gitoriaLog in with ident

tracker

All repositories: gitoria

ReadmeCodePull requestsReleasesTicketsSettings
Commit96ba683a96ba683adeploy.sh: a gate without a 'passed,' line (check-theme) no longer ends the scriptmre96ba683a/plugins/fs/server.hl

4.5 KB

  1. file(String path) {
  2. return __native("fs.file", path)
  3. }
  4. readFile(String path) {
  5. return __native("fs.read_file", path)
  6. }
  7. \* Read a whole file as Bytes (ticket #88): every byte as it is on disk,
  8. nothing decoded — the way back for what writeFile stored from a Bytes. *\
  9. readBytes(String path) {
  10. return toBytes(__native("fs.read_file", path))
  11. }
  12. listDir(String path) {
  13. return __native("fs.list_dir", path)
  14. }
  15. \* Does this path exist? (mission 077) The one call in hl:fs that cannot fail:
  16. it answers true/false, never an Error event — probing a path is a question,
  17. not an operation. Relative paths resolve against the SCRIPT's directory like
  18. every other hl:fs path. *\
  19. exists(String path) {
  20. return __native("fs.exists", path)
  21. }
  22. \* Watch a directory tree for changes (the dev-reload bell): an ordinary
  23. event-loop source — register it and every save arrives as an event.
  24. __native("eventloop.register", watch("."), "changed")
  25. on changed(ev) { … ev.name … }
  26. `exclude` (2026-08-30) names directories the descent must NOT enter — a
  27. path string, or a list of them. Omit it and the whole tree is watched, as
  28. it always was. The paths are compared RESOLVED (relative ones against the
  29. script's directory, like every hl:fs path) and normalised, so `./sessions`,
  30. `sessions` and `var/../sessions` are the same directory, and a sibling
  31. named `sessions-old` is NOT inside `sessions`.
  32. watch(".", "./sessions")
  33. watch(".", ["./sessions", "/var/cache/app"])
  34. A program that writes to a directory it also watches needs this: without
  35. it the only escape was that the walker skips dot-directories (which it
  36. still does — `.git` is the reason), i.e. renaming the directory. *\
  37. watch(String path, exclude) {
  38. return __native("fs.watch", path, exclude)
  39. }
  40. \* Write a whole file, ATOMICALLY (2026-08-28). Everything above this line
  41. reads; hl:fs could report a file's `chmod` and never set one, which is why
  42. nothing in this language could persist anything until now.
  43. The write goes to a sibling temp file, is fsynced, and is then renamed over
  44. the target — so a concurrent reader sees the whole OLD file or the whole NEW
  45. one, never a half of either. That matters more than it sounds: the first
  46. caller is the session store, and a truncated session file does not read as
  47. an error, it reads as a logged-out user.
  48. `mode` defaults to 0600 — a file this program wrote, for this program. Pass
  49. one to loosen it; nothing here consults the umask on your behalf.
  50. `contents` is a String or a Bytes (ticket #88): a Bytes is written raw,
  51. byte for byte — an uploaded image stays an image. *\
  52. writeFile(
  53. String path
  54. contents \\ a String or a Bytes
  55. Number mode
  56. ) {
  57. if (hlTypeName(contents) == 'Bytes') {
  58. return __native("fs.write_file_hex", path, contents.hex(), mode)
  59. }
  60. return __native("fs.write_file", path, contents, mode)
  61. }
  62. \* Append to a file — the log's shape, not the session store's (2026-09-27, a
  63. creator ask after routger's two). Opened O_APPEND, never truncated: the file
  64. is created (at `mode`, default 0600, same as writeFile) if it is not there
  65. yet, and every call after that lands its bytes after whatever is already on
  66. disk. `writeFile`'s atomicity is a RENAME over a temp file — there is no
  67. "old" or "new" whole file to swap here, since an append MUTATES the one
  68. that is already there — but O_APPEND carries its own, narrower guarantee:
  69. the kernel positions a single write() at the CURRENT end of file and
  70. performs it as one operation, so two processes each appending one line
  71. never interleave their bytes mid-line, and a reader never sees a torn
  72. append. That is what "a single write of a line lands whole" buys — call it
  73. once per line (or per record) you want that guarantee for.
  74. `contents` is a String or a Bytes (ticket #88), same as writeFile: a Bytes
  75. is appended raw, byte for byte. *\
  76. appendFile(
  77. String path
  78. contents \\ a String or a Bytes
  79. Number mode
  80. ) {
  81. if (hlTypeName(contents) == 'Bytes') {
  82. return __native("fs.append_file_hex", path, contents.hex(), mode)
  83. }
  84. return __native("fs.append_file", path, contents, mode)
  85. }
  86. \* Make a directory and every missing parent. Already-there is SUCCESS: the
  87. caller wants the directory to exist, not to be the one who created it.
  88. `mode` defaults to 0700. *\
  89. mkDir(String path, Number mode) {
  90. return __native("fs.mk_dir", path, mode)
  91. }
  92. \* Delete a path. IDEMPOTENT: true when the path is gone on return, and a path
  93. that was already gone counts. A sweep must not fail because something else
  94. deleted the file first. *\
  95. remove(String path) {
  96. return __native("fs.remove", path)
  97. }

Branches

Latest commits

  • 96ba683adeploy.sh: a gate without a 'passed,' line (check-theme) no longer ends the scriptmre
  • eb3b9205tracker: report 031mre
  • 9b5d2e89tracker mission 031: README (What it does, Test: four gates + the #32 checks, Files: theme/, new pages), STATUS (real copy, A/B load, how to repeat, open points), LOGmre
  • 39950e4ctracker#32 (mission 031): the WorldAPI theme (theme/ vendored verbatim from layouts.worldapi.org 85b5654; styles.hl inherits it: accent green-dark, type colours 1-6; own base/header rules, row lines, genre-pill and inverted-button frames removed, the season foldable keeps its line; check-theme 21 -> 0, 4th deploy gate; main actions class primary) and the #32 header (theme AppHeader/MainMenu/UserMenu/Sidebar/ContentFirst: desktop brand, search, Series|Shows|Movies|Genres|People, user icon with Unwatched..Settings, Logout; signed out the ident selector, phone the iD icon dropdown; phone menu in the sidebar overlay; marked entry by :has); /find -> /search/<q>, /genres, /people(/<letter>), /settings; main { ContentFirst { slot } } works around the hl:web one-line slot bug; gates 365/0, 32/0, 52/0, check-theme 0mre
  • a386dc92tracker: reports 029 + 030mre
  • 71e0fd7dtracker missions 029 + 030: README (What it does, Files, gate count), STATUS (real-copy numbers, how to repeat, open points), LOGmre
  • d36ea6eatracker#34 + #35 (mission 030): Follow directly under the poster, as wide as the poster (show.hl, styles.hl); the status pill next to a series' title — TVmaze's status (new tvmazeStatus, stored by the sync's TVmaze merge) else TMDB's, TVmaze Ended + TMDB Canceled = Canceled, inverted (filled, dark text, no border), green running / yellow pending / red canceled / muted ended (shows.hl statusOf); the daily delta asks TVmaze's status of an unfollowed series TVmaze's change list names (dailysync.hl syncRunStep, sync.hl syncTvmazeStatus); the status backfill after the details repair (backfill.hl, jobs.hl statusTick; resumable, 550 ms per TVmaze request); gates 354/0, 32/0, 52/0mre
  • 7d7d4487tracker#33 (mission 029): reduced titles — every title TMDB's details never went through this app (no detailsAt, no tmdbSync) is incomplete (shows.hl isIncomplete; the old tracker's migrated rows passed #26's test: 5,697 non-adult on the live copy, 691 series without seasons); the repair job does the visibly reduced first (shows.hl missingParts), the page completes one on open; a title TMDB has no poster for (The Remaining) shows the placeholder; tools/count-incomplete.hl; gate fixtures stand for synced titles (tmdbSync), tests/seed-reduced.hl + #33 checks; gates 347/0, 32/0, 52/0mre
  • 661c2592tracker: report 028mre
  • 27c916fatracker mission 028: README ("Code order", the new file map), STATUS (counts before/after, tests, how to repeat, open), LOGmre
  • d924f398tracker mission 028: comments name the new files (sync.hl, dailysync.hl, backfill.hl, credits.hl, jobs.hl, images.hl …); tools/ref-params.py + tools/lambda-audit.py also scan lib/ (they globbed the root only), lambda-audit counts a plain `x = p` alias like `let x = p`mre
  • 2e89b968tracker mission 028 (code order) 5/5 let: `let` only where a variable is reassigned — 667 never-reassigned lets became plain declarations (project.hl, lib/, components/, tools/, tests/); kept: 264 in loop bodies (a plain declaration there is 'Cannot reassign' on the 2nd pass), 234 reassigned, 27 whose name is also a member/outer/free name (a plain write would rebind it); tools/let-audit.py decides and fixes (README 'Code order'); tests/realdata-m028.{sh,mjs} = the page-output diff on a real copy; gates 342/0, 32/0, 52/0, real-copy pages identicalmre
  • 54796ff2tracker mission 028 (code order) 4/5 thin faces + last copies: the show page's check/follow faces call lib/watches.hl toggleWatched / toggleSeasonWatched (seasonAllWatched moved there) and lib/follows.hl toggleFollowed; both logins (header selector face, /login/callback) share lib/users.hl userOfCode; todayStr/listOf copies in components and the export readers copied into tools/migrate.hl + tools/old-short-ids.hl now once (lib/util.hl, lib/export.hl); gates 342/0, 32/0, 52/0; old-short-ids output byte-identical, migrate output identicalmre
  • 06b078e3tracker mission 028 (code order) 3/5 project.hl is the map: config, routes, wiring and a feature → file index (914 → 258 lines); the background jobs (daily sync run, backfills, details repair, credits job, merge, short ids, collection seed) moved unchanged into lib/jobs.hl (a class: their state is reassigned every step, a static cannot be; one instance made after the server), the login callback into lib/users.hl, poster/photo serving into lib/images.hl, the /shows/<slug> rule into lib/shows.hl showsMovedPath; route handlers are thin wrappers; gates 342/0, 32/0, 52/0, real-copy pages identicalmre
  • 94716fd2tracker mission 028 (code order) 2/5 util + topics: lib/util.hl holds envOr, storageDir, postersDir, profilesDir, newId, hexDigits, todayStr, dateOr, textOr, hasId, listOr, firstOf, sortDesc once (were copied into up to 5 files); tmdbsync.hl split into tmdb.hl (TMDB/TVmaze requests), sync.hl (one title's sync), sync-helpers.hl, backfill.hl; details.hl split into details.hl, credits.hl, credits-helpers.hl (isIncomplete to shows.hl); search-helpers.hl (words, query, ranking, slugs); collections.hl (the TMDB collection seed, out of franchises.hl); deltasync.hl renamed dailysync.hl; no behaviour change: gates 342/0, 32/0, 52/0, real-copy pages identicalmre
  • 186079b0tracker mission 028 (code order) 1/5 move: every root .hl except project.hl into lib/ (styles.hl into components/), import paths only; gates 342/0, 32/0, 52/0; real-copy pages identicalmre
  • 4f47f181tracker: report 027mre
  • dc1d4be4tracker mission 027: Hybriel master 06617221 vendored (plugin allocator fixes 3a781359 + 413f60e4); real copy RSS through first-start jobs + 400 loads flat ~2.55 GB (190aa11d 2.3 -> 5.6 GB), page times <= 1.1x; gates 342/0, 32/0, 52/0mre
  • 84e1b3e1tracker: reports 025 + 026mre
  • dc40d859tracker#31 (mission 026): duplicate titles merged — the 68 type+tmdbId pairs held by 157 records were the old tracker's (all migrated); merge.hl repair job (own clock, before the TMDB jobs) keeps one keeper per title (follows/watches > old short id > oldest), moves follows, watches, seasons, cast, credits, timelines, tombstones the rest (mergedInto, never deleted), slugs + short ids 301 to the keeper; stray seasons merged into their listed twin (Reacher S3 watches) or linked when watched; search import re-checks before its put; deploy.sh waits up to 90 s for 200; real copy 68 -> 0 dup ids, az5b2 follows/watches equal; gates 342/0, 32/0, 52/0mre