args.jsonPath, "utf-8");
|
|
inputJson = JSON.parse(raw);
|
|
} catch (e) {
|
|
console.error("Failed to load JSON from file:", e);
|
|
return { error: "Could not read JSON from file path." };
|
|
}
|
|
}
|
|
// JSON parsen
|
|
if (typeof inputJson === "string") {
|
|
try {
|
|
inputJson = JSON.parse(inputJson);
|
|
} catch (e) {
|
|
console.log("Invalid JSON in summarize-transcription");
|
|
return { error: "Invalid JSON" };
|
|
}
|
|
}
|
|
|
|
const words = inputJson.words;
|
|
if (!Array.isArray(words)) {
|
|
return { error: "No words Array found" };
|
|
}
|
|
|
|
const ENDINGS = [".", "!", "?"]; // '...' auch als Satzende ?
|
|
const ABBREVIATIONS = new Set(["z.B.", "bzw.", "u.a.", "Dr.", "Mr.", "Mrs.", "Prof.", "etc."]); //TODO weitere Ergaenzen
|
|
|
|
const result = [];
|
|
let currentSentence = "";
|
|
let currentSpeaker = null;
|
|
let startTime = null;
|
|
let endTime = null;
|
|
|
|
for (const w of words) {
|
|
if (!currentSpeaker) currentSpeaker = w.speaker;
|
|
if (startTime === null) startTime = w.start;
|
|
endTime = w.end;
|
|
|
|
//speaker changing
|
|
if (currentSpeaker !== w.speaker && currentSentence) {
|
|
result.push({
|
|
speaker: currentSpeaker,
|
|
sentence: currentSentence,
|
|
start: startTime,
|
|
end: endTime
|
|
});
|
|
currentSentence = "";
|
|
startTime = w.start;
|
|
}
|
|
currentSpeaker = w.speaker;
|
|
currentSentence += (currentSentence ? " " : "") + w.text; //sentence beginning or not
|
|
const lastWord = w.text.trim();
|
|
const lastChar = lastWord.slice(-1);
|
|
const isAbbreviation = ABBREVIATIONS.has(lastWord);
|
|
|
|
//sentence ending
|
|
if (ENDINGS.includes(lastChar) && !isAbbreviation) {
|
|
result.push({
|
|
speaker: currentSpeaker,
|
|
sentence: currentSentence,
|
|
start: startTime,
|
|
end: endTime
|
|
});
|
|
currentSentence = "";
|
|
startTime = null;
|
|
endTime = null;
|
|
currentSpeaker = null;
|
|
}
|
|
}
|
|
|
|
// safe last sentence
|
|
if (currentSentence) {
|
|
result.push({
|
|
speaker: currentSpeaker,
|
|
sentence: currentSentence,
|
|
start: startTime,
|
|
end: endTime
|
|
});
|
|
}
|
|
|
|
|
|
// Output as Text
|
|
const output = result.map(r =>
|
|
`Sprecher ${r.speaker} [${r.start.toFixed(2)} - ${r.end.toFixed(2)}]: ${r.sentence}`
|
|
);
|
|
|
|
// Output on cosole
|
|
//console.log("\n------------\nMerged Transcription Result:\n", output, "\n------------\n");
|
|
|
|
try {
|
|
const jsonPath = path.join(outputDir, "transcription_result.json");
|
|
fs.writeFileSync(jsonPath, JSON.stringify(result, null, 2), "utf-8");
|
|
|
|
const txtPath = path.join(outputDir, "transcription_result.txt");
|
|
fs.writeFileSync(txtPath, output.join("\n"), "utf-8");
|
|
|
|
console.log(`Summary successfully saved:\n- ${jsonPath}\n- ${txtPath}`);
|
|
} catch (err) {
|
|
console.error("Error saving Summary:", err);
|
|
}
|
|
}
|
|
}
|