gitoriaLog in with ident

calendar

All repositories: gitoria

ReadmeCodePull requestsReleasesTicketsSettings
Main branchmain2881119fcalendar mission 001 (4/4): code order — README layout + same-output test, STATUS, LOG, report; tests/letcount.py (sub-folders), tests/realdata-baseline.mjs, tests/realdata-compare.pymremain/tests/ports.mjs

2.6 KB

  1. // tests/browser/ports.mjs — the two port questions the CDP driver has to ask,
  2. // and the reasons they are asked this way. They lived in the hl:web browser
  3. // tier's `fixtures.mjs` until that tier moved to `archive/hlweb-v1/`
  4. // (2026-09-10); the driver survives it, so its dependencies do too.
  5. //
  6. // Every process a caller starts is its own to kill BY PID — never `pkill -f`,
  7. // which on this machine matches sibling sessions' runners.
  8. //
  9. // `portOpen` (a plain connect) is NOT here: the driver keeps its own, because
  10. // "something answers" is a different question from "this port is free", and
  11. // only the second one needs the reasons written down.
  12. import { createServer as netServer } from 'node:net';
  13. import { execFileSync } from 'node:child_process';
  14. /**
  15. * A BIND IS THE QUESTION, SO A BIND IS THE TEST (mission 285). A listener on a
  16. * LAN address is invisible to a loopback connect and still takes the port from
  17. * a server that binds 0.0.0.0 — which is how a browser-tier red once looked
  18. * like a hung server: the app's own log said `bind() failed on port N` while
  19. * every "is it open?" probe said no.
  20. *
  21. * Nothing is left behind: a listening socket with no accepted connections
  22. * leaves no TIME_WAIT when it closes.
  23. */
  24. export function portBindable(port) {
  25. return bindProbe(port).then((r) => r.ok);
  26. }
  27. /**
  28. * The same probe with the kernel's own word for WHY. `EADDRINUSE` is a
  29. * collision and is what this file is about; anything else (EACCES, EMFILE …)
  30. * is a different problem and must not be reported as one, because an error
  31. * that names the wrong cause costs the next reader a whole night.
  32. */
  33. function bindProbe(port) {
  34. return new Promise((res) => {
  35. const s = netServer();
  36. s.once('error', (e) => res({ ok: false, code: (e && e.code) || String(e) }));
  37. s.listen(port, '0.0.0.0', () => s.close(() => res({ ok: true, code: null })));
  38. });
  39. }
  40. /**
  41. * Every socket on `port`, in any state, with its owner — `ss -tanp`. LISTEN is
  42. * the state that actually blocks a bind (measured: TIME_WAIT and ESTABLISHED do
  43. * not, on loopback or on the LAN address), so a report showing only listeners
  44. * and finding none says "nothing", which is the least useful thing a diagnostic
  45. * can say. Show the whole port.
  46. */
  47. export function whoHolds(port) {
  48. try {
  49. const out = execFileSync('ss', ['-tanp'], { encoding: 'utf-8' });
  50. const hits = out.split('\n').filter((l) => new RegExp(`:${port}(\\s|$)`).test(l));
  51. return hits.length
  52. ? hits.map((l) => l.trim().replace(/\s+/g, ' ')).join('\n ')
  53. : '(no socket on that port at all by the time `ss` ran — the holder let go between the failed bind and this line)';
  54. } catch { return '(`ss` is not runnable here)'; }
  55. }

Branches

Latest commits

  • 2881119fcalendar mission 001 (4/4): code order — README layout + same-output test, STATUS, LOG, report; tests/letcount.py (sub-folders), tests/realdata-baseline.mjs, tests/realdata-compare.pymre
  • 3b2a4cd0calendar mission 001 (3/4): code order — let only where reassigned (147 dropped incl. components/month-view; 110 left: 67 reassigned, 43 loop-bound); gates 78/0 (3 of 4 runs; 1 known 'next month' flake) + 18/0, live-data run = step 2mre
  • f5ce6b1dcalendar mission 001 (2/4): code order — topics, map, thin faces: lib/util.hl, lib/events(-helpers).hl (+calendarView/settingsView/soonOf), lib/settings.hl, lib/users.hl (+userOfLoginCode, tagOf), lib/api(-helpers).hl; project.hl = map; login route takes &req/&sessions (failed-login reason now kept); gates 78/0 + 18/0mre
  • c6fbd011calendar mission 001 (1/4): code order — files moved: lib/events.hl, lib/users.hl, components/styles.hl (imports only); gates 78/0 + 18/0, live-data run identicalmre
  • d6c59290calendar: Hybriel master 06617221 (plugin allocators 3a781359 + 413f60e4, mpackdb 2cb7ae5e, http1 773de63e); gates 78/0 + 18/0mre
  • 7bd0337ccalendar: Hybriel master 190aa11d (fc838894 GC correctness, #126 closure scopes, #127); gates 78/0 + 18/0mre
  • ff41310ccalendar: Hybriel master 8efba065 (#126 memory, #48 lambda copy; audit: no & needed)mre
  • 14ba08c7antcolony#40: mission references point to the moved missionsmre
  • 99c73346antcolony#40: history (LOG.md), worker briefs (missions/) and reports moved here from antcolony, numbered per project; old numbers in antcolony docs/mission-map.mdmre
  • 76edaa62calendar: Hybriel master ff51cf46 (re-vendor round, static workaround removed)mre
  • 90a3fc2cdeploy.sh: back up live storage/.sessions/.env before every deploy (newest 5 kept)mre
  • 6722b72ddeploy.sh: never send .git or .gitignore to Byrodinmre
  • be099807State of 2026-09-27, before the move to gitoriamre