ndc-content-tools/import-transcript.js

const fs = require('fs')

function importTranscript(srtContent, names) {
  const entries = srtContent.trim().split(/\n\n+/)
  const nameSet = new Set(names.map(name => name.trim()))

  let foundSpeakers = new Set()

  const mdEntries = entries.map(entry => {
    const lines = entry.split(/\n/)
    if (lines.length < 3) return null

    const timeParts = lines[1].split(' --> ')
    if (timeParts.length !== 2) return null

    let content = lines.slice(2).join(' ').trim()
    for (const name of nameSet) {
      const regex = new RegExp(`^${name}:`, 'i')
      if (content.match(regex)) {
        content = content.replace(name, `**${name}**`)
        foundSpeakers.add(name)
        break
      }
    }

      return `==${timeParts[0]}==${content}==${timeParts[1]}==`
  }).filter(Boolean)

  nameSet.forEach(name => {
    if (!foundSpeakers.has(name)) {
      console.warn(`Warning: Speaker ${name} wasn't found. Did you misspell their name?`)
    }
  })

  return mdEntries.join('\n\n')
}

let srtFileName, mdFileName, names = ''

for (let i = 2; i < process.argv.length; i++) {
  switch (process.argv[i]) {
    case '--input':
    case '-i':
      srtFileName = process.argv[++i]
      break
    case '--output':
    case '-o':
      mdFileName = process.argv[++i]
      break
    case '--speakers':
      names = process.argv[++i]
      break
  }
}

if (!srtFileName) {
  console.log("This utility converts valid .srt files to NDC compatible transcripts. Usage: node import-transcripts.js --input <input.srt> --output <output.md> --speakers 'Name1,Name2,...'")
  process.exit(1)
}

const srtContent = fs.readFileSync(srtFileName, 'utf8')
const nameList = names.split(',')
const mdContent = importTranscript(srtContent, nameList)

if (mdFileName) {
  fs.writeFileSync(mdFileName, mdContent, 'utf8')
  console.log(`Converted content written to ${mdFileName}`)
} else {
  console.log(mdContent)
}