Indlela yokuthola umbhalo wokuloba we-YouTube nge-API (curl, Python, JavaScript)

Ukuze uthole umbhalo wokuloba (transcript) we-YouTube nge-API, thumela isicelo se-HTTP esinale-URL noma i-ID yevidiyo kanye nokhiye we-API, bese ufunda umbhalo kuphendulo. Lo mhlahlandlela ukhombisa isicelo ku-curl, Python ne-JavaScript usebenzisa i-API yemibhalo ye-sum.video, bese uchaza amafomethi, amaphutha nezindleko.

Yini ebuyiswa yi-API

I-endpoint ibuyisa izihloko ezikhona zevidiyo yangaphandle ye-YouTube, noma ngabe ezibhalwe umdali noma ezakhiwe ngokuzenzakalelayo yi-YouTube. Ayiguquli umsindo ube umbhalo. Uma ividiyo ingenazo izihloko, uthola iphutha esikhundleni sokuqagela (okwengeziwe ngezansi).

Izisekelo:

• I-endpoint: POST https://sum.video/api/v1/transcript (GET nayo iyasebenza)
• Ukuqinisekisa: ukhiye wakho ku-header ethi X-API-Key, noma Authorization: Bearer <key>
• Okufakwayo: url noma video_id
• lang ongayikhetha (okuzenzakalelayo en)
• format ongayikhetha: json, text noma srt

Impendulo ikhomba ulimi lweqiniso lomzila wezihloko nokuthi ngabe wenziwe ngokuzenzakalelayo yini, ukuze wazi ikhwalithi ongayilindela.

Isinyathelo 1: dala ukhiye we-API

Ngena bese udala ukhiye ku-dashboard ye-API (/dashboard/api ku-sum.video). Ungaba nokhiye abangafika kwabangu-5 abasebenzayo, okusiza ukwehlukanisa ukhiye wokuhlola nokhiye wokukhiqiza. Ukhiye wakho wokuqala uza nezikweletu eziyi-10 zamahhala, ezanele imibhalo emi-5, ngakho ungazama konke okungezansi ngaphandle kwezindleko.

Isinyathelo 2: yenza isicelo sakho sokuqala

curl

curl -X POST https://sum.video/api/v1/transcript \
  -H "X-API-Key: $SUMVIDEO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ", "lang": "en", "format": "text"}'

Njengoba ne-GET isebenza, ungahlola ngisho nasebhalweni olwakha i-URL kuphela. Kungcono ukuthumela ukhiye ku-header ukuze ungapheli emarekhodini (logs).

Python

import os
import requests

resp = requests.post(
    "https://sum.video/api/v1/transcript",
    headers={"X-API-Key": os.environ["SUMVIDEO_API_KEY"]},
    json={"video_id": "dQw4w9WgXcQ", "lang": "en", "format": "text"},
    timeout=30,
)

if resp.status_code == 404:
    print("No captions for this video")
else:
    resp.raise_for_status()
    print(resp.text)

JavaScript (Node 18+ noma isiphequluli)

const resp = await fetch("https://sum.video/api/v1/transcript", {
  method: "POST",
  headers: {
    "X-API-Key": process.env.SUMVIDEO_API_KEY,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    video_id: "dQw4w9WgXcQ",
    lang: "en",
    format: "text",
  }),
});

if (resp.status === 404) {
  console.log("No captions for this video");
} else if (!resp.ok) {
  throw new Error(`Request failed: ${resp.status}`);
} else {
  console.log(await resp.text());
}

Gcina ukhiye kuseva. Ukhiye ekhodini lesiphequluli ungakopishwa yilowo nalowo ovula amathuluzi onjiniyela.

Isinyathelo 3: khetha ifomethi efanele

Uma uhlela ukuxhumanisa emzuzwini othile wevidiyo, sebenzisa i-json bese wakha isixhumanisi usuka ku-offset yengxenye. Ku-LLM, i-text imfushane futhi ilula. Uma kamuva ufuna ukusesha ngesandla ngaphakathi kombhalo, bheka indlela yokusesha embhalweni wevidiyo ye-YouTube.

Isinyathelo 4: khetha ulimi

I-lang ngokuzenzakalelayo ithi en. Setha ulimi olufunayo, njengo-de noma es. Impendulo ikhomba ulimi lomzila obuyiswe ngempela, ngakho yihlole esikhundleni sokucabanga. Ividiyo ivame ukuba nomzila owodwa kuphela wezihloko, futhi i-API ibuyisa izihloko zangempela, ngakho ayihumushi ngawe.

Isinyathelo 5: phatha amaphutha

Hlela lezi zimo kusukela osukwini lokuqala:

• 404 no_transcript: ividiyo ayinazihloko, noma ingukusakaza bukhoma, ividiyo eyimfihlo noma esusiwe. Akukho okufanele kuzanywe futhi. Yeqe, noma ucele umsebenzisi enye ividiyo.
• 402 budget_exceeded: ufinyelele umkhawulo wezikweletu wenyanga owusethile. Bheka isigaba esilandelayo.
• Umkhawulo wesivinini: ukhiye ngamunye uvumela izicelo ezingu-60 ngomzuzu. Hlakaza umsebenzi wamaqoqo bese ucuphela phansi lapho ukhawulelwe.
• Ukwehluleka ohlangothini lwethu: zama futhi ngemuva kwesikhashana. Akunazindleko.

Isiyingi esilula esizama futhi emaphutheni eseva futhi siyeqa ku-404 sivala iningi lezidingo zokukhiqiza.

Kubiza malini

Umbhalo ngamunye ubiza izikweletu ezi-2. Izicelo ezingabuyisi mbhalo noma ezehlulekayo ohlangothini lwe-sum.video azinazindleko, ngakho i-404 ayikaze ikhiphe ibhalansi yakho. Ibhalansi ungayibona nge-GET /api/v1/credits.

Izikweletu zithengiswa ngamaphakheji okuthenga kanye angapheli:

Abasebenzisi be-Pro bathola nezikweletu eziyi-100 zebhonasi njalo ngenyanga. Emisebenzini esindayo, njengokungenisa isiteshi sonke, funda isimo imibhalo ye-YouTube yohlelo lokusebenza lwe-RAG ngezinga elikhulu esihlanganisa ukugcwalisa okuzenzakalelayo nezabelomali. Bheka futhi izintengo.

Imikhawulo okufanele uyazi ngaphambi kokwakha

• Amavidiyo anezihloko kuphela asebenza. Leso yisizathu esikhulu se-404 no_transcript.
• Izihloko ezizenzakalelayo zingaba namaphutha, ikakhulukazi emagameni, ezindlinganisweni zokukhuluma nasemagameni obuchwepheshe.
• I-API ibuyisa umbhalo kuphela. Ayakhi izifinyezo. Uma ufuna isifinyezo, izimpawu zesikhathi nengxoxo ngombhalo, isifinyezi se-YouTube siyakwenza lokho, futhi ithuluzi lemibhalo liyinguqulo engenakhodi yevidiyo eyodwa.
• I-YouTube Data API esemthethweni ayiyona indawo yokufaka esikhundleni lapha. Indlela yayo yokulanda izihloko idinga imvume yokuhlela ividiyo, ngakho ingeyamavidiyo akho (amadokhumenti e-Google).

Imithombo

• YouTube Data API: Captions.download
• I-API yemibhalo ye-sum.video