netanalyzer/utils/chunkReport.js

/**
 * Split a markdown report into Webex-message-sized chunks.
 *
 * Webex caps message bodies at ~7439 chars; we round down to 7000 to leave
 * headroom for the bot framework and any wrapping the client may add.
 *
 * Strategy:
 *   1. If the whole report fits, return it as a single chunk.
 *   2. Otherwise, prefer breaking on section boundaries (`\n\n` immediately
 *      followed by `**`, which is how every section header in our reports is
 *      delimited). This keeps markdown intact across chunks.
 *   3. If a single section is itself larger than the limit, fall back to a
 *      hard slice on the limit so we never lose data.
 */

const DEFAULT_LIMIT = 7000;

function chunkReport(report, maxLen = DEFAULT_LIMIT) {
  if (!report) return [];

  const trimmed = report.trim();
  if (trimmed.length <= maxLen) return [trimmed];

  // Split on `\n\n**` but keep the `**` (consume only the leading `\n\n`).
  const sections = trimmed.split(/\n\n(?=\*\*)/);
  const sectionChunks = [];
  let current = '';

  for (const section of sections) {
    const candidate = current ? `${current}\n\n${section}` : section;
    if (candidate.length > maxLen && current) {
      sectionChunks.push(current);
      current = section;
    } else {
      current = candidate;
    }
  }
  if (current) sectionChunks.push(current);

  // Hard-split any chunk still over the limit (rare — only happens if one
  // section by itself exceeds maxLen, e.g. a store with hundreds of clients).
  const finalChunks = [];
  for (const c of sectionChunks) {
    if (c.length <= maxLen) {
      finalChunks.push(c);
    } else {
      for (let i = 0; i < c.length; i += maxLen) {
        finalChunks.push(c.slice(i, i + maxLen));
      }
    }
  }
  return finalChunks;
}

module.exports = { chunkReport, DEFAULT_LIMIT };