Files
LEDMatrix/web_interface/blueprints/api_v3/__init__.py
T
ChuckandClaude Opus 5 69d408b321 feat(core): one per-element display-customization framework, wired into the web UI (#566)
* fix(sports): rebuild un-shared faces through the pinned layout engine

unshare_element_fonts re-instantiates a duplicate font face so two
elements can be told apart by id(). It did so through bare
ImageFont.truetype, which takes PIL's default layout engine rather than
the one src/common/font_layout.py pins. Raqm and Basic disagree on
fractional advances -- that disagreement is the reason the pin exists,
having broken golden images across machines -- so a rebuilt face could
measure differently from the shared face it replaced, on any host where
Raqm is installed.

These were the only two call sites in src/ bypassing the pin.

The guard asserts that the rebuild goes through the pinned loader rather
than comparing engine values: where Raqm is absent, bare truetype returns
BASIC anyway, so an engine comparison passes whether or not the pin is
honoured. The first draft of this test did exactly that and passed with
the bug reintroduced.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* refactor(web): drop the two dead client-side config-form renderers

generateConfigForm and generateSimpleConfigForm (580 lines) were defined
on the Alpine component and never called: server-side Jinja replaced them,
as pages_v3.py:641 records. Nothing in any template invokes them -- there
is no x-html in the templates and no bracket access on the component.

They carried their own x-widget dispatch, which made them an active trap:
the next person adding a widget would reasonably think both renderers
needed updating.

plugins/config_manager.js (PluginConfigManager, 133 lines) goes for the
same reason -- loaded on every page from base.html, referenced only by
itself and by an archived doc.

Kept, having checked them: widgets/example-color-picker.js is the worked
example docs/widget-guide.md points plugin authors at, and
widgets/plugin-loader.js is the client half of a documented feature
(manifest-declared plugin widgets) whose server route is missing --
soccer-scoreboard already ships a widgets/custom-leagues.js that this
loader is meant to fetch. That is an unfinished feature to complete, not
dead code to delete.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(web): serve plugin-declared widgets, and actually ask for them

LEDMatrixWidgets.loadPluginWidget has always fetched
/static/plugin-widgets/<plugin>/<widget>.js, and docs/widget-guide.md has
always documented that path, but nothing served it. soccer-scoreboard has
shipped a 17KB widgets/custom-leagues.js since August that could never
load. Both halves were missing, not just the route:

- serve_plugin_widget serves the script from the plugin's widgets/
  directory as text/javascript. The manifest is the allowlist -- only a
  widget the plugin declares is reachable -- so installing a plugin does
  not publish everything it ships. Path handling mirrors the sibling
  serve_plugin_web_ui: allowlist regexes, os.path.basename, resolve() +
  relative_to() containment, and the ledmatrix- prefix fallback. The
  declared script name is guarded too, since it comes from the plugin
  rather than the request.

- The config form never requested one. Its x-widget dispatch is a
  hardcoded list of core widget names, so a plugin's own widget fell
  through to a plain text input. An unrecognised x-widget on a string
  field now asks ensureWidget() for it. The text input stays as the
  fallback and is removed only once the widget has actually rendered, so
  a missing or broken widget costs the user an editor rather than their
  configured value on the next save.

- manifest_schema.json gains "widgets", so the declaration is validated
  rather than merely tolerated by additionalProperties.

Verified in a browser against the real partial: a declared widget loads,
registers and renders, and its field posts exactly one value; a field
whose widget 404s keeps its text input and still posts its value.

Not addressed: loadPluginWidgetsFromManifest still has no caller. The
per-field ensureWidget path is lazier and is what the form now uses, so
that bulk helper is dead weight -- worth removing, but left alone here
rather than inventing a call site for it.

Known limitation, documented: only string-typed fields take this path.
object/array/boolean/number fields and enums are dispatched by the
template's own branches, which still only know core widgets.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(element-style): a wrong-size BDF now keeps its font, not its size

BDF fonts are fixed-size bitmap strikes: FreeType accepts only the pixel
size baked into the file and raises for anything else. 32 of the 35
shipped fonts are BDF, so a size picked in the web UI usually is not a
valid strike -- and load_font caught that failure with its generic
"unloadable font" handler, which substitutes PressStart2P. Asking for
5x7.bdf at size 10 therefore rendered a completely different typeface,
silently.

It now falls back to the file's own native size instead, which is what
SportsCore._load_custom_font_from_element_config has always done. The
native size is read via FontManager._read_bdf_native_size rather than a
fourth copy of that parser, matching how core.py already delegates.

Also here, because they are the same code path:

- native_bdf_size() is exposed for the web UI, which needs to know when a
  size field can take effect at all. None means "free choice".
- ElementStyle.font_size now reports the size actually realised rather
  than the one requested. Callers lay out from it, and reserving space
  for a size nothing was drawn at is how this surfaces.
- The module font cache is a bounded LRU (256) instead of an unbounded
  dict. The display process runs for weeks and every config save can add
  a (font, size) pair; every other hot cache in the codebase is bounded
  this way.

Untouched configs are unaffected: the shipped classic fonts are the three
TTFs, so nothing was hitting the substitution path by default.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(element-style): per-mode style and offset overrides

Lets one element be styled differently per situation -- a scoreboard's
live / upcoming / recent cards, weather's current / hourly / daily
screens -- under customization.modes.<mode>.

The mode is bound at construction rather than passed per call. That is
what makes this cheap to adopt: SportsUpcoming and SportsRecent are
already separate instances with distinct SKIN_MODE values, so binding
once makes every existing style()/offset_value() call site mode-aware
without editing any of them. A per-call mode argument exists for the rare
host that renders more than one mode.

The two layers answer different questions, deliberately:

- The base layer keeps the existing "differs from the schema default"
  rule, because the save flow writes the full default object into
  config.json whether or not the user touched it.
- A mode layer is pure override -- its fields default to None, so
  presence is intent. Nothing writes into it unasked, so there is nothing
  for the stricter rule to protect against.

None therefore means inherit, and has to stay distinct from 0: a mode
y_offset of 0 means "sit at the base position", not "no preference".
This is the same distinction scroll_card.switch_* draws with "inherit".

A malformed mode value falls back to the resolved base value rather than
to the caller's default -- caught by the degradation tests, which is what
they are for: resolving the mode first let one bad string in a mode block
silently discard a good base offset.

With no modes block, and for every existing caller, resolution is
unchanged.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(element-style): declare per-mode overrides in config_schema.json

A plugin adds "x-style-modes": ["live", "upcoming", "recent"] alongside
its x-style-elements declaration and gets a customization.modes.<mode>
group per mode, with every field of every declared element repeated as an
override.

Those override fields are typed nullable and default to null, which is
the whole trick. The save flow writes schema defaults into config.json
wholesale, so giving a mode field the base element's default would make
every mode a frozen copy of the base the first time a user pressed Save,
and the base would stop reaching them. Null means inherit. The mutation
test for this is explicit: with concrete defaults, a base font_size of 14
resolves as 10 with user_forced set.

min/max from the declaration carry into the mode blocks, so an
out-of-range override is rejected by validation rather than clamped
silently at render time.

Also: the emitted font field now carries "x-widget": "font-selector". The
widget already shipped and the config form already allowlisted it -- the
hint was simply never emitted, so the field rendered as a bare text box
that the user had to type a font filename into.

Verified through the real SchemaManager path -- load_schema, defaults
extraction, merge_with_defaults, validation, then resolution -- rather
than against a hand-built dict, since the thing at risk is what that
pipeline does to a null.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(web): render the config form from the schema the save route validates

The form read config_schema.json with a raw json.load while
api_v3.save_plugin_config went through SchemaManager. Those are not the
same schema: SchemaManager applies expand_style_elements, which turns a
compact customization.x-style-elements declaration into the per-element
blocks the form knows how to render.

Without it, that customization object has an x-style-elements key and no
"properties", so the template's object branch matched nothing and the
section rendered as empty space -- while saving still validated against
the expanded shape. of-the-day ships the compact form, so its
customization section has been invisible in the web UI.

pages_v3 gains a schema_manager the way it already has config_manager and
plugin_manager. use_cache=False matches the save route, so an edited
schema is not served stale during plugin development. The raw read stays
as a fallback for callers that register this blueprint without one.

Checked before making the change: load_schema does nothing here except
read, validate and expand -- inject_skin_selector is a separate method it
does not call -- so this is not a behaviour change for schemas without
the declaration.

The test pair renders the same compact schema with and without a
SchemaManager, so it documents exactly what was broken as well as what is
fixed.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(web): style-editor widget -- a row per element instead of 65 accordions

Rendered element by element, a realistic scoreboard's customization block
is 65 nested sections, and reaching one per-mode font size takes five
levels of expanding. The widget collapses that to one compact row per
element -- font, size, colour, X, Y -- with a tab per declared mode.

It emits ordinary inputs under the same dotted names the generic renderer
would produce, so the save/validate/merge pipeline is untouched: no hidden
JSON blob and no new server-side parsing. It is driven entirely by the
schema block it is handed, so fields added to the schema later appear
without editing the widget. If it fails to load or throws, the generic
nested rendering it replaces is left in place.

Fixing two things the save path got wrong for nullable fields, found by
posting what the widget actually emits:

- The indexed-array recombiner (text_color.0/.1/.2 -> one list) compared
  the declared type to the string 'array', so a per-mode colour, typed
  ["array", "null"], was never reassembled and failed validation on save.
  _parse_form_value_with_schema had the same comparison.
- A blank nullable field became [] rather than None, which then failed the
  minItems the colour array declares. Null is the inherit sentinel, so it
  has to survive.

And two things the widget itself got wrong, found by looking at it:

- An unset base control fell back to the select's first option, so an
  untouched scoreboard claimed every element used 10x20.bdf -- and the
  size box then locked itself to that bitmap font's fixed size. Base
  controls now show the schema default; mode controls stay blank, because
  blank there means inherit.
- Elements arrived alphabetised (Detail and Odds above Score). Flask's
  JSON provider sorts keys, so declaration order has to be stated
  explicitly; expand_style_elements now emits x-propertyOrder, which the
  generic renderer already honoured too.

Size is disabled and shown as fixed for a bitmap font, using the
scalable/native_size the font catalog now reports.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(element-style): visibility, alignment and scale per element

Completes the customization vocabulary: hide an element, align it, and
resize a logo, alongside the font/size/colour/offset that already existed.
All three per mode.

They resolve to "change nothing" until the user asks for something -- True,
None and 1.0 -- rather than to whatever the schema declares. That is the
same invariant the font fields keep: a caller that honours them still
renders an untouched config exactly as it did before they existed. A
schema default therefore does not count as a choice, which matters because
the save flow writes that default into config either way.

scale sits in the layout block with the offsets rather than in the element
block, because it is geometry: a logo has a scale and no font. The widget's
columns come from the schema, so a logo row shows visibility, offsets and
scale and no empty font cell.

Two bugs found by the tests rather than by reading:

- A nullable enum needs null in its enum list, not just in its type. The
  mode copy of `align` defaulted to null and then failed its own schema, so
  a plugin declaring any enum field with modes could not save at all. Six
  tests failed on this before any of them reached what they were testing.
- defaults_from_schema only ever extracted font/font_size/text_color, so
  the schema defaults for the new fields were invisible to the resolver and
  a declared default read as a user choice.

Widget: the table scrolls horizontally and pins the element-name column.
Nine columns do not fit the config panel, and clipping them hid the offsets
entirely while scrolling them made every row anonymous.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(element-style): resolve elements under the names plugins actually use

Two naming conventions collided as the scoreboards grew. Counted across
the published schemas: the style block names elements with a _text suffix
(score_text, status_text, detail_text), while the layout block mostly uses
the bare noun (score, date, time, odds) -- except status_text, which kept
the suffix in seven plugins and lost it in two. records vs record splits
seven to two the same way.

A lookup now tries the exact name first and then the spellings that mean
the same thing. Exact-first is what makes this inert for any config that
already matches; the aliases only decide cases that resolved to nothing
before.

This is also what makes migrating to the compact declaration form safe.
That form uses one key for both blocks, so a scoreboard adopting it asks
for layout.score_text while its users have layout.score saved -- without
the aliases, every offset they had dialled in would silently become 0.

Applies to the style block, the layout block, the schema defaults and the
per-mode overrides, since the drift shows up in all four.

Not attempting to canonicalise on write: renaming keys in config.json
would break the plugins still reading the old spelling from their own
bundled code, and the drift costs a dict miss rather than correctness.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(plugins): BasePlugin.styles -- per-element styling every plugin inherits

Adopting the element-style system meant repeating three things in every
plugin: a guarded import, finding its own config_schema.json, and
rebuilding the resolver when on_config_change swapped the config dict.
This is those three things once, on the class all 45 plugins already
inherit from.

    title = self.styles.style('title_text',
                              classic_font='PressStart2P-Regular.ttf',
                              classic_size=8, classic_color=(255, 255, 255))

The classic_* arguments are the adoption contract: with nothing configured
they come back verbatim, so a plugin that switches to this renders exactly
as before until a user changes something.

A plugin with one instance per display mode sets STYLE_MODE on the class
and every existing lookup becomes mode-aware without a call site changing
-- which is the point of binding the mode to the resolver rather than
passing it per call. styles_for() covers a plugin that renders several
modes from one instance.

Schema discovery reads the concrete class's own module rather than this
file, because this file lives in src/plugin_system where no plugin schema
exists -- the same trap SportsCore._config_schema_path documents. The
first mutation test for that passed anyway: an installed plugin's module
directory and its entry under plugins_dir are the same path, so the test
could not tell the two apart. The case where they diverge is a plugin
symlinked in for development, and the test now forces that shape.

Getting discovery wrong is silent rather than loud: with no schema the
resolver has no defaults to compare against, so every configured value
reads as a deliberate override and the plugin quietly stops honouring its
own shipped styling.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(element-style): adopt hand-written customization blocks, and widen the font list

Nineteen plugins spell their style elements out longhand instead of
declaring them -- football's block is 701 lines for seven elements -- and
predate this system entirely. Core now recognises that shape, so they pick
up the row-per-element editor and the real font picker on a core update
rather than on a plugin release. Checked against every published schema:
21 plugins adopt, and the defaults of each still validate against the
schema generated for it.

Detection requires *every* field in a block to be one this system
understands. A looser "has at least one style field" rule sweeps in
baseball's `count`, which carries a text_color beside geometry that means
nothing here. That distinction took three attempts to test: the first two
assertions passed under both rules, because an over-eager rule leaves a
fontless block looking untouched and only surfaces as an extra row in the
editor.

The hardcoded font enum is replaced rather than extended. Football lists
five of the thirty-five installed fonts, which is why a font a user
uploads can never appear in one. It is not a curated safe set -- it omits
some twenty other faces that fit the declared size cap just as well -- it
is the fonts that happened to exist when it was written.

Widening it does need a guard, though, and not the one the schema already
has: a bitmap font ignores font_size and renders at its size baked into
the file, so `maximum: 16` cannot stop a 27px face. The picker now filters
out fixed-size fonts taller than the element's own declared ceiling, which
drops exactly the four that would overflow a 32px panel and keeps the
other thirty.

Per-mode overrides stay opt-in: core cannot invent a plugin's display
modes, so `x-style-modes` remains the one line that unlocks them. Their
layout half covers every positionable element rather than only those with
a style block -- the two namespaces do not line up in a hand-written
schema, and football positions six things (logos, timeouts, possession)
that have no style block at all.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* refactor(web): remove the two Fonts-tab panels that reported invented data

"Element Font Overrides" let a user configure an override, showed a
success toast, and changed nothing. All three endpoints behind it were
stubs -- GET returned a hardcoded {}, POST and DELETE returned success
without calling anything -- each marked "This would integrate with the
actual font system".

Wiring them to FontManager would not have fixed it. The machinery there is
real (_load_overrides/_save_overrides persist config/font_overrides.json,
resolve_font applies them, and the countdown plugin genuinely consumes
it), but the panel's element dropdown offered eleven invented keys --
nfl.live.score, clock.time, weather.current -- that no plugin has ever
read. An override saved against one of those would have persisted
correctly and still done nothing.

"Detected Manager Fonts" goes for the same reason. It claimed to show
"fonts currently in use by managers (auto-detected)"; its own comment said
"we'll simulate this", and it listed every font in the catalog with a
hardcoded usage_count of 1 -- the panel beside it, with fabricated
numbers attached.

Per-element font choice now lives in each plugin's own config editor,
against the elements that plugin actually has, and covers size, colour,
offsets, visibility, alignment and scale rather than family and size.

Kept: the font library (upload, preview, delete), which works, and
/fonts/tokens, which is a stub but genuinely feeds the preview's size
dropdown. FontManager's override methods are untouched -- countdown uses
them.

Verified in a browser with the tab's JS running: no console errors, 35
fonts listed, upload and preview intact. Removing the panel meant unwiring
it from populateFontSelects too, which would otherwise have bailed out
early on the missing select and left the preview dropdown empty.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* refactor(sports): one reader for element colours and layout offsets

There were two copies of the per-element colour read and three of the
layout-offset read. They had already drifted -- the scroll-card renderer
carries a comment about having ignored offsets its own schema advertised
-- and each new capability had to be added to all of them or silently work
in some places and not others.

All of them now go through src.element_style, which is what carries the
alias handling and the per-mode lookup. That lands immediately for the
nine plugins importing these modules: a scoreboard asking for `score_text`
offsets finds the `layout.score` its users configured, and a Live instance
resolves its own colours through SKIN_MODE without any call site passing a
mode.

_normalize_color learned "#RRGGBB" in the process. The scoreboards' own
readers have always accepted it, so the shared one had to, or consolidating
would have quietly dropped a form users' configs may hold. _coerce_offset
picked up the non-finite guard the scroll-card reader had and the other two
did not.

_get_layout_offset is promoted onto SportsCoreSharedMixin. Each plugin
still carries its own copy in its bundled sports.py, which wins by MRO --
so adopting this is a deletion in the plugin, and until that deletion
nothing changes for it.

Note for whoever runs the suite next: test_display_dirty_tracking.py is
order-dependent. Fifteen of its tests failed in one full run and passed in
the next with no change in between, and pass in isolation. Pre-existing,
unrelated to this, but it makes a full-run diff untrustworthy until it is
fixed.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* docs(changelog): record the element-style work under Unreleased

This file's own preamble asks for it: a plugin may delete its bundled
fallback copy of a core module only when its manifest floors on the first
release that shipped that module, which requires the additions to be
recorded here against a version.

Names a plugin can now import and floor on -- the stateless layout_offset
and element_color readers, alias_keys, native_bdf_size, the resolver's mode
binding, BasePlugin.styles, and the promoted
SportsCoreSharedMixin._get_layout_offset -- plus the schema and web-UI
changes, the four fixes and the three removals.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(fonts): log the BDF native-size read failure instead of swallowing it

The bdf-native-size lookup in get_fonts_catalog() caught any exception
and silently discarded it. Every other guarded read added in this PR
(the manifest parse in _declared_widget_script, the SchemaManager
fallback in _load_plugin_config_partial) logs before falling through
to the same degraded behavior. This one didn't, which is the shape a
silent-exception-swallow lint rule flags. Behavior is unchanged --
native_size still comes back None -- but a corrupt or unreadable BDF
file now leaves a trace.

Verified: font-related tests (140) and the full suite still pass,
with only the 2 pre-existing Europe/Kiev/Asia/Calcutta tzdata-alias
failures already present on origin/main.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix: address CodeRabbit findings on the style-editor/font-selector PR

- Fix _load_font_sized double-wrapping the (font, size) tuple on the
  missing-font path, which handed callers a tuple instead of a font.
- Fix _set_nested_value skipping an explicit None when the key already
  existed, which silently kept stale overrides when a user cleared a
  nullable per-mode field or blanked all channels of an indexed color.
- Preserve BDF scalable/native_size metadata through fetchFontCatalog's
  catalog-format mapping so maxFixedSize filtering actually applies.
- Stop caching an empty array on a failed font-catalog fetch so a later
  call can retry instead of being stuck with the failed result.
- Keep a saved font selected in the style editor even when it no longer
  fits a newly declared maxFixedSize, instead of silently deselecting it.
- Don't drop in-progress user edits to fallback fields when a plugin
  widget finishes loading asynchronously and takes over the form.
- Tighten the removed font-override endpoint test to assert 405, not
  just != 200.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(web): a partial save no longer switches off checkboxes it never showed

An HTML checkbox posts nothing when unchecked, so the save route walked the
schema and forced every boolean missing from the form to False. That is right
for the rendered form and wrong for every other caller: a script, the MQTT
bridge or a curl against the documented endpoint never rendered a checkbox, and
reading its silence as "all off" turns a one-field save into a mass disable.

Found on hardware. Posting four customization.* keys to a live device switched
off nfl.enabled, ncaa_fb.enabled and every display-mode toggle in one request.

The form now reports the top-level sections it drew (__rendered_section), and
inside those an absent checkbox still means unchecked -- including a section
whose only fields are checkboxes that are all off, which no heuristic could
recover. A post with no marker only touches objects it actually posted a field
from. Meta fields are dropped before form keys are treated as config paths,
because unknown keys are otherwise written straight into config.json.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* feat(sports): resolve element colour by name, and honour visible/align/scale

Two of the three gaps this framework shipped with.

Colour by name. A draw resolved its colour by comparing the *identity* of the
font object it was handed, which cannot tell two elements apart when they share
a face -- so those draws went out white. Every bitmap font is in that case,
because a freetype.Face cannot be re-instantiated to un-share it, which is how
an element rendered in any of the 32 shipped BDF fonts silently lost a colour
its picker had offered all along. _draw_text_with_outline now takes
element="score_text" and reads the colour by name; the identity path remains
for un-annotated callers, but narrows before giving up -- one configured colour
among the sharers is the only thing the user can have meant.

Visible, align and scale. The resolver has understood these since the
framework landed and nothing consumed them: an element could be marked hidden
in the web UI and still render. Adds the stateless readers, the mixin
accessors, and a scale parameter on the one shared logo-sizing seam (keyed into
the cache, so two elements scaled differently cannot be served each other's
image). Naming an element in a draw also honours its visibility.

Untouched configs are unaffected: every new parameter defaults to today's
behaviour, and all ten affected plugins render pixel-identically to main across
every harness size.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* docs(plugins): how to declare styleable elements; harden the widget's lookups

The plugin-author guide for the compact x-style-elements declaration -- what
each key does, how to read values back without breaking the "user-forced only
when it differs from the default" rule, and why a hand-written block needs no
changes to be adopted.

Also clears the static-analysis findings on style-editor.js. Every lookup in
that file is keyed by something out of a schema or a saved config, so a key of
__proto__ or constructor would walk the prototype chain and hand back a
function instead of a schema; reads now go through an own-property helper. The
panel registry became a list, and the flagged vars moved to their function
roots.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(web): clear the remaining static-analysis findings

Five, all on lines this branch touched.

The Python one is not a new defect: _set_missing_booleans_to_false's first
parameter was always named `config`, which shadows the `config` submodule
imported for its side effects at the bottom of this module. Editing the
signature simply put the existing warning on a changed line. The parameter is
the plugin's config dict, so `plugin_config` is what it should have been called
anyway; callers pass it positionally and are unaffected.

The JavaScript ones are the object-injection rule firing on reads keyed by
data. own() now goes through a property descriptor, so the one unavoidable
data-keyed read is no longer a computed member access; at() consumes its path
instead of indexing it; and the column set is a Map, which has no prototype to
pollute and needs no guarded reads at all.

Verified the widget still renders identically against football's real schema:
29 element rows, all four mode tabs, values populated, no console errors.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(web): drop the hasOwnProperty alias the descriptor read made redundant

own() now reads through Object.getOwnPropertyDescriptor, so the alias it used to call has no remaining reference.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
2026-09-13 11:53:50 -04:00

2161 lines
92 KiB
Python

"""The /api/v3 blueprint.
This was one 10,469-line module. It is now a package, but deliberately not a
set of *separate* blueprints: every route module beside this one decorates the
single `api_v3` object defined here, so endpoint names stay `api_v3.<function>`
and the URL map is unchanged. web_interface/app.py registers it as before.
The shared imports, constants and helpers stay in this file rather than moving
to a _common submodule, because tests monkeypatch some of them by module
attribute -- `monkeypatch.setattr(api_v3_module, "_BACKUP_EXPORT_DIR", ...)`.
Keeping them here means that keeps working exactly as it did. Route modules
read the mutable ones back through this module (see _pkg below) rather than
binding them by value, for the same reason.
The route modules are imported at the *bottom*: they import names from here, so
everything they need has to exist first.
"""
from flask import Blueprint, request, jsonify, Response
import contextlib
import json
import os
import re
import stat
import sys
import shutil
import subprocess
import tempfile
import time
import hashlib
import uuid
import logging
from datetime import datetime
from pathlib import Path
from typing import Dict, Any, Optional, Tuple, Type
from urllib.parse import urlparse, urlunparse
logger = logging.getLogger(__name__)
# Import new infrastructure
from src.web_interface.api_helpers import success_response, error_response, validate_request_json
from src.web_interface.errors import ErrorCode
from src.web_interface.secret_helpers import (find_secret_fields, mask_all_secret_values,
merge_secrets, remove_empty_secrets,
separate_secrets,
strip_masked_values)
from src.web_interface.error_handler import describe_exception, redact_text
from src.plugin_system.operation_types import OperationType
from src.web_interface.validators import (
validate_file_upload
)
from src.error_aggregator import get_error_aggregator
from src.common.permission_utils import install_requirements_file
from src.common.path_safety import resolve_under
_SUDO = shutil.which('sudo')
_JOURNALCTL = shutil.which('journalctl')
_GIT = shutil.which('git')
# Cap subprocess output returned to the browser — pip can produce MBs on build failures.
_MAX_OUTPUT_BYTES = 51_200 # 50 KB
def _truncate_output(stdout: str, stderr: str) -> str:
"""Combine stdout+stderr and truncate to _MAX_OUTPUT_BYTES (keeping the tail)."""
combined = (stdout + stderr).strip()
if len(combined) > _MAX_OUTPUT_BYTES:
combined = '[...output truncated...]\n' + combined[-_MAX_OUTPUT_BYTES:]
return combined
def _pip_install_requirements(req_file: Path, timeout: int) -> subprocess.CompletedProcess:
"""Install a requirements.txt file, preferring the vetted sudo wrapper so
the packages are visible to root-run ledmatrix.service — not just to
whichever non-root user runs this web process. Falls back to installing
for the current process only if the wrapper isn't set up yet (i.e. the
admin hasn't run scripts/install/configure_web_sudo.sh since upgrading),
so the button still does *something* useful rather than hard-failing.
Thin wrapper around the shared implementation in permission_utils so the
Plugin Store's own dependency installation (store_manager.py) follows the
exact same root-visible install path instead of a divergent one.
"""
return install_requirements_file(req_file, timeout=timeout)
def _scrub_git_remote_url(url: str) -> str:
"""Strip embedded username/password from an HTTPS remote URL before returning it to the UI."""
try:
p = urlparse(url)
if p.scheme in ('http', 'https') and (p.username or p.password):
netloc = p.hostname or ''
if p.port:
netloc += f':{p.port}'
return urlunparse(p._replace(netloc=netloc))
except Exception:
pass
return url
# Will be initialized when blueprint is registered
# NOTE: the managers live on the blueprint object (app.py sets
# api_v3.config_manager / api_v3.plugin_manager). Deliberately not
# mirrored as module globals: a bare `config_manager` used to resolve to
# a None that was never assigned, which silently disabled the /health
# checks and made /display/current fall back to a hardcoded 128x64.
plugin_store_manager = None
saved_repositories_manager = None
cache_manager = None
schema_manager = None
operation_queue = None
plugin_state_manager = None
operation_history = None
sync_manager = None # Optional DisplaySyncManager instance (set by app.py if available)
# Get project root directory (web_interface/../..)
# web_interface/blueprints/api_v3/_common.py -> up four to the project root.
# This was three levels when everything lived in web_interface/blueprints/api_v3.py;
# the split moved the file one directory deeper and silently pointed PROJECT_ROOT
# at web_interface/ instead. Nothing failed at import -- it surfaced as routes
# 404ing and "installation script not found", because every path built from it
# was wrong. Asserted in test_api_v3_url_map.py so the next move cannot repeat it.
PROJECT_ROOT = Path(__file__).resolve().parents[3]
# System fonts that cannot be deleted (used by catalog API and delete endpoint)
SYSTEM_FONTS = frozenset([
'pressstart2p-regular', 'pressstart2p',
'4x6-font', '4x6',
'5by7.regular', '5by7', '5x7',
'5x8', '6x9', '6x10', '6x12', '6x13', '6x13b', '6x13o',
'7x13', '7x13b', '7x13o', '7x14', '7x14b',
'8x13', '8x13b', '8x13o',
'9x15', '9x15b', '9x18', '9x18b',
'10x20',
'matrixchunky8', 'matrixlight6', 'tom-thumb',
'clr6x12', 'helvr12', 'texgyre-27'
])
api_v3 = Blueprint('api_v3', __name__)
def _get_plugin_version(plugin_id: str) -> str:
"""Read the installed version from a plugin's manifest.json.
Returns the version string on success, or '' if the manifest
cannot be read (missing, corrupt, permission denied, etc.) or if
``plugin_id`` is not a plain directory name. Several callers pass an id
that arrived in a request body, so the name is validated here rather
than relying on each of them to have done it.
"""
manifest_path = resolve_under(
api_v3.plugin_store_manager.plugins_dir, plugin_id, "manifest.json"
)
if manifest_path is None:
logger.warning("[PluginVersion] Rejected unsafe plugin id %r", plugin_id)
return ''
try:
with open(manifest_path, 'r', encoding='utf-8') as f:
manifest = json.load(f)
return manifest.get('version', '')
except (FileNotFoundError, PermissionError, OSError) as e:
logger.warning("[PluginVersion] Could not read manifest for %s at %s: %s", plugin_id, manifest_path, e)
except json.JSONDecodeError as e:
logger.warning("[PluginVersion] Invalid JSON in manifest for %s at %s: %s", plugin_id, manifest_path, e)
return ''
def _is_plugin_update_available(installed_version: str, latest_version: str) -> bool:
"""Return True when the registry's ``latest_version`` is strictly newer
than the installed version.
Thin alias for the shared comparator in
`src.plugin_system.compatibility.is_update_available` — the store's
`update_plugin` uses the same function, so the UI badge and the actual
reinstall decision can never disagree.
"""
from src.plugin_system.compatibility import is_update_available
return is_update_available(installed_version, latest_version)
def _ensure_cache_manager():
"""Ensure cache manager is initialized."""
global cache_manager
if cache_manager is None:
from src.cache_manager import CacheManager
cache_manager = CacheManager()
return cache_manager
def _save_config_atomic(config_manager, config_data, create_backup=True):
"""
Save configuration using atomic save if available, fallback to regular save.
Returns:
tuple: (success: bool, error_message: str or None)
"""
if hasattr(config_manager, 'save_config_atomic'):
result = config_manager.save_config_atomic(config_data, create_backup=create_backup)
if result.status.value != 'success':
return False, result.message
return True, None
else:
try:
config_manager.save_config(config_data)
return True, None
except Exception as e:
return False, str(e)
def _coerce_to_bool(value):
"""
Coerce a form value to a proper Python boolean.
HTML checkboxes send string values like "true", "on", "1" when checked.
This ensures we store actual booleans in config JSON, not strings.
Args:
value: The form value (string, bool, int, or None)
Returns:
bool: True if value represents a truthy checkbox state, False otherwise
"""
if value is None:
return False
if isinstance(value, bool):
return value
if isinstance(value, int):
return value == 1
if isinstance(value, str):
return value.lower() in ('true', 'on', '1', 'yes')
return False
def _get_display_service_status():
"""Return status information about the ledmatrix service."""
try:
result = subprocess.run(
['systemctl', 'is-active', 'ledmatrix'],
capture_output=True,
text=True,
timeout=3
)
return {
'active': result.stdout.strip() == 'active',
'returncode': result.returncode,
'stdout': result.stdout.strip(),
'stderr': result.stderr.strip()
}
except subprocess.TimeoutExpired:
return {
'active': False,
'returncode': -1,
'stdout': '',
'stderr': 'timeout'
}
except Exception as err:
return {
'active': False,
'returncode': -1,
'stdout': '',
'stderr': str(err)
}
def _run_systemctl_command(args):
"""Run a systemctl command safely."""
try:
result = subprocess.run(
args,
capture_output=True,
text=True,
timeout=15
)
return {
'returncode': result.returncode,
'stdout': result.stdout,
'stderr': result.stderr
}
except subprocess.TimeoutExpired:
return {
'returncode': -1,
'stdout': '',
'stderr': 'timeout'
}
except Exception as err:
return {
'returncode': -1,
'stdout': '',
'stderr': str(err)
}
def _ensure_display_service_running():
"""Ensure the ledmatrix display service is running."""
status = _get_display_service_status()
if status.get('active'):
status['started'] = False
return status
result = _run_systemctl_command(['sudo', 'systemctl', 'start', 'ledmatrix.service'])
service_status = _get_display_service_status()
result['started'] = result.get('returncode') == 0
result['active'] = service_status.get('active')
result['status'] = service_status
return result
def _stop_display_service():
"""Stop the ledmatrix display service."""
result = _run_systemctl_command(['sudo', 'systemctl', 'stop', 'ledmatrix.service'])
status = _get_display_service_status()
result['active'] = status.get('active')
result['status'] = status
return result
#: Field names whose value is a credential. Matched by name because this
#: endpoint returns the whole config, core keys included, and core config has
#: no schema to carry x-secret markers.
_CREDENTIAL_NAME_PARTS = ("password", "passwd", "secret", "token", "api_key",
"apikey", "access_key", "private_key", "client_secret")
def _looks_like_a_credential(name: str) -> bool:
lowered = name.lower()
return any(part in lowered for part in _CREDENTIAL_NAME_PARTS)
def _redact_credentials(value):
"""A copy of `value` with credential-named fields blanked.
/config/main returned the raw config to anyone who could reach the port,
and this interface has no authentication. On one rig that meant a 40-char
GitHub token, a 183-char Home Assistant token and five API keys were
readable by anything on the LAN.
The x-secret masking used by the plugin config endpoints does not help
here: this endpoint never consults a schema, and core keys such as
github.api_token have no schema to mark. Matching on the field name is
blunt, but for a whole-config dump the right default is that anything
named like a credential does not leave the process.
Blanked rather than removed, and safe to blank: POST /config/main merges
into the loaded config and only writes the keys it was given, so a client
that round-trips this response cannot erase a secret it never saw.
"""
if isinstance(value, dict):
return {k: (_blank_credential_value(v) if _looks_like_a_credential(k)
else _redact_credentials(v))
for k, v in value.items()}
if isinstance(value, list):
return [_redact_credentials(item) for item in value]
return value
def _blank_all_scalars(value):
"""Blank every scalar reached from `value`, at any depth.
Unlike `_blank_credential_value`, this never delegates back to
`_redact_credentials`'s name-based walk: an object reached through a
credential-owned list (e.g. `tokens: [{"value": "secret"}]`) has no
field name of its own to test, so every scalar inside it is blanked
regardless of what its keys are called.
"""
if isinstance(value, dict):
return {k: _blank_all_scalars(v) for k, v in value.items()}
if isinstance(value, list):
return [_blank_all_scalars(item) for item in value]
return ""
def _blank_credential_value(value):
"""Blank a value that sits under a credential-shaped key.
A dict is still walked -- `secrets: {api_key: ..., note: ...}` is a
section name, not a value to blank in one go, so sub-fields that are not
themselves credential-named survive (see
test_a_credential_shaped_container_is_still_walked). A list has no field
names to test for its items, though, so every scalar reached through it
is blanked outright: a bare list of secrets (`tokens: ["a", "b"]`) and a
list of credential-shaped objects (`tokens: [{"value": "secret"}]`) are
both blanked at any depth via `_blank_all_scalars`.
"""
if isinstance(value, dict):
return _redact_credentials(value)
if isinstance(value, list):
return [_blank_all_scalars(item) for item in value]
return ""
def _validate_time_format(time_str):
"""Validate time format is HH:MM"""
try:
datetime.strptime(time_str, '%H:%M')
return True, None
except (ValueError, TypeError):
return False, f"Invalid time format: {time_str}. Expected HH:MM format."
def _git_current_branch(project_dir):
"""Current branch name, or '' when detached or git fails."""
try:
r = subprocess.run(['git', 'branch', '--show-current'],
capture_output=True, text=True, timeout=10, cwd=str(project_dir))
return r.stdout.strip() if r.returncode == 0 else ''
except (subprocess.TimeoutExpired, OSError):
return ''
def _git_upstream(project_dir):
"""Configured upstream for the current branch (e.g. 'origin/main'), or ''."""
try:
r = subprocess.run(['git', 'rev-parse', '--abbrev-ref', '--symbolic-full-name', '@{u}'],
capture_output=True, text=True, timeout=10, cwd=str(project_dir))
return r.stdout.strip() if r.returncode == 0 else ''
except (subprocess.TimeoutExpired, OSError):
return ''
def _git_remote_branch_exists(project_dir, branch):
"""True when origin/<branch> exists locally as a remote-tracking ref."""
if not branch:
return False
try:
r = subprocess.run(
['git', 'show-ref', '--verify', '--quiet', f'refs/remotes/origin/{branch}'],
capture_output=True, text=True, timeout=10, cwd=str(project_dir))
return r.returncode == 0
except (subprocess.TimeoutExpired, OSError):
return False
def resolve_pull_command(project_dir):
"""Work out how to pull, for branches with and without an upstream.
A plain ``git pull --rebase`` fails outright on a branch that has no
upstream ("There is no tracking information for the current branch"),
which is easy to end up on: checking out a branch by name, restoring a
backup, or following an install guide that names one. The update button
then reports a failure the user cannot act on.
``--autostash`` is passed for the same reason. Rebase refuses to start
when any tracked file is modified, and on these installs something always
is: first_time_install.sh chmods five scripts that git tracked as 644, so
every machine that ran the installer carries five permanent mode changes
and the update button reports "cannot pull with rebase: You have unstaged
changes". Those modes are corrected in this commit, but a user cannot pull
the correction while the pull is what is blocked, and any other local edit
would reproduce it anyway. Autostash reapplies the changes afterwards.
Returns ``(args, note, error)``. When ``origin/<branch>`` exists the pull
is made explicit against it, so the update proceeds and the branch is
given tracking information afterwards.
"""
upstream = _git_upstream(project_dir)
if upstream:
return ['git', 'pull', '--rebase', '--autostash'], '', None
branch = _git_current_branch(project_dir)
if not branch:
return None, '', (
"This checkout is in a detached HEAD state, so there is no branch "
"to update. Switch to a branch first (Tools -> Switch branch)."
)
if _git_remote_branch_exists(project_dir, branch):
return (
['git', 'pull', '--rebase', '--autostash', 'origin', branch],
f"Branch '{branch}' had no upstream; pulled from origin/{branch} and set it as the upstream.",
None,
)
return None, '', (
f"Branch '{branch}' has no upstream and there is no origin/{branch} to "
f"pull from. Use Switch branch to move to a branch that exists on the "
f"remote, or push this one first."
)
_BRANCH_NAME_RE = re.compile(r'[A-Za-z0-9._/-]{1,200}')
def is_valid_branch_name(name):
"""Accept only plain branch names.
This value becomes a subprocess argument, so anything exotic is refused
rather than escaped. '..' is excluded because it is range syntax to git.
"""
if not name or not _BRANCH_NAME_RE.fullmatch(name):
return False
return '..' not in name and not name.startswith('-')
def checkout_branch(project_dir, target, stash=False):
"""Switch the checkout to `target`, returning (payload, http_status).
Split out of the route so it can be tested against real repositories.
Attaches tracking when the branch exists on origin, so the next
Pull Latest is a plain `git pull` rather than the no-upstream fallback.
"""
target = (target or '').strip()
if not target:
return {'status': 'error', 'message': 'Branch name required'}, 400
if not is_valid_branch_name(target):
return {'status': 'error', 'message': 'Invalid branch name'}, 400
try:
subprocess.run(['git', 'fetch', 'origin', '--prune'],
capture_output=True, text=True, timeout=60, cwd=project_dir)
local_exists = subprocess.run(
['git', 'show-ref', '--verify', '--quiet', f'refs/heads/{target}'],
capture_output=True, text=True, timeout=10, cwd=project_dir).returncode == 0
remote_exists = _git_remote_branch_exists(project_dir, target)
if not local_exists and not remote_exists:
return {'status': 'error',
'message': f"No branch '{target}' locally or on origin"}, 404
# Local edits block a checkout. Pull Latest already stashes for the
# same reason, so offer it here too -- but only when asked, never
# silently: putting someone's edits away unasked is worse than
# refusing the switch.
stash_note = ''
if stash:
stashed = subprocess.run(['git', 'stash', 'push', '-m', f'switch to {target}'],
capture_output=True, text=True, timeout=60, cwd=project_dir)
if stashed.returncode == 0 and 'No local changes' not in stashed.stdout:
stash_note = ' Local changes were stashed (recover them with git stash list).'
if local_exists:
co = subprocess.run(['git', 'checkout', target],
capture_output=True, text=True, timeout=60, cwd=project_dir)
else:
# -B so a stale local ref does not block the checkout.
co = subprocess.run(['git', 'checkout', '-B', target, f'origin/{target}'],
capture_output=True, text=True, timeout=60, cwd=project_dir)
if co.returncode != 0:
logger.warning("git checkout %s failed: %s", target, co.stderr)
return {
'status': 'error',
'message': f"Could not switch to '{target}'.",
# Keep git's full list of blocking files: naming them is the
# difference between an error the user can act on and one they
# cannot.
'detail': (co.stderr or '').strip(),
'can_retry_with_stash': 'would be overwritten by checkout' in (co.stderr or ''),
}, 200
if remote_exists:
subprocess.run(['git', 'branch', f'--set-upstream-to=origin/{target}', target],
capture_output=True, text=True, timeout=10, cwd=project_dir)
logger.info("Switched checkout to branch %s", target)
return {
'status': 'success',
'message': f"Now on '{target}'.{stash_note} Use Pull Latest to fetch its newest code.",
}, 200
except subprocess.TimeoutExpired:
return {'status': 'error', 'message': 'Timed out talking to git'}, 504
except OSError as exc:
logger.error("checkout_branch failed: %s", exc, exc_info=True)
return {'status': 'error', 'message': 'Could not switch branch'}, 500
def get_git_version(project_dir=None):
"""Get git version information from the repository"""
if project_dir is None:
project_dir = PROJECT_ROOT
try:
# Try to get tag description (e.g., v2.4-10-g123456)
result = subprocess.run(
['git', 'describe', '--tags', '--dirty'],
capture_output=True,
text=True,
timeout=5,
cwd=str(project_dir)
)
if result.returncode == 0:
version_str = result.stdout.strip()
if re.match(r'^[a-zA-Z0-9._\-]+$', version_str):
return version_str
# Fallback to short commit hash
result = subprocess.run(
['git', 'rev-parse', '--short', 'HEAD'],
capture_output=True,
text=True,
timeout=5,
cwd=str(project_dir)
)
if result.returncode == 0:
version_str = result.stdout.strip()
if re.match(r'^[a-zA-Z0-9._\-]+$', version_str):
return version_str
return 'Unknown'
except Exception:
return 'Unknown'
_update_check_cache: Dict[str, Any] = {'result': None, 'ts': 0.0}
_UPDATE_CHECK_TTL = 300 # 5 minutes — avoids a git fetch on every page load
def _update_check_failed(detail: str) -> Dict[str, Any]:
"""A check that could not run is not the same as being up to date.
Reporting update_available=False on a git failure hides the banner, and
the banner is the only route to the update button -- so a checkout git
refuses to touch looks exactly like a current one, permanently. The most
common cause is an install performed as root: git then reports "dubious
ownership" and every command fails, including the fetch here.
"""
return {'update_available': False, 'remote_sha': 'unknown',
'commits_behind': 0, 'check_failed': True, 'error': detail}
def _describe_git_failure(stderr: str) -> str:
"""Turn git's stderr into something the user can act on."""
text = (stderr or '').strip()
if 'dubious ownership' in text or 'detected dubious ownership' in text:
return ("This checkout is owned by a different user than the one "
"running the web interface, so git refuses to use it. It is "
"usually the result of installing as root. Fix the ownership "
"and the update will work: sudo chown -R $USER:$USER "
+ str(PROJECT_ROOT))
if 'could not resolve host' in text.lower() or 'network is unreachable' in text.lower():
return "Could not reach GitHub to check for updates."
return "Could not check for updates: " + (text.splitlines()[0] if text else "git failed")
def _installed_plugin_ids():
"""Best-effort list of installed plugin IDs for the web process.
Health/metrics state is written by the separate display service to the
shared on-disk cache, so the tracker's in-memory set is empty here. We
enumerate the installed plugins and read each one's persisted summary by ID
instead of relying on the tracker's in-memory `get_all_*` view.
"""
pm = api_v3.plugin_manager
manifests = getattr(pm, 'plugin_manifests', None)
if not manifests:
# Only pay for a discovery scan when we haven't discovered anything yet;
# subsequent polls reuse the already-populated manifest map.
try:
pm.discover_plugins()
except Exception:
logger.debug('discover_plugins failed while listing plugin ids', exc_info=True)
manifests = getattr(pm, 'plugin_manifests', None)
try:
return list(manifests.keys()) if manifests else []
except Exception:
logger.debug('listing plugin_manifests failed while building plugin ids', exc_info=True)
return []
def _do_transactional_uninstall(plugin_id, preserve_config):
"""Execute an uninstall with snapshot-based rollback.
Order of operations:
1. Snapshot main config + secrets (abort on unexpected errors, proceed on expected I/O errors).
2. Clean up plugin config (abort with 500 if this raises — avoids orphaned files).
3. Unload plugin from runtime if loaded (rollback + 500 if this raises).
4. Remove plugin files (rollback + 500 if this returns False or raises).
5. Finish (remove state, invalidate caches).
Rollback restores the config snapshot and, if the plugin had been
loaded before unload, calls load_plugin to restore runtime state.
Returns (True, None) on success or (False, error_message) on failure.
"""
from src.exceptions import ConfigError
# --- Step 1: snapshot main + secrets ---
main_snapshot = None
secrets_snapshot = None
try:
main_snapshot = api_v3.config_manager.get_raw_file_content('main')
except (OSError, ConfigError):
pass # Proceed without snapshot; narrow catch preserves TypeError/AttributeError
try:
secrets_snapshot = api_v3.config_manager.get_raw_file_content('secrets')
except (OSError, ConfigError):
pass
# --- Step 2: cleanup config first (abort before touching filesystem) ---
if not preserve_config:
api_v3.config_manager.cleanup_plugin_config(plugin_id, remove_secrets=True)
# Record whether the plugin was running before we touch anything.
was_loaded = (
api_v3.plugin_manager is not None
and plugin_id in api_v3.plugin_manager.plugins
)
def _rollback(reload_plugin):
if main_snapshot is not None:
try:
api_v3.config_manager.save_raw_file_content('main', main_snapshot)
except Exception as restore_err:
logger.error("Failed to restore main config snapshot for %s: %s", plugin_id, restore_err)
if secrets_snapshot is not None:
try:
api_v3.config_manager.save_raw_file_content('secrets', secrets_snapshot)
except Exception as restore_err:
logger.error("Failed to restore secrets snapshot for %s: %s", plugin_id, restore_err)
if reload_plugin and api_v3.plugin_manager is not None:
try:
api_v3.plugin_manager.load_plugin(plugin_id)
except Exception as reload_err:
logger.error("Failed to reload plugin %s during rollback: %s", plugin_id, reload_err)
# --- Step 3: unload ---
if was_loaded:
try:
api_v3.plugin_manager.unload_plugin(plugin_id)
except Exception as unload_err:
_rollback(reload_plugin=False) # unload failed — runtime state unchanged
return False, f"Failed to unload plugin {plugin_id}: {unload_err}"
# --- Step 4: remove files ---
try:
success = api_v3.plugin_store_manager.uninstall_plugin(plugin_id)
except Exception as remove_err:
_rollback(reload_plugin=was_loaded)
return False, f"Failed to remove plugin {plugin_id}: {remove_err}"
if not success:
_rollback(reload_plugin=was_loaded)
return False, f"Failed to uninstall plugin {plugin_id}"
# --- Step 5: finish ---
if api_v3.schema_manager:
api_v3.schema_manager.invalidate_cache(plugin_id)
if api_v3.plugin_state_manager:
api_v3.plugin_state_manager.remove_plugin_state(plugin_id)
# Persistently record the uninstall so a later core `git pull` update
# cannot resurrect a built-in plugin (committed under plugin-repos/) that
# the user removed. Best-effort: never fail the uninstall over this.
try:
api_v3.plugin_store_manager.record_uninstalled_plugin(plugin_id)
except Exception as record_err:
logger.warning("Could not record uninstall for %s: %s", plugin_id, record_err)
return True, None
def deep_merge(base_dict, update_dict):
"""
Deep merge update_dict into base_dict.
For nested dicts, recursively merge. For other types, update_dict takes precedence.
Lists are intentionally REPLACED wholesale, never index-merged: form posts
carry complete arrays, and index-merging would resurrect items the user
deleted. This also applies to the parallel secrets lists produced by
separate_secrets — a newly saved secrets list is authoritative.
"""
result = base_dict.copy()
for key, value in update_dict.items():
if key in result and isinstance(result[key], dict) and isinstance(value, dict):
# Recursively merge nested dicts
result[key] = deep_merge(result[key], value)
else:
# For non-dict values or new keys, use the update value
result[key] = value
return result
def _parse_form_value(value):
"""
Parse a form value into the appropriate Python type.
Handles booleans, numbers, JSON arrays/objects, and strings.
"""
import json
if value is None:
return None
# Handle string values
if isinstance(value, str):
stripped = value.strip()
# Check for boolean strings
if stripped.lower() == 'true':
return True
if stripped.lower() == 'false':
return False
if stripped.lower() in ('null', 'none') or stripped == '':
return None
# Try parsing as JSON (for arrays and objects) - do this BEFORE number parsing
# This handles RGB arrays like "[255, 0, 0]" correctly
if stripped.startswith('[') or stripped.startswith('{'):
try:
return json.loads(stripped)
except json.JSONDecodeError:
pass
# Try parsing as number
try:
if '.' in stripped:
return float(stripped)
return int(stripped)
except ValueError:
pass
# Return as string (original value, not stripped)
return value
return value
def _get_schema_property(schema, key_path):
"""
Get the schema property for a given key path (supports dot notation).
Args:
schema: The JSON schema dict
key_path: Dot-separated path like "customization.time_text.font"
Returns:
The property schema dict or None if not found
"""
if not schema or 'properties' not in schema:
return None
parts = key_path.split('.')
current = schema['properties']
i = 0
while i < len(parts):
# Try progressively longer candidates, longest first, so schema keys that
# themselves contain dots (e.g. league keys like "fifa.world") are matched
# instead of being mistaken for nested "fifa" -> "world" objects.
matched = False
for j in range(len(parts), i, -1):
candidate = '.'.join(parts[i:j])
if isinstance(current, dict) and candidate in current:
prop = current[candidate]
# Consumed all remaining parts — this is the target property.
if j == len(parts):
return prop
# Navigate deeper through an object with properties.
if isinstance(prop, dict) and 'properties' in prop:
current = prop['properties']
i = j
matched = True
break
# Matched a non-object before consuming the path — can't go deeper.
return None
if not matched:
return None
return None
def _is_field_required(key_path, schema):
"""
Check if a field is required according to the schema.
Args:
key_path: Dot-separated path like "mqtt.username"
schema: The JSON schema dict
Returns:
True if field is required, False otherwise
"""
if not schema or 'properties' not in schema:
return False
parts = key_path.split('.')
if len(parts) == 1:
# Top-level field
required = schema.get('required', [])
return parts[0] in required
else:
# Nested field - navigate to parent object
parent_path = '.'.join(parts[:-1])
field_name = parts[-1]
# Get parent property
parent_prop = _get_schema_property(schema, parent_path)
if not parent_prop or 'properties' not in parent_prop:
return False
# Check if field is required in parent
required = parent_prop.get('required', [])
return field_name in required
# Sentinel object to indicate a field should be skipped (not set in config)
_SKIP_FIELD = object()
def _schema_type_is(prop, wanted):
"""Whether a schema property is of ``wanted`` type.
JSON Schema allows a union (``["array", "null"]``), which the per-element
style system uses for its per-mode override fields: null there means
"inherit the base", so the type genuinely is "an array or nothing". A
bare ``prop.get('type') == 'array'`` reads False for those, which meant
the indexed colour inputs a form posts as ``...text_color.0/.1/.2`` were
never recombined into a list.
"""
if not isinstance(prop, dict):
return False
declared = prop.get('type')
if isinstance(declared, list):
return wanted in declared
return declared == wanted
def _schema_allows_null(prop):
"""Whether a schema property's declared type includes null."""
return _schema_type_is(prop, 'null')
def _parse_form_value_with_schema(value, key_path, schema):
"""
Parse a form value using schema information to determine correct type.
Handles arrays (comma-separated strings), objects, and other types.
Args:
value: The form value (usually a string)
key_path: Dot-separated path like "category_order" or "customization.time_text.font"
schema: The plugin's JSON schema
Returns:
Parsed value with correct type, or _SKIP_FIELD to indicate the field should not be set
"""
import json
# Get the schema property for this field
prop = _get_schema_property(schema, key_path)
# Handle None/empty values
if value is None or (isinstance(value, str) and value.strip() == ''):
# A nullable field left blank means null, not an empty container.
# This is the inherit sentinel for per-mode style overrides: an
# empty list there would read as "the user chose no colour" rather
# than "follow the base element".
if _schema_allows_null(prop):
return None
# If schema says it's an array, return empty array instead of None
if prop and _schema_type_is(prop, 'array'):
return []
# If schema says it's an object, return empty dict instead of None
if prop and _schema_type_is(prop, 'object'):
return {}
# If it's an optional string field, preserve empty string instead of None
if prop and prop.get('type') == 'string':
if not _is_field_required(key_path, schema):
return "" # Return empty string for optional string fields
# For number/integer fields, check if they have defaults or are required
if prop:
prop_type = prop.get('type')
if prop_type in ('number', 'integer'):
# If field has a default, use it
if 'default' in prop:
return prop['default']
# If field is not required and has no default, skip setting it
if not _is_field_required(key_path, schema):
return _SKIP_FIELD
# If field is required but empty, return None (validation will fail, which is correct)
return None
return None
# Handle string values
if isinstance(value, str):
stripped = value.strip()
# Check for boolean strings
if stripped.lower() == 'true':
return True
if stripped.lower() == 'false':
return False
# "on"/"off" come from HTML checkboxes — only coerce when schema says boolean
if prop and prop.get('type') == 'boolean':
if stripped.lower() == 'on':
return True
if stripped.lower() == 'off':
return False
# Handle arrays based on schema
if prop and _schema_type_is(prop, 'array'):
# Try parsing as JSON first (handles "[1,2,3]" format)
if stripped.startswith('['):
try:
return json.loads(stripped)
except json.JSONDecodeError:
pass
# Otherwise, treat as comma-separated string
if stripped:
# Split by comma and strip each item
items = [item.strip() for item in stripped.split(',') if item.strip()]
# Try to convert items to numbers if schema items are numbers
items_schema = prop.get('items', {})
if items_schema.get('type') in ('number', 'integer'):
try:
return [int(item) if '.' not in item else float(item) for item in items]
except ValueError:
pass
return items
return []
# Handle objects based on schema
if prop and _schema_type_is(prop, 'object'):
# Try parsing as JSON
if stripped.startswith('{'):
try:
return json.loads(stripped)
except json.JSONDecodeError:
pass
# If it's not JSON, return empty dict (form shouldn't send objects as strings)
return {}
# Try parsing as JSON (for arrays and objects) - do this BEFORE number parsing
if stripped.startswith('[') or stripped.startswith('{'):
try:
return json.loads(stripped)
except json.JSONDecodeError:
pass
# Handle numbers based on schema
if prop:
prop_type = prop.get('type')
if prop_type == 'integer':
try:
return int(stripped)
except ValueError:
return prop.get('default', 0)
elif prop_type == 'number':
try:
return float(stripped)
except ValueError:
return prop.get('default', 0.0)
# Try parsing as number (fallback) — skip when schema explicitly says string
if not (prop and prop.get('type') == 'string'):
try:
if '.' in stripped:
return float(stripped)
return int(stripped)
except ValueError:
pass
# Return as string
return value
return value
def _resolve_key_segments(key_path, config):
"""Split a dot-notation path into segments, greedily preserving keys that
themselves contain dots (e.g. league keys like "fifa.world").
At each level the longest candidate that matches a key already present in the
config wins; otherwise the path splits on the next dot (the normal
nested-create case). Because dotted keys such as ``leagues."fifa.world"``
always exist in the saved config being updated, this routes the value to the
real league object instead of fabricating a ``leagues.fifa.world`` tree.
"""
parts = key_path.split('.')
segments = []
node = config
i = 0
while i < len(parts):
matched = False
if isinstance(node, dict):
for j in range(len(parts), i, -1):
candidate = '.'.join(parts[i:j])
if candidate in node:
segments.append(candidate)
node = node[candidate]
i = j
matched = True
break
if not matched:
part = parts[i]
segments.append(part)
node = node.get(part) if isinstance(node, dict) else None
i += 1
return segments
def _set_nested_value(config, key_path, value):
"""
Set a value in a nested dict using dot notation path.
Handles existing nested dicts correctly by merging instead of replacing.
Keys containing dots (e.g. league keys like "fifa.world") are preserved when
they already exist in the config rather than being split into nested objects.
Args:
config: The config dict to modify
key_path: Dot-separated path (e.g., "customization.period_text.font")
value: The value to set (or _SKIP_FIELD to skip setting)
"""
# Skip setting if value is the sentinel
if value is _SKIP_FIELD:
return
segments = _resolve_key_segments(key_path, config)
current = config
# Navigate/create intermediate dicts
for seg in segments[:-1]:
if seg not in current:
current[seg] = {}
elif not isinstance(current[seg], dict):
# If the existing value is not a dict, replace it with a dict
current[seg] = {}
current = current[seg]
# Set the final value. _SKIP_FIELD (checked above) is the only sentinel
# for "leave the existing value alone" -- an explicit None here is a real
# value (e.g. the per-mode "inherit the base" override) and must overwrite
# whatever was already stored.
current[segments[-1]] = value
#: Hidden field the rendered plugin form repeats once per top-level section it
#: drew. Named with a leading underscore pair so the save path can drop it (and
#: anything else meta) before treating form keys as config paths.
_RENDERED_SECTION_FIELD = '__rendered_section'
def _boolean_is_in_scope(full_path, prefix, sections, submitted_parents):
"""Whether a missing checkbox at ``full_path`` may be forced to False.
"Missing" only means "unchecked" for a form that actually rendered the
control. A caller that posts a handful of fields -- a script, the MQTT
bridge, a curl against the documented endpoint -- never rendered anything,
and reading its silence as "every other checkbox is off" turns a one-field
save into a mass disable. That is not hypothetical: a partial post of four
``customization.*`` keys switched off ``nfl.enabled``, ``ncaa_fb.enabled``
and every display-mode toggle on a live device.
Two ways to be in scope:
* ``sections`` -- the rendered form lists the top-level sections it drew
(``__rendered_section``). Anything it drew is fair game, including a
section whose only fields are checkboxes that are all unchecked, which
is the case no heuristic can recover.
* ``submitted_parents`` -- no marker, so fall back to evidence: the
containing object must have had at least one field posted.
"""
if sections is not None:
return full_path.split('.', 1)[0] in sections
if submitted_parents is None:
return True
return prefix in submitted_parents
def _submitted_parents(form_keys):
"""The object paths a form actually posted a field from ('' = top level)."""
parents = set()
for key in form_keys:
parents.add(key.rsplit('.', 1)[0] if '.' in key else '')
return parents
def _set_missing_booleans_to_false(plugin_config, schema_props, form_keys, prefix='', config_node=None,
sections=None, submitted_parents=None):
"""Walk schema and set missing boolean form fields to False.
HTML checkboxes don't submit values when unchecked. When saving plugin plugin_config,
the backend starts from existing plugin_config (to support partial form updates), which
means an unchecked checkbox's old ``True`` value persists. This function detects
boolean schema properties not present in the form submission and explicitly sets
them to ``False`` -- but only where that silence is evidence, see
:func:`_boolean_is_in_scope`.
The top-level ``enabled`` field is excluded because it has its own preservation
logic in the save endpoint.
Handles boolean fields inside nested objects and inside arrays of objects
(e.g. ``feeds.custom_feeds.0.enabled``).
Args:
plugin_config: The root plugin plugin_config dict (used for pure-dict paths)
schema_props: Schema ``properties`` dict at the current nesting level
form_keys: Set of form field names that were submitted
prefix: Dot-notation prefix for the current nesting level
config_node: The current plugin_config subtree when inside an array item (avoids
using _set_nested_value which corrupts lists)
sections: Top-level sections the form reported rendering, or None when it
reported none (then submitted_parents decides)
submitted_parents: Object paths with at least one posted field; computed
on the first call when there are no section markers
"""
if sections is None and submitted_parents is None:
submitted_parents = _submitted_parents(form_keys)
# Determine which plugin_config node to operate on
node = config_node if config_node is not None else plugin_config
for prop_name, prop_schema in schema_props.items():
if not isinstance(prop_schema, dict):
continue
full_path = f"{prefix}.{prop_name}" if prefix else prop_name
prop_type = prop_schema.get('type')
if prop_type == 'boolean' and full_path != 'enabled':
# If this boolean wasn't submitted in the form, it's an unchecked
# checkbox -- provided the form drew it at all.
if (full_path not in form_keys
and _boolean_is_in_scope(full_path, prefix, sections,
submitted_parents)):
if config_node is not None:
# Inside an array item — set directly on the item dict
node[prop_name] = False
else:
# Pure dict path — use helper
_set_nested_value(plugin_config, full_path, False)
elif prop_type == 'object' and 'properties' in prop_schema:
# Recurse into nested objects
if config_node is not None:
# Inside an array item — ensure nested dict exists in item
if prop_name not in node or not isinstance(node[prop_name], dict):
node[prop_name] = {}
_set_missing_booleans_to_false(
plugin_config, prop_schema['properties'], form_keys, full_path,
config_node=node[prop_name],
sections=sections, submitted_parents=submitted_parents
)
else:
_set_missing_booleans_to_false(
plugin_config, prop_schema['properties'], form_keys, full_path,
sections=sections, submitted_parents=submitted_parents
)
elif prop_type == 'array':
# Handle arrays of objects that may contain boolean fields
# Form keys use indexed notation: "path.0.field", "path.1.field"
items_schema = prop_schema.get('items', {})
if isinstance(items_schema, dict) and items_schema.get('type') == 'object' and 'properties' in items_schema:
array_prefix = f"{full_path}."
# Collect unique item indices from submitted form keys
indices = set()
for k in form_keys:
if k.startswith(array_prefix):
# Extract index: "path.0.field" -> "0"
rest = k[len(array_prefix):]
idx = rest.split('.', 1)[0]
if idx.isdigit():
indices.add(int(idx))
if not indices:
continue
# Navigate to the array in the plugin_config (create if missing)
if config_node is not None:
if prop_name not in node or not isinstance(node[prop_name], list):
node[prop_name] = []
array_list = node[prop_name]
else:
# Navigate from root plugin_config through dict keys to get the list
parts = full_path.split('.')
current = plugin_config
for part in parts[:-1]:
if part not in current or not isinstance(current[part], dict):
current[part] = {}
current = current[part]
arr_key = parts[-1]
if arr_key not in current or not isinstance(current[arr_key], list):
current[arr_key] = []
array_list = current[arr_key]
# Recurse into each array item so its missing booleans get set to False
for idx in indices:
# Ensure list is long enough and item is a dict
while len(array_list) <= idx:
array_list.append({})
if not isinstance(array_list[idx], dict):
array_list[idx] = {}
item_prefix = f"{full_path}.{idx}"
_set_missing_booleans_to_false(
plugin_config, items_schema['properties'], form_keys, item_prefix,
config_node=array_list[idx],
sections=sections, submitted_parents=submitted_parents
)
def _enhance_schema_with_core_properties(schema):
"""
Enhance schema with core plugin properties (enabled, display_duration, live_priority).
These properties are system-managed and should always be allowed even if not in the plugin's schema.
Args:
schema: The original JSON schema dict
Returns:
Enhanced schema dict with core properties injected
"""
import copy
if not schema:
return schema
# Core plugin properties that should always be allowed
# These match the definitions in SchemaManager.validate_config_against_schema()
core_properties = {
"enabled": {
"type": "boolean",
"default": True,
"description": "Enable or disable this plugin"
},
"display_duration": {
"type": "number",
"default": 15,
"minimum": 1,
"maximum": 300,
"description": "How long to display this plugin in seconds"
},
"live_priority": {
"type": "boolean",
"default": False,
"description": "Enable live priority takeover when plugin has live content"
}
}
# Create a deep copy of the schema to modify (to avoid mutating the original)
enhanced_schema = copy.deepcopy(schema)
if "properties" not in enhanced_schema:
enhanced_schema["properties"] = {}
# Inject core properties if they're not already defined in the schema
for prop_name, prop_def in core_properties.items():
if prop_name not in enhanced_schema["properties"]:
enhanced_schema["properties"][prop_name] = copy.deepcopy(prop_def)
return enhanced_schema
def _filter_config_by_schema(config, schema, prefix=''):
"""
Filter config to only include fields defined in the schema.
Removes fields not in schema, especially important when additionalProperties is false.
Args:
config: The config dict to filter
schema: The JSON schema dict
prefix: Prefix for nested paths (used recursively)
Returns:
Filtered config dict containing only schema-defined fields
"""
if not schema or 'properties' not in schema:
return config
filtered = {}
schema_props = schema.get('properties', {})
for key, value in config.items():
if key not in schema_props:
# Field not in schema, skip it
continue
prop_schema = schema_props[key]
# Handle nested objects recursively
if isinstance(value, dict) and prop_schema.get('type') == 'object' and 'properties' in prop_schema:
filtered[key] = _filter_config_by_schema(value, prop_schema, f"{prefix}.{key}" if prefix else key)
else:
# Keep the value as-is for non-object types
filtered[key] = value
return filtered
_MAX_CREDENTIAL_BACKUPS = 5
def _prune_credential_backups(plugin_dir: Path) -> None:
"""Keep only the newest _MAX_CREDENTIAL_BACKUPS credential backups.
Every re-upload copies the previous credentials.json aside. Without
pruning those accumulate for the life of the install — each one a
complete set of OAuth client credentials sitting in the plugin
directory.
"""
backups = sorted(
plugin_dir.glob('credentials.json.backup.*'),
key=lambda p: p.stat().st_mtime,
reverse=True,
)
for stale in backups[_MAX_CREDENTIAL_BACKUPS:]:
try:
stale.unlink()
except OSError:
logger.warning("Could not remove old credential backup %s", stale.name)
# calendarList.list pages at 250 entries maximum. Ten pages is far past any
# real account and exists only so a malformed nextPageToken cannot spin here.
_CALENDAR_LIST_MAX_PAGES = 10
def _calendar_plugin_dir() -> Optional[Path]:
"""Where the calendar plugin is installed, or None if it is not."""
if api_v3.plugin_manager:
plugin_dir = api_v3.plugin_manager.get_plugin_directory('calendar')
else:
plugin_dir = PROJECT_ROOT / 'plugins' / 'calendar'
if not plugin_dir:
return None
plugin_dir = Path(plugin_dir)
return plugin_dir if plugin_dir.exists() else None
def _run_calendar_registration(plugin_dir: Path, stdin_payload: str):
"""Run the plugin's OAuth script and return the JSON object it prints.
The script decides between web and terminal mode by whether stdin is a
tty, so it must be given a pipe. It emits one JSON object on stdout; the
last parsable line is taken, because an import warning or a library's
stderr redirection can land in front of it.
Returns (payload, error_message). Exactly one is None.
"""
script = plugin_dir / 'calendar_registration.py'
if not script.exists():
return None, 'Authentication script not found in the calendar plugin'
try:
result = subprocess.run( # nosec B603 - fixed script path inside the plugin dir
[sys.executable, str(script)],
input=stdin_payload,
capture_output=True,
text=True,
timeout=120,
cwd=str(plugin_dir),
)
except subprocess.TimeoutExpired:
return None, 'Authentication timed out after 120s'
except OSError as e:
logger.error('Could not run calendar_registration.py', exc_info=True)
return None, 'Could not run the authentication script: %s' % describe_exception(e)
for line in reversed((result.stdout or '').splitlines()):
line = line.strip()
if not line:
continue
try:
payload = json.loads(line)
except json.JSONDecodeError:
continue
if isinstance(payload, dict):
return payload, None
raw = (result.stderr or result.stdout or '').strip()
# Redacted in the log too: this is a script that handles OAuth client
# secrets, and its stderr can quote them verbatim (CWE-532).
if raw:
logger.error('calendar_registration.py failed (exit %s): %s',
result.returncode, redact_text(raw))
return None, 'Authentication script produced no result%s' % (
': %s' % redact_text(raw) if raw else '')
def _resolve_backup_export_dir() -> Path:
"""Where exported backups live: beside the install, not inside it.
They used to be written to ``<project>/config/backups/exports``. That is
inside the directory a reinstall deletes, so the documented recovery path
-- export a backup, then reinstall -- destroyed the backup it had just
told the user to make. Anyone who downloaded the ZIP was fine; anyone
relying on the on-device copy was not.
Falls back to the old location when the parent directory is not writable,
so an unusual layout degrades to previous behaviour instead of failing to
export at all.
"""
preferred = PROJECT_ROOT.parent / "ledmatrix-backups"
fallback = PROJECT_ROOT / "config" / "backups" / "exports"
try:
preferred.mkdir(parents=True, exist_ok=True)
with tempfile.NamedTemporaryFile(dir=preferred, prefix=".writetest-"):
pass
return preferred
except OSError as e:
logger.warning(
f"[Backup] Export dir {preferred} is not writable ({e}); "
f"falling back to {fallback}, which a reinstall will delete"
)
return fallback
_BACKUP_EXPORT_DIR = _resolve_backup_export_dir()
def _safe_backup_path(filename: str) -> Path:
"""Resolve a filename to an absolute path inside the export dir,
rejecting any traversal attempts. Returns None if unsafe."""
# Use basename first (CodeQL-recognized sanitizer) then validate format
filename = os.path.basename(filename or '')
if not filename or not re.match(r'^[a-zA-Z0-9][a-zA-Z0-9._-]{0,200}\.zip$', filename):
return None
path = (_BACKUP_EXPORT_DIR / filename).resolve()
try:
path.relative_to(_BACKUP_EXPORT_DIR.resolve())
except ValueError:
return None
return path
_STARLARK_APPS_DIR = PROJECT_ROOT / 'starlark-apps'
_STARLARK_MANIFEST_FILE = _STARLARK_APPS_DIR / 'manifest.json'
# A dedicated, never-replaced file to flock -- see _starlark_manifest_lock.
_STARLARK_MANIFEST_LOCK_FILE = _STARLARK_APPS_DIR / 'manifest.json.lock'
def _get_starlark_plugin() -> Optional[Any]:
"""Get the starlark-apps plugin instance, or None."""
if not api_v3.plugin_manager:
return None
return api_v3.plugin_manager.get_plugin('starlark-apps')
def _find_pixlet_binary(explicit_path: Optional[str] = None) -> Optional[str]:
"""Find pixlet binary: explicit path → bundled binary → system PATH."""
import platform
if explicit_path and os.path.isfile(explicit_path) and os.access(explicit_path, os.X_OK):
return explicit_path
bin_dir = PROJECT_ROOT / "bin" / "pixlet"
system = platform.system().lower()
machine = platform.machine().lower()
if system == "linux":
if "aarch64" in machine or "arm64" in machine:
name = "pixlet-linux-arm64"
elif "x86_64" in machine or "amd64" in machine:
name = "pixlet-linux-amd64"
else:
name = None
elif system == "darwin":
name = "pixlet-darwin-arm64" if "arm64" in machine else "pixlet-darwin-amd64"
else:
name = None
if name:
bundled = bin_dir / name
if bundled.is_file():
if os.access(str(bundled), os.X_OK):
return str(bundled)
try:
bundled.chmod(0o755)
except OSError:
logger.warning("Could not make pixlet bundled binary executable (%s); falling back to PATH", bundled)
else:
if os.access(str(bundled), os.X_OK):
return str(bundled)
logger.warning("Pixlet bundled binary still not executable after chmod (%s); falling back to PATH", bundled)
return shutil.which("pixlet")
@contextlib.contextmanager
def _starlark_manifest_lock():
"""Hold an exclusive lock across a standalone starlark-manifest read-modify-write.
StarlarkAppsPlugin._update_manifest_safe (plugin-repos/starlark-apps/manager.py)
already does this -- fcntl.flock held for the whole read-modify-write cycle --
when the plugin instance is loaded. These routes fall back to reading and
writing manifest.json directly when it is not, and did so with no lock: each
_write_starlark_manifest() call is atomic on its own (temp file + rename), but
two concurrent requests can each read the manifest, mutate their own copy, and
write it back, and the second write silently discards the first's change.
Locks _STARLARK_MANIFEST_LOCK_FILE, a sidecar that is never written to or
renamed over -- not manifest.json itself. manifest.json is replaced by an
atomic rename on every write (here and in the plugin), which swaps in a
fresh inode; a lock held on the old inode does not exclude a second locker
that opens the path afresh right after the rename and gets the new inode,
so two writers could still race each other despite both "holding a lock".
A stable sidecar path always resolves to the same inode, so every locker
-- standalone or plugin-owned -- contends for the same lock. The plugin
must lock this same sidecar file for that guarantee to hold across both.
Callers should do their read, mutation and _write_starlark_manifest() call
entirely inside the `with` block, mirroring the plugin's lock scope.
"""
# Imported here, not at module scope: fcntl is POSIX-only, and a top-level
# import made the whole api_v3 package unimportable on Windows -- which the
# monolithic blueprint never was, so the split would have broken local dev
# and the test suite there. Deliberately NOT degraded to a no-op lock off
# POSIX: the docstring above describes a real lost-update race, and silently
# not locking would be worse than failing loudly on a platform that cannot
# run the display anyway.
import fcntl
_STARLARK_APPS_DIR.mkdir(parents=True, exist_ok=True)
lock_fd = os.open(str(_STARLARK_MANIFEST_LOCK_FILE), os.O_CREAT | os.O_RDWR, 0o644)
try:
fcntl.flock(lock_fd, fcntl.LOCK_EX)
try:
yield
finally:
fcntl.flock(lock_fd, fcntl.LOCK_UN)
finally:
os.close(lock_fd)
def _read_starlark_manifest() -> Dict[str, Any]:
"""Read the starlark-apps manifest.json directly from disk."""
try:
if _STARLARK_MANIFEST_FILE.exists():
with open(_STARLARK_MANIFEST_FILE, 'r') as f:
return json.load(f)
except (json.JSONDecodeError, OSError) as e:
logger.error(f"Error reading starlark manifest: {e}")
return {'apps': {}}
def _starlark_github_token() -> Optional[str]:
"""The GitHub token the Starlark store should authenticate with.
These routes used to read `github_token` off config.json, a key that is
written nowhere and offered by no setting -- so the store always ran
unauthenticated at 60 requests/hour, on the same per-IP budget every
plugin update check spends, while the token the user had actually
configured sat in config_secrets.json raising the same budget to 5000.
The store going blank was that budget running out.
Prefer the store manager's token, which is the one the settings UI
writes and validates; keep the config.json key as a fallback so a
hand-edited config still works.
"""
token = getattr(api_v3.plugin_store_manager, 'github_token', None)
if token:
return token
try:
config = api_v3.config_manager.load_config() if api_v3.config_manager else {}
return config.get('github_token')
except Exception:
logger.warning("[Starlark] Could not read config for a GitHub token", exc_info=True)
return None
def _get_tronbyte_repository_class() -> Type[Any]:
"""Import TronbyteRepository from plugin-repos directory."""
import importlib.util
import importlib
module_path = PROJECT_ROOT / 'plugin-repos' / 'starlark-apps' / 'tronbyte_repository.py'
if not module_path.exists():
raise ImportError(f"TronbyteRepository module not found at {module_path}")
# If already imported, return cached class
if "tronbyte_repository" in sys.modules:
return sys.modules["tronbyte_repository"].TronbyteRepository
spec = importlib.util.spec_from_file_location("tronbyte_repository", str(module_path))
if spec is None:
raise ImportError(f"Failed to create module spec for tronbyte_repository at {module_path}")
module = importlib.util.module_from_spec(spec)
if module is None:
raise ImportError("Failed to create module from spec for tronbyte_repository")
sys.modules["tronbyte_repository"] = module
try:
spec.loader.exec_module(module)
except BaseException:
# A module that failed to execute must not stay in sys.modules: the
# cache branch above would hand back the half-initialised object for
# the rest of the process, so one transient failure would disable
# this path permanently and surface as AttributeError, not ImportError.
sys.modules.pop("tronbyte_repository", None)
raise
return module.TronbyteRepository
def _get_pixlet_renderer_class() -> Type[Any]:
"""Import PixletRenderer from plugin-repos directory."""
import importlib.util
import importlib
module_path = PROJECT_ROOT / 'plugin-repos' / 'starlark-apps' / 'pixlet_renderer.py'
if not module_path.exists():
raise ImportError(f"PixletRenderer module not found at {module_path}")
# If already imported, return cached class
if "pixlet_renderer" in sys.modules:
return sys.modules["pixlet_renderer"].PixletRenderer
spec = importlib.util.spec_from_file_location("pixlet_renderer", str(module_path))
if spec is None:
raise ImportError(f"Failed to create module spec for pixlet_renderer at {module_path}")
module = importlib.util.module_from_spec(spec)
if module is None:
raise ImportError("Failed to create module from spec for pixlet_renderer")
sys.modules["pixlet_renderer"] = module
try:
spec.loader.exec_module(module)
except BaseException:
# A module that failed to execute must not stay in sys.modules: the
# cache branch above would hand back the half-initialised object for
# the rest of the process, so one transient failure would disable
# this path permanently and surface as AttributeError, not ImportError.
sys.modules.pop("pixlet_renderer", None)
raise
return module.PixletRenderer
def _validate_and_sanitize_app_id(app_id: Optional[str], fallback_source: Optional[str] = None) -> Tuple[Optional[str], Optional[str]]:
"""Validate and sanitize app_id to a safe slug."""
if not app_id and fallback_source:
app_id = fallback_source
if not app_id:
return None, "app_id is required"
if '..' in app_id or '/' in app_id or '\\' in app_id:
return None, "app_id contains invalid characters"
sanitized = re.sub(r'[^a-z0-9_]', '_', app_id.lower()).strip('_')
if not sanitized:
sanitized = f"app_{hashlib.sha256(app_id.encode()).hexdigest()[:12]}"
if sanitized[0].isdigit():
sanitized = f"app_{sanitized}"
return sanitized, None
def _validate_timing_value(value: Any, field_name: str, min_val: int = 1, max_val: int = 86400) -> Tuple[Optional[int], Optional[str]]:
"""Validate and coerce timing values."""
if value is None:
return None, None
try:
int_value = int(value)
except (ValueError, TypeError):
return None, f"{field_name} must be an integer"
if int_value < min_val:
return None, f"{field_name} must be at least {min_val}"
if int_value > max_val:
return None, f"{field_name} must be at most {max_val}"
return int_value, None
def _validate_starlark_app_path(app_id: str) -> Tuple[Optional[Path], Optional[str]]:
"""The app's directory, or an error if app_id could escape the base dir.
Returns the *resolved* path rather than a boolean, and every caller uses
what it returns instead of re-joining ``_STARLARK_APPS_DIR / app_id``
afterwards. The old shape validated in one place and rebuilt the path in
another, which is two things that have to stay in step -- and is why
CodeQL reported twenty-four path-injection alerts across these handlers
even though the guard was effective: a boolean is not a sanitiser it can
follow, and the value reaching the filesystem was the raw one.
The name is unchanged so the call sites read the same.
"""
if not isinstance(app_id, str) or not app_id:
return None, "Invalid app_id"
# Reject the traversal characters outright before touching the filesystem.
if '..' in app_id or '/' in app_id or '\\' in app_id:
return None, "Invalid app_id: contains path traversal characters"
# os.path.basename strips any directory component, so what is joined below
# cannot carry one. The equality check means this rejects rather than
# silently truncates -- behaviour is identical to the character test above,
# and it is the sanitiser CodeQL's path-injection query actually follows.
# relative_to() alone is a check it cannot trace, which is why twenty-four
# of these stayed flagged after the value was threaded through properly.
safe_name = os.path.basename(app_id)
if safe_name != app_id or safe_name in ('', '.', '..'):
return None, "Invalid app_id: contains path traversal characters"
try:
base_path = _STARLARK_APPS_DIR.resolve()
app_path = (base_path / safe_name).resolve()
try:
app_path.relative_to(base_path)
except ValueError:
return None, "Invalid app_id: path traversal attempt"
return app_path, None
except OSError as e:
logger.warning("Path validation error for app_id %r: %s", app_id, e)
return None, "Invalid app_id"
def _standalone_render_starlark_app(app_id: str) -> Tuple[bool, int, Optional[str]]:
"""Render a Starlark app via pixlet directly (no plugin required).
Reads the .star file and config from starlark-apps/{app_id}/, runs pixlet,
and saves the output to cached_render.webp in the same directory.
This is the web-service fallback when starlark-apps plugin is not loaded.
Returns (success, http_status_code, error_message).
"""
manifest = _read_starlark_manifest()
if not isinstance(manifest, dict):
return False, 400, "Invalid manifest shape: expected object with 'apps' mapping"
apps = manifest.get('apps', {})
if not isinstance(apps, dict):
return False, 400, "Invalid manifest shape: expected object with 'apps' mapping"
app_data = apps.get(app_id)
if not app_data:
return False, 404, f"App not found: {app_id}"
# Validated here as well as at the handler: this is reachable on its own,
# and a path built from app_id should never be assembled without it.
app_dir, path_error = _validate_starlark_app_path(app_id)
if path_error:
return False, 400, path_error
star_file = app_dir / app_data.get('star_file', f'{app_id}.star')
if not star_file.exists():
return False, 404, f"Star file not found: {star_file}"
full_config = api_v3.config_manager.load_config() if api_v3.config_manager else {}
plugin_config = full_config.get('starlark-apps', {})
pixlet_path = _find_pixlet_binary(plugin_config.get('pixlet_path'))
if not pixlet_path:
return False, 503, "Pixlet binary not found — install pixlet first"
magnify = plugin_config.get('magnify')
if magnify is None:
hw = full_config.get('display', {}).get('hardware', {})
cols = hw.get('cols', 64)
chain = hw.get('chain_length', 1)
rows = hw.get('rows', 32)
magnify = max(1, min(8, int(min((cols * chain) / 64, rows / 32))))
else:
try:
magnify = max(1, min(8, int(magnify)))
except (ValueError, TypeError):
magnify = 1
config_file = app_dir / 'config.json'
app_config: Dict[str, Any] = {}
if config_file.exists():
try:
with open(config_file) as f:
app_config = json.load(f)
except json.JSONDecodeError as e:
logger.warning("Invalid config.json for %r at %s: %s", app_id, config_file, e)
return False, 400, f"Invalid config.json for {app_id}"
except OSError as e:
logger.warning("Cannot read config.json for %r at %s: %s", app_id, config_file, e)
return False, 400, f"Cannot read config.json for {app_id}"
if not isinstance(app_config, dict):
return False, 400, (
f"config.json for {app_id} must be a JSON object, "
f"got {type(app_config).__name__}"
)
INTERNAL_KEYS = {'render_interval', 'display_duration'}
pixlet_config = {k: v for k, v in app_config.items() if k not in INTERNAL_KEYS}
output_path = str(app_dir / 'cached_render.webp')
cmd = [pixlet_path, 'render', str(star_file)]
for key, value in pixlet_config.items():
if not re.match(r'^[a-zA-Z_][a-zA-Z0-9_]*$', key):
continue
value_str = 'true' if value is True else 'false' if value is False else str(value)
if re.search(r'[`$|<>&;\x00]|\$\(', value_str):
continue
cmd.append(f'{key}={value_str}')
cmd.extend(['-o', output_path, '-m', str(magnify)])
try:
result = subprocess.run(cmd, capture_output=True, text=True, timeout=30, cwd=str(app_dir))
if result.returncode == 0 and os.path.isfile(output_path):
return True, 200, None
return False, 502, f"Pixlet failed (exit {result.returncode}): {result.stderr.strip()}"
except subprocess.TimeoutExpired:
return False, 504, "Render timed out after 30s"
except Exception as e:
logger.exception("Starlark render failed for %r", app_id)
return False, 500, "Render error"
def _write_starlark_manifest(manifest: Dict[str, Any]) -> bool:
"""Write the starlark-apps manifest.json to disk with atomic write."""
temp_file = None
try:
_STARLARK_APPS_DIR.mkdir(parents=True, exist_ok=True)
# Atomic write: unique temp file in the target directory, then rename.
# with_suffix('.tmp') gave every caller the same manifest.tmp, and
# Flask serves concurrently -- upload, uninstall, config, toggle and
# the plugin toggle all reach here. Two writers shared one file,
# interleaved their json.dump output, and both renamed it, so the
# rename was atomic over content that was a mix of two manifests.
fd, temp_name = tempfile.mkstemp(
dir=str(_STARLARK_APPS_DIR), prefix='.manifest.', suffix='.tmp')
temp_file = Path(temp_name)
with os.fdopen(fd, 'w') as f:
json.dump(manifest, f, indent=2)
f.flush()
os.fsync(f.fileno()) # Ensure data is written to disk
os.chmod(temp_name, 0o644) # mkstemp creates 0600; match a normal write
# Atomic rename (overwrites destination)
temp_file.replace(_STARLARK_MANIFEST_FILE)
return True
except OSError as e:
logger.error(f"Error writing starlark manifest: {e}")
# Clean up temp file if it exists
if temp_file and temp_file.exists():
try:
temp_file.unlink()
except Exception:
pass
return False
def _install_star_file(app_id: str, star_file_path: str, metadata: Dict[str, Any], assets_dir: Optional[str] = None) -> bool:
"""Install a .star file and update the manifest (standalone, no plugin needed)."""
import shutil
import json
app_dir, path_error = _validate_starlark_app_path(app_id)
if path_error:
logger.warning("Refusing to install %r: %s", app_id, path_error)
return False
app_dir.mkdir(parents=True, exist_ok=True)
dest = app_dir / f"{app_id}.star"
shutil.copy2(star_file_path, str(dest))
# Copy asset directories if provided (images/, sources/, etc.)
if assets_dir and Path(assets_dir).exists():
assets_path = Path(assets_dir)
for item in assets_path.iterdir():
if item.is_dir():
# Copy entire directory (e.g., images/, sources/)
dest_dir = app_dir / item.name
if dest_dir.exists():
shutil.rmtree(dest_dir)
shutil.copytree(item, dest_dir)
logger.debug(f"Copied assets directory: {item.name}")
logger.info(f"Installed assets for {app_id}")
# Try to extract schema using PixletRenderer
schema = None
try:
PixletRenderer = _get_pixlet_renderer_class()
pixlet = PixletRenderer()
if pixlet.is_available():
_, schema, _ = pixlet.extract_schema(str(dest))
if schema:
schema_path = app_dir / "schema.json"
with open(schema_path, 'w') as f:
json.dump(schema, f, indent=2)
logger.info(f"Extracted schema for {app_id}")
except Exception as e:
logger.warning(f"Failed to extract schema for {app_id}: {e}")
# Create default config — pre-populate with schema defaults
default_config = {}
if schema:
fields = schema.get('fields') or schema.get('schema') or []
for field in fields:
if isinstance(field, dict) and 'id' in field and 'default' in field:
default_config[field['id']] = field['default']
# Create config.json file
config_path = app_dir / "config.json"
with open(config_path, 'w') as f:
json.dump(default_config, f, indent=2)
with _starlark_manifest_lock():
manifest = _read_starlark_manifest()
manifest.setdefault('apps', {})[app_id] = {
'name': metadata.get('name', app_id),
'enabled': True,
'render_interval': metadata.get('render_interval', 300),
'display_duration': metadata.get('display_duration', 15),
'config': metadata.get('config', {}),
# The filename, not the full path. Readers join this to the app's own
# directory and fall back to a bare '<app_id>.star', so an absolute
# value gave the key two meanings -- and Path.__truediv__ discards the
# left side when the right is absolute, which pinned the manifest to
# whatever PROJECT_ROOT installed it. Moving or redeploying the install
# then left the app unable to find its own file.
'star_file': dest.name,
}
return _write_starlark_manifest(manifest)
def _starlark_virtual_plugins() -> list:
"""Installed Starlark apps, shaped like plugin entries.
#253 surfaced these alongside real plugins so an installed .star app can
be seen, enabled and disabled from the same list as everything else; #330
dropped it with the rest of the Starlark code, which is why an app
installs successfully and then appears nowhere.
Reads the loaded plugin when there is one and the on-disk manifest
otherwise, so the list is right before starlark-apps has been loaded too.
"""
entries = []
base = {
'version': 'starlark', 'category': 'Starlark App', 'tags': ['starlark'],
'verified': False, 'last_updated': None, 'last_commit': None,
'last_commit_message': None, 'branch': None, 'web_ui_actions': [],
'vegas_mode': 'fixed', 'vegas_content_type': 'multi',
'is_starlark_app': True,
}
try:
plugin = _get_starlark_plugin()
if plugin is not None and hasattr(plugin, 'apps'):
for app_id, app in plugin.apps.items():
m = getattr(app, 'manifest', {}) or {}
entries.append({**base,
'id': f'starlark:{app_id}',
'name': m.get('name', app_id),
'author': m.get('author', 'Tronbyte Community'),
'description': m.get('summary', 'Starlark app'),
'enabled': app.is_enabled(),
'loaded': True})
return entries
for app_id, data in (_read_starlark_manifest().get('apps', {}) or {}).items():
entries.append({**base,
'id': f'starlark:{app_id}',
'name': data.get('name', app_id),
'author': data.get('author', 'Tronbyte Community'),
'description': data.get('summary', 'Starlark app'),
'enabled': data.get('enabled', True),
'loaded': False})
except Exception:
# Never let a Starlark problem empty the whole plugins list.
logger.exception('Could not build Starlark virtual plugin entries')
return entries
def _toggle_starlark_app(app_id: str, enabled: bool):
"""Enable or disable one Starlark app, loaded or not."""
# Check for traversal, but toggle the key that was listed.
# _starlark_virtual_plugins publishes the raw manifest key, while
# _validate_and_sanitize_app_id lowercases it and rewrites every character
# outside [a-z0-9_]: an app stored as 'My-App' was offered to the UI as
# 'starlark:My-App' and looked up here as 'my_app', so toggling an app the
# page had just drawn answered 404. Keys written by _install_star_file are
# already sanitised; ones written by the plugin, or edited by hand, are
# not. _validate_starlark_app_path rejects traversal without rewriting.
_, err = _validate_starlark_app_path(app_id)
if err:
# err already names app_id; do not prefix it a second time.
return jsonify({'status': 'error', 'message': err}), 400
safe_id = app_id
plugin = _get_starlark_plugin()
if plugin is not None and safe_id in getattr(plugin, 'apps', {}):
def _update(manifest):
# setdefault rather than indexing: the app is loaded, but its
# on-disk entry need not exist, and _update_manifest_safe does not
# catch KeyError -- it would escape as a 500 rather than the error
# this returns.
manifest.setdefault('apps', {}).setdefault(safe_id, {})['enabled'] = enabled
if plugin._update_manifest_safe(_update) is False:
return jsonify({'status': 'error',
'message': 'Failed to save app state'}), 500
# Only now is the in-memory copy allowed to disagree with disk.
plugin.apps[safe_id].manifest['enabled'] = enabled
else:
with _starlark_manifest_lock():
manifest = _read_starlark_manifest()
app_data = manifest.get('apps', {}).get(safe_id)
if not app_data:
return jsonify({'status': 'error',
'message': f'Starlark app not found: {safe_id}'}), 404
app_data['enabled'] = enabled
if not _write_starlark_manifest(manifest):
return jsonify({'status': 'error', 'message': 'Failed to save manifest'}), 500
return jsonify({'status': 'success',
'message': f"Starlark app {'enabled' if enabled else 'disabled'}",
'enabled': enabled})
_PIXLET_EDITOR_SCRIPT = PROJECT_ROOT / 'scripts' / 'utils' / 'pixlet_config_editor.sh'
# Deliberately under /tmp: a session cannot survive a reboot, so neither should
# the record of one.
_PIXLET_EDITOR_STATE = Path(tempfile.gettempdir()) / 'ledmatrix_pixlet_editor.json'
_PIXLET_EDITOR_DEFAULT_PORT = 8080
_PIXLET_EDITOR_DEFAULT_TIMEOUT = 1800
_PIXLET_EDITOR_MAX_TIMEOUT = 14400
def _read_pixlet_editor_state() -> Optional[Dict[str, Any]]:
try:
if _PIXLET_EDITOR_STATE.is_file():
with open(_PIXLET_EDITOR_STATE, encoding='utf-8') as handle:
state = json.load(handle)
return state if isinstance(state, dict) else None
except (OSError, json.JSONDecodeError):
logger.debug('Unreadable pixlet editor state; treating as no session', exc_info=True)
return None
def _clear_pixlet_editor_state() -> None:
with contextlib.suppress(OSError):
_PIXLET_EDITOR_STATE.unlink()
def _pixlet_editor_alive(pid: Optional[int]) -> bool:
"""Is the recorded session still running?
os.kill(pid, 0) is not enough on its own: the script is a child of this
process, so once it exits it stays a zombie until reaped, and signal 0
succeeds against a zombie. Left at that, a finished session would read as
running forever and the UI would keep offering a Stop button for it.
"""
if not pid:
return False
# Reap it if it is ours and already finished; harmless if it is not.
with contextlib.suppress(ChildProcessError, OSError):
reaped, _ = os.waitpid(pid, os.WNOHANG)
if reaped == pid:
return False
try:
os.kill(pid, 0)
except ProcessLookupError:
return False
except PermissionError:
# Exists but is not ours to signal, which still counts as running.
return True
# Not our child (e.g. the web service restarted under a live session), so
# waitpid told us nothing -- ask /proc whether it is merely a zombie.
with contextlib.suppress(OSError, IndexError, ValueError):
with open(f'/proc/{pid}/stat', encoding='utf-8') as handle:
# The comm field can contain spaces and parens; everything after
# the final ')' is positional, and state is the first of those.
fields = handle.read().rsplit(')', 1)[1].split()
if fields and fields[0] == 'Z':
return False
return True
def _pixlet_editor_status() -> Dict[str, Any]:
"""Current session, reconciled against reality.
The state file records what we started; the process may have ended on its
own (its timeout, a crash, a manual kill). Anything stale is cleared here so
the UI never offers a Stop button for a session that is already over.
"""
state = _read_pixlet_editor_state()
if not state:
return {'running': False}
if not _pixlet_editor_alive(state.get('pid')):
_clear_pixlet_editor_state()
return {'running': False}
remaining = None
deadline = state.get('deadline')
if isinstance(deadline, (int, float)):
remaining = max(0, int(deadline - time.time()))
return {
'running': True,
'app_id': state.get('app_id'),
'port': state.get('port', _PIXLET_EDITOR_DEFAULT_PORT),
'pid': state.get('pid'),
'started_at': state.get('started_at'),
'timeout': state.get('timeout'),
'seconds_remaining': remaining,
'host_bound': state.get('host', '0.0.0.0'),
}
_MQTT_BRIDGE_DIR = PROJECT_ROOT / 'integrations' / 'mqtt_bridge'
_MQTT_BRIDGE_CONFIG = _MQTT_BRIDGE_DIR / 'bridge_config.json'
_MQTT_BRIDGE_EXAMPLE = _MQTT_BRIDGE_DIR / 'bridge_config.example.json'
_MQTT_BRIDGE_SERVICE = 'ledmatrix-mqtt-bridge.service'
# Mirrors DEFAULTS in ledmatrix_mqtt_bridge.py. Duplicated rather than imported
# because that module pulls in paho-mqtt, which the web process does not need
# installed just to render a settings form.
_MQTT_BRIDGE_DEFAULTS = {
'mqtt_host': 'localhost',
'mqtt_port': 1883,
'mqtt_username': None,
'mqtt_client_id': 'ledmatrix-mqtt-bridge',
'mqtt_topic': 'ledmatrix/command',
'mqtt_tls': False,
'mqtt_tls_insecure': False,
# Opt-in acknowledgement that credentials may cross an untrusted
# network in cleartext. Off by default: the save is refused instead.
'allow_insecure_mqtt': False,
'ledmatrix_api_base': 'http://localhost:5000',
'request_timeout': 15,
'on_demand_duration': None,
'log_level': 'INFO',
}
_MQTT_BRIDGE_LOG_LEVELS = ('DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL')
def _mqtt_bridge_service_state() -> Dict[str, Any]:
"""installed / active / enabled for the bridge unit."""
state = {'installed': False, 'active': False, 'enabled': False}
try:
listed = _run_systemctl_command(
['systemctl', 'list-unit-files', _MQTT_BRIDGE_SERVICE, '--no-legend'])
state['installed'] = bool((listed.get('stdout') or '').strip())
state['active'] = _run_systemctl_command(
['systemctl', 'is-active', _MQTT_BRIDGE_SERVICE]).get('stdout', '').strip() == 'active'
state['enabled'] = _run_systemctl_command(
['systemctl', 'is-enabled', _MQTT_BRIDGE_SERVICE]).get('stdout', '').strip() == 'enabled'
except Exception:
logger.debug('Could not read %s state', _MQTT_BRIDGE_SERVICE, exc_info=True)
return state
def _read_mqtt_bridge_config() -> Dict[str, Any]:
"""Stored config overlaid on the defaults. Missing file is not an error.
Named `settings`, not `config`: this module imports a submodule called
`config` at the bottom for its route side effects, and a local of the same
name shadows it.
"""
settings = dict(_MQTT_BRIDGE_DEFAULTS)
settings['mqtt_password'] = None
try:
if _MQTT_BRIDGE_CONFIG.is_file():
with open(_MQTT_BRIDGE_CONFIG, encoding='utf-8') as handle:
stored = json.load(handle)
if isinstance(stored, dict):
settings.update(stored)
except (OSError, json.JSONDecodeError) as err:
logger.warning('Could not read %s: %s', _MQTT_BRIDGE_CONFIG, err)
return settings
def _coerce_mqtt_bridge_value(key: str, raw: Any) -> Tuple[Any, Optional[str]]:
"""Validate one submitted field. Returns (value, error)."""
if key in ('mqtt_port',):
try:
port = int(raw)
except (TypeError, ValueError):
return None, 'Port must be a whole number'
if not 1 <= port <= 65535:
return None, 'Port must be between 1 and 65535'
return port, None
if key in ('request_timeout',):
try:
timeout = int(raw)
except (TypeError, ValueError):
return None, 'Request timeout must be a whole number'
if not 1 <= timeout <= 300:
return None, 'Request timeout must be between 1 and 300 seconds'
return timeout, None
if key == 'on_demand_duration':
if raw in (None, ''):
return None, None
try:
duration = int(raw)
except (TypeError, ValueError):
return None, 'On-demand duration must be a whole number of seconds'
if not 1 <= duration <= 86400:
return None, 'On-demand duration must be between 1 and 86400 seconds'
return duration, None
if key in ('mqtt_tls', 'mqtt_tls_insecure', 'allow_insecure_mqtt'):
return bool(raw) if isinstance(raw, bool) else str(raw).lower() in ('1', 'true', 'yes', 'on'), None
if key == 'log_level':
level = str(raw or '').upper()
if level not in _MQTT_BRIDGE_LOG_LEVELS:
return None, f"Log level must be one of {', '.join(_MQTT_BRIDGE_LOG_LEVELS)}"
return level, None
if key == 'ledmatrix_api_base':
base = str(raw or '').strip().rstrip('/')
if not base.startswith(('http://', 'https://')):
return None, 'API base must start with http:// or https://'
if len(base) > 300:
return None, 'API base is too long'
return base, None
# Remaining keys are free text; empty means "unset" for the optional ones.
text = '' if raw is None else str(raw).strip()
if len(text) > 300:
return None, f'{key} is too long'
if key == 'mqtt_username' and not text:
return None, None
if key in ('mqtt_host', 'mqtt_client_id', 'mqtt_topic') and not text:
return None, f'{key.replace("_", " ")} cannot be empty'
return text, None
# Imported last, and for their side effect: each registers its routes on
# api_v3. They import names from this module, so this module must be fully
# executed before they run.
from web_interface.blueprints.api_v3 import ( # noqa: E402,F401
backup,
config,
display,
fonts,
misc,
plugins,
starlark,
system,
wifi,
)
__all__ = ["api_v3"]