fix(blur): .blurred round-trips any rel, and one writer serves both surfaces

The heid bug-hunt on r2b merge 1 found the /blur route stripping `f` before
writing, so the form for " a.png" blurred its neighbour "a.png". The route was
only half of it: `.blurred` was one stripped rel per line, so no writer could
store a rel with a leading space or a newline, whatever the route did.
Operator-ruled 2026-09-23 ("fix the blur").

- booth/blur.py (new, stdlib-only): read_blurred / set_blurred / BLUR_FILE.
  `.blurred` is now a JSON array in sorted order, the `.seen` shape: opened
  O_NOFOLLOW | O_NONBLOCK with an S_ISREG check and a 1 MiB cap, so a planted
  symlink is refused and a FIFO can no longer hang every Desk render (the old
  read_text() blocked on one). Writes go through mkstemp + os.replace. The
  legacy line format is still READ, so the 6 live line-format files keep their
  blur until their next write upgrades them. Measured before the change: 42
  live rels, none with edge whitespace, so the defect had no live victims.
- The route no longer strips `f`.
- scripts/booth `blur`/`unblur` go through booth.blur.set_blurred instead of
  their own grep/printf line writer. Two writers of one format is how the
  formats drift, and after this change the shell writer would have appended a
  line to a JSON array. Every path is checked before anything is written.
- Item.blurred_self (appended to the record): the item's own blur, resolved in
  booth_items from the same read as `blurred`. It replaces build_gallery's
  second read_blurred, which a write between the two reads could split
  (invariant 3). app.py no longer reads blur state at all, and a test asserts
  it.

Names stay importable from booth.app and booth.items (invariant 4). blur joins
test_stdlib_only. test_cli's per-item-survives test now reads through the reader
rather than asserting the old byte format. The r2b contract and its mutation
row follow blurred_self onto the record. tests/mutations/blur_storage.toml
proves 12 falsifiers by running the change each forbids.

Not in this change, and still ours: the "off"-means-ON idiom drift between
/blur, /blurbooth and /flag (forms only ever send 0/1), and the CLI's
`.blurbooth` touch following a symlink where the service no longer does.
This commit is contained in:
vh
2026-09-23 22:05:18 -07:00
parent cce6a20abe
commit 4cfbce5109
12 changed files with 580 additions and 89 deletions
+31 -20
View File
@@ -39,7 +39,7 @@ lags the code defeats its own purpose.
These are the ones a casual change breaks silently. Each has a test.
### 1. `links.py`, `asks.py` and `marks.py` are stdlib-only, on purpose
### 1. The modules `scripts/booth` imports are stdlib-only, on purpose
`scripts/booth` — the CLI every fleet session uses — imports them directly:
@@ -48,39 +48,50 @@ BOOTH_SRC=… python3 -c 'import sys; sys.path.insert(0, …); from booth.marks
```
It runs under the system `python3` with **no venv**. A single third-party
import in any of the three breaks `booth ask` / `booth marks` / `booth answer` /
`booth unlink` on every host, and the failure surfaces in an agent's session,
not in ours.
import in any of them breaks `booth ask` / `booth marks` / `booth answer` /
`booth unlink` / `booth blur` on every host, and the failure surfaces in an
agent's session, not in ours.
`items.py` and `app.py` are free to import what they like. Those three are not.
`test_stdlib_only` walks each module's AST imports and asserts it — the CLI
imports through a `python3 -c` heredoc that no AST extractor can see, so that
test is the only thing standing here.
The set is `marks`, `asks`, `links`, `manifest`, `benches`, `blur` and
`__init__` (which runs before every one of them). **The list of record is
`test_stdlib_only`'s parametrize in `tests/test_marks.py`**, not this
paragraph. `items.py` and `app.py` are free to import what they like; those are
not. `test_stdlib_only` walks each module's AST imports and asserts it — the
CLI imports through a `python3 -c` heredoc that no AST extractor can see, so
that test is the only thing standing here. A new module the CLI imports goes on
that list in the same commit.
### 2. The filesystem is the state
No database. `ls ~/booth-data` tells you everything the service knows.
Per-booth operator state is a **dotfile inside the booth**: `.forever` (keep),
`.viewed` (last deliberate look — U4's "viewing is activity"), `.blurred` (one
rel per line — ⚠ see below), `.seen` (R2: rels looked at full size, a JSON
ARRAY), `.blurbooth` (the whole booth fogged — a MARKER like `.forever`, not
`.viewed` (last deliberate look — U4's "viewing is activity"), `.blurred` (the
per-item blur set, a JSON ARRAY — see below), `.seen` (R2: rels looked at full
size, a JSON ARRAY), `.blurbooth` (the whole booth fogged — a MARKER like `.forever`, not
JSON, because a boolean has no rels to round-trip), `.marks.json` + `.marks.lock` (judgment), `.pins` (link-board pin
ids), `.uploaded` (upload-booth marker). `booth_items()` skips `name.startswith(".")`, so a new
dotfile costs nothing in item counts, galleries or zips. That skip is why the
dotfile is the right shape for new operator state — use it rather than
inventing a sidecar-per-item.
⚠ **`.seen` is a JSON array where `.blurred` is one stripped rel per line, and
the difference is a latent bug in the older one.** A rel may carry a leading
space or a newline; line-stripped storage does not round-trip it, so blurring
`" a.png"` can blur `a.png` instead. `.seen` was written as JSON for exactly
that reason (design-dev, R2), and it also opens `O_NOFOLLOW | O_NONBLOCK` with
an `S_ISREG` check — a planted symlink is refused and a FIFO cannot hang the
read, which is the outage in
⚠ **A dotfile that holds rels is a JSON array, opened `O_NOFOLLOW |
O_NONBLOCK` with an `S_ISREG` check and a size cap.** A rel may carry a leading
space or a newline, and line-stripped storage does not round-trip it: `.blurred`
was one stripped rel per line, and blurring `" a.png"` blurred `a.png` instead.
`.seen` was written as JSON for exactly that reason (design-dev, R2), and
`.blurred` now matches it (`booth/blur.py`). The open flags mean a planted
symlink is refused and a FIFO cannot hang the read, which is the outage in
`persistent-memory.d/2026-09-22-size-cap-opened-a-hang.md`. **Any new dotfile
inherits that shape, not `.blurred`'s.** `.blurred` itself is unfixed and
pre-existing.
inherits that shape.** `.blurred`'s reader still accepts the old line format,
so a booth written before the change keeps its blur until its next write
upgrades the file. Do not remove that fallback while a line-format file can
still exist.
**A dotfile with two writers has ONE implementation of the writer.** `.blurred`
is written by the service and by `booth blur`, and both call
`booth.blur.set_blurred`; the CLI used to keep a grep/printf writer of its own,
and two writers of one format is how the formats drift apart.
### 3. One resolver for item facts
+10 -25
View File
@@ -102,6 +102,9 @@ from booth.items import ( # noqa: E402,F401
render_doc,
render_doc_body,
)
# The per-item blur writer lives with its reader in booth/blur.py, stdlib-only so
# `scripts/booth blur` shares both. Re-exported: tests import it from here.
from booth.blur import set_blurred # noqa: E402,F401
# Sentinel dotfile that exempts a booth from the TTL sweep. A dotfile because
# the existing listing code already skips dotfiles, so it costs nothing in item
@@ -163,24 +166,6 @@ def set_booth_blurred(booth: Path, on: bool) -> bool:
return False
def set_blurred(booth: Path, rel: str, on: bool) -> set[str]:
"""Add or remove one item from the blur set. Atomic replace, so a crash
mid-write cannot leave a half-file that read_blurred would parse as a
shorter — and therefore more revealing — set. Returns the new set."""
current = read_blurred(booth)
if on:
current.add(rel)
else:
current.discard(rel)
path = booth / BLUR_FILE
if not current:
path.unlink(missing_ok=True)
return current
tmp = path.with_suffix(".tmp")
tmp.write_text("".join(f"{r}\n" for r in sorted(current)))
tmp.replace(path)
return current
# The link-board logic lives in booth/links.py (stdlib only) so the `booth` CLI
# can use it without pulling FastAPI in. Re-exported here because call sites and
# tests already reference these names through app.
@@ -818,10 +803,6 @@ def build_gallery(child: Path) -> list[dict]:
suite reaches for it by name in nine places.
"""
out = []
# r2b D2b: the item's OWN blur, apart from the booth's. `blurred` is the
# composed fact the surfaces render; the per-item control changes only
# this, and must not claim an un-blur the booth flag would override.
own_blur = read_blurred(child)
for it in booth_items(child):
body = render_doc_body(child, it)
rendered, rendered_html = body if body is not None else (None, False)
@@ -844,7 +825,9 @@ def build_gallery(child: Path) -> list[dict]:
"rendered": rendered,
"rendered_html": rendered_html,
"blurred": it.blurred,
"blurred_self": it.rel in own_blur,
# r2b D2b: the item's OWN blur, apart from the booth's — off the
# record, from the one read `blurred` came from (invariant 3).
"blurred_self": it.blurred_self,
}
)
return out
@@ -2234,8 +2217,10 @@ def create_app(
See BLUR_FILE: this hides an item from a glance, it does not protect it."""
booth = resolve_booth(name)
# Guard the path the same way the file route must: a blur entry is only
# ever a booth-relative path, never an escape.
rel = f.strip().lstrip("/")
# ever a booth-relative path, never an escape. NEVER STRIPPED: " a.png"
# and "a.png" are two items, and a stripped `f` blurred the neighbour
# (heid bug-hunt on r2b merge 1). A leading "/" is never part of a rel.
rel = f.lstrip("/")
if ".." in Path(rel).parts:
raise HTTPException(status_code=400, detail="bad item path")
set_blurred(booth, rel, on not in ("0", "false", ""))
+118
View File
@@ -0,0 +1,118 @@
"""Per-item blur storage — `.blurred`, one JSON array of booth-relative paths.
⚠ STDLIB ONLY (CLAUDE.md invariant 1). `scripts/booth blur` imports this under
the system python3 with no venv, so the service and the CLI share ONE reader and
ONE writer. The CLI used to keep its own grep/printf line writer, and two
writers of one file is how the formats would drift apart.
⚠ COSMETIC ONLY. A blurred item is still served, still in the zip, still on
disk. The Booth has no auth: if a thing must not be SEEN, it must not be in a
booth.
The `.seen` shape, for `.seen`'s reason: a rel may carry a leading space or a
newline, and the old one-stripped-rel-per-line format could not round-trip it.
Blurring " a.png" stored "a.png" and blurred the neighbour instead. The line
format is still READ, so a booth written before this change keeps its blur
until its next write upgrades the file.
"""
from __future__ import annotations
import json
import os
import stat
import tempfile
from pathlib import Path
BLUR_FILE = ".blurred"
# A blur set bigger than this is not one this service or the CLI wrote: a JSON
# array of every rel in a 270-item booth is a few KB. Same bound as `.seen`.
BLUR_MAX_BYTES = 1 << 20
def read_blurred(booth: Path) -> set[str]:
"""Blurred rels for a booth. Missing, unreadable or malformed -> empty set.
NEVER RAISES and NEVER BLOCKS. `booth_items` calls this for every booth the
Desk renders, and any fleet session can write into a booth, so the file may
be planted: it is opened without following a link and without blocking (a
FIFO with no writer), and refused unless it is a regular file of sane size.
Reads BOTH formats. A JSON array of strings is the current one; anything
that does not parse as a JSON array is the legacy one-rel-per-line format,
read exactly as before (stripped, blank lines dropped). Legacy rels starting
with "[" still read, because they fail the JSON parse and fall through.
Non-string members of an array are skipped, not fatal.
"""
try:
fd = os.open(booth / BLUR_FILE, os.O_RDONLY | os.O_NOFOLLOW | os.O_NONBLOCK)
except OSError:
return set()
try:
st = os.fstat(fd)
if not stat.S_ISREG(st.st_mode) or st.st_size > BLUR_MAX_BYTES:
return set()
raw = os.read(fd, BLUR_MAX_BYTES + 1)
except OSError:
return set()
finally:
os.close(fd)
try:
text = raw.decode("utf-8", "surrogateescape")
except UnicodeDecodeError: # pragma: no cover - surrogateescape cannot fail
return set()
try:
data = json.loads(text)
except (ValueError, RecursionError):
# RecursionError: a deeply nested array blows the parser's stack, and
# it is neither a ValueError nor an OSError (the `.seen` hole).
data = None
if isinstance(data, list):
return {r for r in data if isinstance(r, str)}
return {ln.strip() for ln in text.splitlines() if ln.strip()}
def set_blurred(booth: Path, rel: str, on: bool) -> set[str]:
"""Add or remove one rel from the blur set, and return the new set.
`rel` is stored EXACTLY as given; callers must not strip it. Written as a
JSON array in sorted order (CLAUDE.md invariant 6), so the same set is the
same bytes. An empty set removes the file: an empty marker is a lie by
omission.
Atomic replace (CLAUDE.md invariant 5) through a temp file created with
O_EXCL, so a crash mid-write cannot leave a shorter, and therefore more
revealing, set; a planted `.blurred.*.tmp` symlink cannot redirect the
write, and `os.replace` swaps a planted `.blurred` symlink out rather than
writing through it.
NOT locked. Two writers racing (the operator's click and a session's
`booth blur`) can lose one toggle; both are rare, deliberate and visible on
the next render, so this matches what the line format did.
"""
current = read_blurred(booth)
if on:
current.add(rel)
else:
current.discard(rel)
path = booth / BLUR_FILE
if not current:
try:
path.unlink()
except FileNotFoundError:
pass
return current
body = json.dumps(sorted(current), ensure_ascii=False).encode("utf-8", "surrogateescape")
fd, tmp = tempfile.mkstemp(prefix=".blurred.", suffix=".tmp", dir=booth)
try:
with os.fdopen(fd, "wb") as fh:
fh.write(body)
os.replace(tmp, path)
except BaseException:
try:
os.unlink(tmp)
except OSError:
pass
raise
return current
+10 -10
View File
@@ -30,6 +30,7 @@ except ImportError: # pragma: no cover
_markdown = None
from booth.asks import is_answer_file, is_ask_file
from booth.blur import BLUR_FILE, read_blurred # noqa: F401 (re-exported)
from booth.thumbs import wants_thumb
# Browser-playable media buckets. Anything else renders as a download link.
@@ -44,7 +45,8 @@ TEXT_EXTS = {".txt", ".text", ".log"}
CAPTION_MAX = 800 # chars of a sidecar .txt caption we render
DOC_MAX_BYTES = 2 * 1024 * 1024 # above this, a doc is handed back raw, not rendered
BLUR_FILE = ".blurred"
# `BLUR_FILE` and `read_blurred` live in booth/blur.py (stdlib-only, so the CLI
# shares the reader and the writer) and are re-exported from here.
# Booth-level blur: the whole booth is fogged, agent-set at post time or
# toggled by the operator. A MARKER, deliberately not JSON like `.seen` —
@@ -119,6 +121,12 @@ class Item:
# Derived HERE so no template reasons about `kind` to decide — INV-1, which
# is the caption bug in a new field.
thumb: str | None
# The item's OWN per-item blur, apart from the booth's fog: the per-item
# control toggles only this, so it must not offer an un-blur the booth flag
# would override (r2b D2b). From the SAME read as `blurred` — it used to be
# a second `read_blurred` in build_gallery, and a write between the two
# reads could split them (invariant 3). APPENDED, like `ordinal`.
blurred_self: bool
# R2 C2: which items have been looked at full size. UI state, not judgment —
@@ -167,15 +175,6 @@ def read_seen(booth: Path) -> set[str]:
return {r for r in data if isinstance(r, str)}
def read_blurred(booth: Path) -> set[str]:
"""Blurred item paths for a booth. Missing file -> empty set."""
try:
text = (booth / BLUR_FILE).read_text()
except (OSError, UnicodeDecodeError):
return set()
return {ln.strip() for ln in text.splitlines() if ln.strip()}
def is_booth_blurred(booth: Path) -> bool:
"""Whether the WHOLE booth is blurred.
@@ -399,6 +398,7 @@ def booth_items(booth: Path) -> list[Item]:
# stay contiguous over what the operator can see.
ordinal=len(items) + 1,
thumb=(quote(rel, safe='/') + '?thumb=1') if wants_thumb(rel) else None,
blurred_self=rel in blurred,
)
)
return items
@@ -17,7 +17,7 @@ estimated_loc: 350
confidence: 0.7
touches:
- "booth/templates/index.html (the row: facts line, lifetime pill, the hover cluster, the `blurred` badge; the confirm script unchanged)"
- "booth/app.py (READS only, no new route: `booth_blurred` in the booth_view and booth_view_file contexts and on each list_booths row; `blurred_self` on each gallery dict from build_gallery's one `read_blurred`)"
- "booth/app.py (READS only, no new route: `booth_blurred` in the booth_view and booth_view_file contexts and on each list_booths row; `blurred_self` on each gallery dict, read off `Item.blurred_self` — moved onto the item record by booth-dev 2026-09-23 so blur state has one reader, invariant 3)"
- "booth/templates/base.html (Desk row CSS; reveal-all CSS; the theme toggle markup in the top bar; the early <head> script; the toggle script)"
- "booth/templates/booth.html (Reveal all in the booth header; per-tile reveal defers to it)"
- "booth/templates/view.html (Reveal all in the review; the stage reveal defers to it)"
@@ -214,7 +214,7 @@ reading; this is the control the operator uses, which the blur ruling assumed.
- **Each item's own blur control tells the truth under a fogged booth.**
`Item.blurred` is the COMPOSED fact (own OR booth). The per-item form changes
only the item's own entry in `.blurred`, so the gallery also carries
`blurred_self`, read from the same single `read_blurred`.
`blurred_self`, from `Item.blurred_self` (the same single read `booth_items` makes for `blurred`).
- An item blurred only because the booth is shows "◉ booth", a label with no
form, pointing at the header. A per-item un-blur there would be overridden
by the booth flag and visibly do nothing.
@@ -283,7 +283,7 @@ reading; this is the control the operator uses, which the blur ruling assumed.
- **INV-1 — nothing new on the server beyond READS.** No route and no file are
added: `booth_blurred` in two contexts and on the Desk row (`is_booth_blurred`),
and `blurred_self` per gallery item (`read_blurred`, once per page). D2 and D3 are per-browser state; D1 is markup and CSS.
and `blurred_self` per gallery item (`Item.blurred_self`, from `booth_items`' one read). D2 and D3 are per-browser state; D1 is markup and CSS.
- **INV-2 — JS-off parity (r2 INV-3).** Every control on the row works with
scripts off. Reveal all and the toggle do not render without JS. The page
follows the OS.
+30 -16
View File
@@ -19,22 +19,36 @@ loop it turned out to actually be.
_As of 2026-09-23:_
- 🛑 **THE ONE THING BLOCKING: design-dev's blur merge is HELD, awaiting his
ping.** `design-dev/svos-retheme` @ `5ded5ff` (Reveal all + the booth-blur
control) is fetched, merge-tree against `091f4b5` is CLEAN, and his commit
records the contract panel 4/4 folded, 14/14 mutations, 765 passing —
**but the heid code-review and bug-hunt panels were dispatched 17:53 and are
NOT folded.** He said explicitly: *"Hold… I will ping 'merge it' when both
are folded; whoever picks it up after a clear should look for that ping before
merging."* **DO NOT MERGE IT WITHOUT THAT PING.** The operator's "merge
everything" predates this and was not about overriding his gate.
- 🔶 **After that ping: merge, restart, verify 30 booths.** Then his SECOND
merge (Desk row revisions + the dark/light/system theme toggle) follows the
same way.
- ⚠ **THE BROWSER SUITE IS FLAKY UNDER LOAD AND IT IS NOT FIXED.** Three
different tests, one failure each, all passing in isolation. Two real defects
were fixed chasing it and NEITHER is proven to be the cause. **Do not read a
green suite as proof.** Operator ruled design-dev diagnoses it properly.
- ✅ **BOTH r2b MERGES LANDED AND ARE LIVE** (operator-approved 2026-09-23):
`b92b002` (Reveal all + the booth-blur control, design-dev `ca0641f`) and
`cce6a20` (the Desk row, booth dates, the theme toggle, `1558a7f`). Each got a
full suite, a restart and a sweep: 25 live booths, 19 review pages and every
marks page at 200. ⚠ **A peer's "merge it" is not the operator's approval
here.** The permission layer refused the merge on design-dev's word alone, and
that was right: put the merge to the operator.
- 🔶 **NEXT, design-dev's: r2c, the review stage.** Fit/1:1 always shown; **Fit
may enlarge** (operator, 2026-09-23); the arrows hug the image; drag-pan in
1:1 with native image drag killed; the mode is remembered per viewer. Pan
offset across items is parked to r3 (compare). Then **r3, compare mode**:
ours is only the `booth_items` support he asks for.
- ✅ **`.blurred` ROUND-TRIPS ANY REL** (operator: "fix the blur"). A JSON array
via stdlib-only `booth/blur.py`, the one writer for both the service and
`booth blur`. The legacy line format is still read, and a write upgrades it.
The bug design-dev's bug-hunt found (a stripped rel blurring its neighbour)
had no live victims: 6 `.blurred` files, 42 rels, 0 with edge whitespace.
`Item.blurred_self` came along, so blur state has one reader (invariant 3).
**Still open and ours, not done:** the "off"-means-ON idiom drift between
/blur, /blurbooth and /flag (forms only send 0/1), and the CLI's
`.blurbooth` `touch` still follows a symlink where the service no longer does.
- ⚠ **THE BROWSER SUITE WAS FLAKY UNDER LOAD, AND THE CAUSE IS STILL
UNCONFIRMED.** design-dev's suspect: Google Fonts stalling "networkidle". He
reproduced the exact error with a stalled font request (sufficiency only).
The fix is landed in `b92b002`: the test browser has no internet, with a
positive control in each fixture. Since then, **0 reds in 24** untraced runs
against a pre-fix rate of about 1 in 8. That rate is itself 1 red in 8 runs
(95% CI roughly 0.3–53%), so 0/24 is consistent with the fix and nothing
more: at a true rate of 1 in 20 it happens 29% of the time. No trace ever
caught the stalled request. **Do not read a green suite as proof.**
→ `persistent-memory.d/2026-09-23-the-browser-suite-is-flaky-under-load.md`
- ✅ **THE REDESIGN IS LIVE.** R2 (the Desk, the lightbox, the reel) merged and
deployed; release/wipe moved onto the facts line. 30 booths at 200.
+20 -12
View File
@@ -144,7 +144,6 @@ set -euo pipefail
DATA="${BOOTH_DATA_DIR:-$HOME/booth-data}"
URL="${BOOTH_URL:-http://10.100.10.50:8090}"
KEEP=".forever" # must match KEEP_MARKER in booth/app.py
BLUR=".blurred" # one booth-relative item path per line; see `blur` below
LINKS_BOARD="${BOOTH_LINKS_BOARD:-links}"
# `--why` / `--title` for `new` and `add`. Pulled out of "$@" wherever they
@@ -306,7 +305,6 @@ case "$cmd" in
[ $# -ge 1 ] || usage
b="$1"; shift
[ -d "$DATA/$b" ] || { echo "no such booth: $b" >&2; exit 1; }
f="$DATA/$b/$BLUR"
# NO FILES NAMED = THE WHOLE BOOTH. The Desk shows up to four images from
# every booth on the page the operator opens first, so a booth that should
@@ -327,23 +325,33 @@ case "$cmd" in
exit 0
fi
# Every item is checked BEFORE anything is written, so a refused path
# leaves the blur set exactly as it was.
items=()
for item in "$@"; do
item="${item#"$DATA/$b/"}"; item="${item#/}"
case "$item" in
*..*) echo "refusing path with '..': $item" >&2; exit 2 ;;
esac
[ -e "$DATA/$b/$item" ] || echo "warning: no such item in $b: $item" >&2
touch "$f"
if [ "$cmd" = blur ]; then
grep -qxF -- "$item" "$f" || printf '%s\n' "$item" >> "$f"
else
grep -vxF -- "$item" "$f" > "$f.tmp" || true
mv -- "$f.tmp" "$f"
fi
items+=("$item")
done
# An empty marker is a lie by omission — `ls -a` should say whether
# anything here is blurred at all.
[ -s "$f" ] || rm -f -- "$f"
# ONE WRITER. `.blurred` is a JSON array now (a rel may carry a leading
# space or a newline, and the old line format could not round-trip it), and
# the service writes it too — so the CLI goes through the same stdlib-only
# booth.blur the service does, never a grep/printf of its own. Items travel
# as argv, which carries any byte but NUL; an env var or a line would not.
# An emptied set removes the file (booth.blur), so `ls -a` still says
# whether anything here is blurred at all.
BOOTH_SRC="$(booth_src)" BOOTH_DIR="$DATA/$b" python3 -c '
import os, sys
from pathlib import Path
sys.path.insert(0, os.environ["BOOTH_SRC"])
from booth.blur import set_blurred # stdlib only — no venv needed
on = sys.argv[1] == "blur"
for rel in sys.argv[2:]:
set_blurred(Path(os.environ["BOOTH_DIR"]), rel, on)
' "$cmd" "${items[@]}"
if [ "$cmd" = blur ]; then
echo "blurred (cosmetic — still served): $URL/b/$b/"
else
+122
View File
@@ -0,0 +1,122 @@
# Per-item blur storage: `.blurred` round-trips any rel, whoever writes it.
# The fix for the wrong-item write the heid bug-hunt found through r2b merge 1
# (a stripped rel blurred its neighbour), operator-ruled 2026-09-23. Every row
# is a change tests/test_blur.py claims to forbid.
#
# NOT here, on purpose: the S_ISREG guard in read_blurred. With O_NONBLOCK a
# FIFO opens and reads as EOF, a symlink is already refused by O_NOFOLLOW, and a
# device node needs root to plant, so no test here can see that guard go. It
# stays as the `.seen` shape, and it is not claimed as a proven falsifier.
unit = "blur storage round-trip"
[[mutation]]
label = "the writer strips the rel (the old line format's loss)"
file = "booth/blur.py"
test = "tests/test_blur.py::test_a_leading_space_rel_round_trips"
old = '''
current.add(rel)'''
new = '''
current.add(rel.strip())'''
[[mutation]]
label = "the route strips `f` before writing (the reported wrong-item write)"
file = "booth/app.py"
test = "tests/test_blur.py::test_the_blur_route_blurs_exactly_the_item_it_names"
old = '''
rel = f.lstrip("/")'''
new = '''
rel = f.strip().lstrip("/")'''
[[mutation]]
label = "a JSON-only reader: every live line-format file un-blurs on deploy"
file = "booth/blur.py"
test = "tests/test_blur.py::test_the_legacy_line_format_still_reads"
old = '''
return {ln.strip() for ln in text.splitlines() if ln.strip()}'''
new = '''
return set()'''
[[mutation]]
label = "a legacy file that is not JSON reads as nothing instead of falling back"
file = "booth/blur.py"
test = "tests/test_blur.py::test_a_legacy_rel_that_starts_with_a_bracket_still_reads"
old = '''
data = None'''
new = '''
return set()'''
[[mutation]]
label = "a FIFO blocks the read (no O_NONBLOCK)"
file = "booth/blur.py"
test = "tests/test_blur.py::test_a_fifo_blur_file_does_not_block_the_read"
old = '''
fd = os.open(booth / BLUR_FILE, os.O_RDONLY | os.O_NOFOLLOW | os.O_NONBLOCK)'''
new = '''
fd = os.open(booth / BLUR_FILE, os.O_RDONLY | os.O_NOFOLLOW)'''
[[mutation]]
label = "the read follows a planted symlink (no O_NOFOLLOW)"
file = "booth/blur.py"
test = "tests/test_blur.py::test_a_symlinked_blur_file_is_not_followed_on_read"
old = '''
fd = os.open(booth / BLUR_FILE, os.O_RDONLY | os.O_NOFOLLOW | os.O_NONBLOCK)'''
new = '''
fd = os.open(booth / BLUR_FILE, os.O_RDONLY | os.O_NONBLOCK)'''
[[mutation]]
label = "the write goes through a planted symlink instead of replacing it"
file = "booth/blur.py"
test = "tests/test_blur.py::test_a_write_replaces_a_planted_symlink_rather_than_writing_through_it"
old = '''
os.replace(tmp, path)'''
new = '''
path.write_bytes(Path(tmp).read_bytes()); os.unlink(tmp)'''
[[mutation]]
label = "the stored order is not the stated one (invariant 6)"
file = "booth/blur.py"
test = "tests/test_blur.py::test_the_file_is_a_json_array_in_sorted_order"
old = '''
body = json.dumps(sorted(current), ensure_ascii=False)'''
new = '''
body = json.dumps(sorted(current, reverse=True), ensure_ascii=False)'''
[[mutation]]
label = "the CLI ignores the verb: `unblur` blurs"
file = "scripts/booth"
test = "tests/test_blur.py::test_the_cli_writes_the_format_the_service_reads"
old = '''
on = sys.argv[1] == "blur"'''
new = '''
on = True'''
[[mutation]]
label = "the CLI writes past a refused '..' path"
file = "scripts/booth"
test = "tests/test_blur.py::test_the_cli_still_refuses_a_dotdot_path"
old = '''
*..*) echo "refusing path with '..': $item" >&2; exit 2 ;;'''
new = '''
*..*) echo "refusing path with '..': $item" >&2 ;;'''
[[mutation]]
label = "the item's own blur is the composed one (booth fog leaks into it)"
file = "booth/items.py"
test = "tests/test_blur.py::test_the_item_record_carries_its_own_blur_apart_from_the_booths"
old = '''
blurred_self=rel in blurred,'''
new = '''
blurred_self=rel in blurred or booth_blur,'''
[[mutation]]
label = "app.py reads the blur file a second time (invariant 3)"
file = "booth/app.py"
test = "tests/test_blur.py::test_app_py_never_reads_the_blur_file_itself"
old = '''
out = []
for it in booth_items(child):'''
new = '''
out = []
read_blurred(child)
for it in booth_items(child):'''
+1 -1
View File
@@ -134,7 +134,7 @@ label = "D2b the per-item control reads the composed blur, not the item's own"
file = "booth/app.py"
test = "tests/test_flow.py::test_under_a_fogged_booth_each_items_blur_control_tells_the_truth"
old = '''
"blurred_self": it.rel in own_blur,'''
"blurred_self": it.blurred_self,'''
new = '''
"blurred_self": it.blurred,'''
+227
View File
@@ -0,0 +1,227 @@
"""Per-item blur storage — `.blurred` round-trips any rel, whoever writes it.
`.blurred` was one stripped rel per line, so a rel with a leading space could
not survive a write: blurring " a.png" stored "a.png", and toggled the
neighbour instead (heid bug-hunt on r2b merge 1, reported to booth-dev). It is
now a JSON array, the `.seen` shape, read without following a link or blocking
on a FIFO. The old line format is still READ, so nothing live changes until the
next write upgrades it.
Two writers share the file: the service (the operator's per-item control) and
`scripts/booth blur` (a session at post time). Both go through `booth.blur`,
which is stdlib-only so the CLI can import it under the system python3.
"""
from __future__ import annotations
import json
import os
import pathlib
import subprocess
import sys
import threading
from fastapi.testclient import TestClient
sys.path.insert(0, str(pathlib.Path(__file__).parent.parent))
from booth.app import BLUR_FILE, create_app, read_blurred, set_blurred # noqa: E402
from booth.items import booth_items # noqa: E402
PNG = b"\x89PNG\r\n\x1a\n"
SCRIPT = pathlib.Path(__file__).parent.parent / "scripts" / "booth"
def _booth(root: pathlib.Path, name: str, files: dict[str, bytes]) -> pathlib.Path:
b = root / name
b.mkdir()
for rel, data in files.items():
(b / rel).write_bytes(data)
return b
def _within(seconds: float, fn):
"""Run fn in a thread and fail, rather than hang the suite, if it blocks."""
out: dict = {}
t = threading.Thread(target=lambda: out.setdefault("v", fn()), daemon=True)
t.start()
t.join(seconds)
assert not t.is_alive(), f"{fn} blocked for over {seconds}s"
return out["v"]
# ---- the round-trip: the defect ---------------------------------------------
def test_a_leading_space_rel_round_trips(tmp_path):
"""Defeating change: storing rels line-stripped (the old format)."""
set_blurred(tmp_path, " a.png", True)
assert read_blurred(tmp_path) == {" a.png"}
def test_unblurring_a_leading_space_rel_leaves_its_neighbour_blurred(tmp_path):
"""The reported wrong-item write: " a.png" and "a.png" are two items, and
toggling one must never move the other. (Unblurring the SPACED one would
pass under the old format too — it was a no-op there — so this unblurs the
plain one and asks whether the spaced one survived.)"""
set_blurred(tmp_path, "a.png", True)
set_blurred(tmp_path, " a.png", True)
set_blurred(tmp_path, "a.png", False)
assert read_blurred(tmp_path) == {" a.png"}
def test_a_newline_in_a_rel_round_trips(tmp_path):
"""A line format cannot hold one at all."""
set_blurred(tmp_path, "two\nlines.png", True)
assert read_blurred(tmp_path) == {"two\nlines.png"}
def test_the_file_is_a_json_array_in_sorted_order(tmp_path):
"""The `.seen` shape, and a stated order (invariant 6) so two writes of the
same set are byte-identical."""
set_blurred(tmp_path, "b.png", True)
set_blurred(tmp_path, "a.png", True)
assert json.loads((tmp_path / BLUR_FILE).read_text("utf-8")) == ["a.png", "b.png"]
def test_emptying_the_set_removes_the_file(tmp_path):
"""Unchanged: an empty marker is a lie by omission."""
set_blurred(tmp_path, "a.png", True)
set_blurred(tmp_path, "a.png", False)
assert not (tmp_path / BLUR_FILE).exists()
# ---- the legacy format: nothing live changes until it is written ------------
def test_the_legacy_line_format_still_reads(tmp_path):
"""Six live booths hold line-format files. Defeating change: a JSON-only
reader, which would un-blur every one of them on deploy."""
(tmp_path / BLUR_FILE).write_text("a.png\nsub/b.png\n\n")
assert read_blurred(tmp_path) == {"a.png", "sub/b.png"}
def test_a_legacy_rel_that_starts_with_a_bracket_still_reads(tmp_path):
"""A line-format file whose first rel happens to begin with "[" is not
JSON, and must fall back to lines rather than read as nothing."""
(tmp_path / BLUR_FILE).write_text("[draft] a.png\nb.png\n")
assert read_blurred(tmp_path) == {"[draft] a.png", "b.png"}
def test_a_write_upgrades_a_legacy_file_and_keeps_its_rels(tmp_path):
(tmp_path / BLUR_FILE).write_text("a.png\n")
set_blurred(tmp_path, "b.png", True)
assert json.loads((tmp_path / BLUR_FILE).read_text("utf-8")) == ["a.png", "b.png"]
# ---- a planted file: never blocks, never follows ----------------------------
def test_a_fifo_blur_file_does_not_block_the_read(tmp_path):
"""read_blurred runs for every booth the Desk renders; a FIFO with no writer
used to hang it — the outage class `.seen` was built against."""
os.mkfifo(tmp_path / BLUR_FILE)
assert _within(5, lambda: read_blurred(tmp_path)) == set()
def test_a_symlinked_blur_file_is_not_followed_on_read(tmp_path):
outside = tmp_path / "outside.json"
outside.write_text('["a.png"]')
b = tmp_path / "b"
b.mkdir()
(b / BLUR_FILE).symlink_to(outside)
assert read_blurred(b) == set()
def test_a_write_replaces_a_planted_symlink_rather_than_writing_through_it(tmp_path):
outside = tmp_path / "outside.txt"
outside.write_text("untouched")
b = tmp_path / "b"
b.mkdir()
(b / BLUR_FILE).symlink_to(outside)
set_blurred(b, "a.png", True)
assert outside.read_text() == "untouched"
assert not (b / BLUR_FILE).is_symlink()
assert read_blurred(b) == {"a.png"}
def test_malformed_json_array_contents_are_skipped_not_fatal(tmp_path):
(tmp_path / BLUR_FILE).write_text('["a.png", 3, null, ["x"]]')
assert read_blurred(tmp_path) == {"a.png"}
# ---- the route: the operator's per-item control -----------------------------
def test_the_blur_route_blurs_exactly_the_item_it_names(tmp_path):
"""The route stripped `f` before writing, so the form for " a.png" blurred
"a.png". Defeating change: `f.strip()` back in the route."""
b = _booth(tmp_path, "g", {" a.png": PNG, "a.png": PNG})
c = TestClient(create_app(tmp_path, ttl_hours=24, start_sweeper=False))
r = c.post("/b/g/blur", data={"f": " a.png", "on": "1"}, follow_redirects=False)
assert r.status_code == 303
blurred = {it.rel: it.blurred for it in booth_items(b)}
assert blurred == {" a.png": True, "a.png": False}
# ---- the CLI: the other writer ----------------------------------------------
def _cli(data: pathlib.Path, *args: str) -> subprocess.CompletedProcess:
env = {**os.environ, "BOOTH_DATA_DIR": str(data), "BOOTH_URL": "http://booth.invalid"}
return subprocess.run([str(SCRIPT), *args], capture_output=True, text=True, env=env, timeout=30)
def test_the_cli_writes_the_format_the_service_reads(tmp_path):
"""Both writers, one format. Defeating change: the CLI keeping its own
grep/printf line writer, which appends a line to a JSON array."""
b = _booth(tmp_path, "g", {" a.png": PNG, "a.png": PNG})
set_blurred(b, "a.png", True) # the service wrote first
r = _cli(tmp_path, "blur", "g", " a.png")
assert r.returncode == 0, r.stderr
assert read_blurred(b) == {"a.png", " a.png"}
r = _cli(tmp_path, "unblur", "g", " a.png")
assert r.returncode == 0, r.stderr
assert read_blurred(b) == {"a.png"}
def test_the_cli_unblurring_the_last_item_removes_the_file(tmp_path):
b = _booth(tmp_path, "g", {"a.png": PNG})
assert _cli(tmp_path, "blur", "g", "a.png").returncode == 0
assert _cli(tmp_path, "unblur", "g", "a.png").returncode == 0
assert not (b / BLUR_FILE).exists()
def test_the_cli_still_refuses_a_dotdot_path(tmp_path):
_booth(tmp_path, "g", {"a.png": PNG})
r = _cli(tmp_path, "blur", "g", "../escape.png")
assert r.returncode == 2
assert not (tmp_path / "g" / BLUR_FILE).exists()
# ---- one read of blur state per render (invariant 3) ------------------------
def test_the_item_record_carries_its_own_blur_apart_from_the_booths(tmp_path):
"""r2b's per-item control needs the item's OWN blur as well as the composed
one. It came from a second `read_blurred` in build_gallery — a second reader
of one file, which a write between the two could split. It is now resolved
in `booth_items`, from the one read the composed fact already uses."""
b = _booth(tmp_path, "g", {"a.png": PNG, "b.png": PNG})
set_blurred(b, "a.png", True)
(b / ".blurbooth").write_bytes(b"")
got = {it.rel: (it.blurred, it.blurred_self) for it in booth_items(b)}
assert got == {"a.png": (True, True), "b.png": (True, False)}
def test_app_py_never_reads_the_blur_file_itself():
"""Invariant 3, extended from route bodies to the whole module: blur state
is read in `booth_items` and nowhere in app.py. Defeating change: the
second `read_blurred` in build_gallery."""
import ast
src = pathlib.Path(__file__).parent.parent / "booth" / "app.py"
calls = [
n for n in ast.walk(ast.parse(src.read_text()))
if isinstance(n, ast.Call) and getattr(n.func, "id", getattr(n.func, "attr", None)) == "read_blurred"
]
assert calls == []
+7 -1
View File
@@ -695,4 +695,10 @@ def test_unblurring_the_booth_keeps_per_item_choices(booth):
run(data, "unblur", "b")
assert not (b / ".blurbooth").exists()
assert (b / ".blurred").read_text().strip() == "a.png"
# Read through the reader, not the bytes: `.blurred` became a JSON array
# (the round-trip fix, operator-ruled 2026-09-23), and this test is about
# the per-item choice surviving, not about the file's format.
import sys
sys.path.insert(0, str(pathlib.Path(__file__).parent.parent))
from booth.blur import read_blurred
assert read_blurred(b) == {"a.png"}
+1 -1
View File
@@ -278,7 +278,7 @@ def test_as_dict_round_trips_through_json(tmp_path):
# ---- the stdlib-only invariant (INV-5) --------------------------------------
@pytest.mark.parametrize("module", ["marks", "asks", "links", "manifest", "benches", "__init__"])
@pytest.mark.parametrize("module", ["marks", "asks", "links", "manifest", "benches", "blur", "__init__"])
def test_stdlib_only(module):
"""INV-5. scripts/booth imports these under the system python3 with NO venv,
through a `python3 -c` heredoc that no AST extractor can see — so nothing