| Age | Commit message (Collapse) | Author |
|---|
|
Composer::getLoop() answers, so a plugin reaches the graph's own downloader
and executor through the route Composer's own docblocks point plugin authors
at, rather than through an instance of its own.
wait() drains the promises it is given and rethrows the first rejection once
the group is done, which is what React\Promise\all() hands PHP. abortJobs()
has nothing to cancel while every request settles before the call that
started it returns. A non-null $progress is an explicit error: a ProgressBar
is a symfony/console object each world runs its own implementation of, so
there is none to hand across.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
add() and addCopy() answer with a promise, and enableAsync(), wait() and
countActiveJobs() answer alongside them. The Rust future runs to completion
before the promise is handed over, so requests a plugin starts together run
one after another rather than overlapping; overlapping them needs a promise
representation that crosses the boundary unresolved.
Everything else the surface does is preserved. add() still refuses a
downloader outside a Loop, and it does so by throwing out of the call the way
PHP does, where a failed request instead arrives as a rejection the caller
handles — __shirabe_rejected_promise is the failure half of the resolved-
promise helper. wait() and countActiveJobs() answer for a downloader with no
outstanding job, which, once every request settles before its call returns,
it never has.
This is where HttpDownloader parts company with ProcessExecutor, whose async
surface stays an explicit error: executeAsync() resolves its promise with a
Symfony Process, whose state is the proc_open() resource of whichever process
called start(), where a request resolves its promise with a Response.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
A response is built for one request and the graph never retains it, so there
is no entity for a handle to point at. The child holds a real instance
instead, revived from the object record the wire carries, and collect() frees
the copy each world holds — which is what that method is for. The
value-object rule rejects the class only because collect() assigns to $this,
so the category comes from an overrides.list entry.
HttpDownloader::get() and copy() answer with one.
Two gaps stay: decodeJson() reaches Composer\Json\JsonFile, which a guard
shadows, and Composer answers a curl request with the CurlResponse subclass
where this port flattens the value into a Response.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
A downloader carries its own options, TLS defaults and request backends, and
what it shares with the graph is the IO it collects authentication into and
the config it reads. Plugin code writing `new HttpDownloader($io, $config)`
therefore allocates a Rust-side entity of its own rather than a second
downloader the graph knows nothing about, and the guard that shadowed the
class in the worker is gone.
The request surface is not served yet: get() and copy() have no wire
representation for the Response they return, and the async surface resolves
its promises with one.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
A Rust-side failure reached plugin code as a RuntimeException whose message
carried the name of the call that failed, so `catch (TransportException $e)`
never matched and the status code the plugin branches on was gone.
The Throw frame now names the class the exception was thrown as and carries
the state that class declares beyond message and code.
\Shirabe\MaterializedThrowable rebuilds it in the child: `new $class($message,
$code)` for a class whose constructor has \Exception's shape, then the
properties by reflection. A class the child cannot build that way keeps the
RuntimeException shape.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
Composer reaches this class two ways: the object graph hands one out through
Composer::getLoop()->getProcessExecutor(), and plugins write
`new ProcessExecutor($io)` freely. Both bind to a Rust-side entity, so the
timeout the run shares -- seeded from process-timeout and rewritten while the
run is in flight -- has one value instead of one per world, and the executor
can still be passed to the classes that take one (`new Filesystem($process)`).
Three things the stub generator was missing came with it:
- By-ref parameters. The call carries their positions and the answer carries
what each holds afterwards; a position the answer omits was never assigned
to, which is what PHP does with an untouched by-ref parameter.
ProcessExecutor::execute is the only one on a proxied class.
- Argument arity, reproduced where the real body reads func_num_args().
execute($cmd) forwards the child's output and execute($cmd, $out) captures
it, and nothing but the argument count separates the two.
- Static methods that cannot run in the worker. One that reads a static
property the Rust side owns, or that reaches a guarded class, forwards
through __shirabeCallStatic instead of being materialized. That also fixes
Filesystem::isLocalPath and getPlatformPath, whose materialized bodies
called the guarded Composer\Util\Platform.
The async surface stays an explicit error. executeAsync resolves its promise
with a Symfony Process, whose proc_open() resource and pipes belong to
whichever process called start(), so a Rust-side spawn has none to hand back;
running the real start() in the worker needs a promise representation that
crosses the boundary unresolved.
The fixture project drives the whole synchronous surface from plugin code and
compares the trace against upstream Composer byte for byte.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
|
|
Comments that pointed at design notes kept outside the repository are dead
ends for anyone reading only the tree, so what each of them explained now
lives in a tagged TODO at the site it applies to. Several of those sites
also stated something the implementation does not do, and the TODOs record
the actual gap instead: the two halves of the codec recognize handle
descriptors by different rules, the scripts Command path drops the exception
class and collects output in a BufferedOutput that cannot carry an
interactive command, find_shortest_path panics where PHP throws, and the
package dispatch hand-rolls the variant selection AnyPackage should own.
The classifier document likewise described rust-snapshot,
plugin-constructible and several of the open questions as designed rather
than as built.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
|
|
|
|
The shim carried the PCRE A (anchored) modifier alongside every compiled
pattern so preg_match2 could honour it by searching the sub-slice at the
offset. Only two call sites ever passed such a pattern, and each can cut
that slice itself, so the flag is gone from the cache, ResolvedPattern,
PregPattern and the php_regex! macro, and a pattern still carrying A is
now rejected rather than silently searched unanchored.
StringInput::tokenize and PhpFileCleaner::match search from their cursor
with a `^`-prefixed pattern instead. The rule is written down in
docs/dev/regex-porting.md.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
The worker is a long-lived child holding the environment it was handed
at spawn, so `@putenv`, the bin dir the event dispatcher prepends to
PATH, and COMPOSER_DEV_MODE never reached the PHP code running in it.
The shim now journals every write to the three storages PHP exposes, and
the outermost rpc_call replays the entries the worker has not seen yet
through __shirabe_sync_env. Replaying the writes rather than pushing a
whole snapshot keeps the worker's own $_SERVER entries intact.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
The worker's autoloader fell through to the real Composer source for every
Rust-owned FQCN without a proxy stub, so plugin code doing `new Filesystem()`
or subclassing `LibraryInstaller` silently ran on a second instance the Rust
side never sees. An unimplemented part of the plugin API has to fail with an
explicit error naming it, not quietly work on a disconnected copy.
The stub generator now emits a guard class for each of those FQCNs: the real
declaration, hierarchy and constants, with every constructor and method
raising an explicit error. References satisfied by the declaration alone
(`instanceof`, `X::class`, `Link::TYPE_REQUIRE`) keep working. Two FQCNs stay
resolvable to the real class, each listed with the worker-side mechanism that
makes a natively constructed instance correct.
The error had nowhere to go: `Installer::run` dropped the `Result` of both
`dispatch_script` calls, so an exception from a listener ended in exit 0.
Both propagate now, the way the exception does upstream.
Three real-plugin E2E comparisons stop at a guard and are ignored, each
naming the class it needs.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
Split date() into date_utc() and date_local(), the latter resolving the
system's local timezone through the tzfile crate ($TZ, then
/etc/localtime, falling back to UTC when neither is readable). The
timestamps Composer renders for humans -- the GitHub OAuth token note,
the GitHub API rate limit reset time, the Perforce client spec fields and
the "today" check of the show command -- now go through date_local().
PHP resolves its default timezone from the date.timezone ini setting,
which Shirabe does not read, so date_default_timezone_get/set have no
input left to model and are dropped from the shim and its callers. The
resulting difference is recorded in docs/known-incompatibilities.md.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
The binary called itself Composer everywhere: the application name, the
logo, --version, about, and every warning that talks about the running
program. Prompts to file a bug also pointed at Composer's issue tracker.
Add SHIRABE_VERSION and SHIRABE_RELEASE_DATE next to the Composer version
constants and report those, naming the Composer version this port tracks
alongside them. Composer::VERSION and getVersion() are untouched, so the
composer platform package, composer-runtime-api and the HTTP User-Agent
keep the value plugins and package repositories expect.
build.rs stamps the release date with the UTC date of the HEAD commit,
the way Composer's Compiler fills in @release_date@ when building the
phar. It now also fails the build when git cannot be read, instead of
letting COMPOSER_DEV_WARNING_TIME fall back to the tagged-release value
and suppress the outdated-build warning forever.
Messages about the Composer ecosystem keep their wording. Two of them are
pinned by upstream installer fixtures (Rule's "cannot be modified by
Composer" and SolverProblemsException's "you can run Composer with") and
stay as they are so those fixtures can keep being used verbatim.
The e2e list comparison against upstream Composer now skips the banner,
which cannot match by design, and compares everything below it as before.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
`__halt_compiler` is a PHP keyword, so a stub may spell the token in any
letter case. The native phar reader compared the bytes exactly and rejected
such an archive, and the lint that keeps a second token out of the
executable missed lowercase ones, which would shadow the embedded bundle.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
Plugins and scripts need the real `Composer\` classes and the packages
Composer depends on, which so far came from a checkout found through
SHIRABE_COMPOSER_PHP_DIR or a path next to the workspace. Neither exists
for a distributed binary.
The build script now archives those PHP sources into a phar the way
Compiler.php does and the executable carries it. The worker maps it with
Phar::loadPhar and reads a content-addressed sentinel back to tell a
bundle it can use from one it cannot; where its PHP cannot open the phar,
the bundle is unpacked once into the cache directory and autoloaded from
there. SHIRABE_COMPOSER_PHP_DIR still overrides both for development.
PHP locates a phar's manifest by the first __HALT_COMPILER(); token in
the file, so the executable must hold no other copy of it: phar.rs builds
the token at run time, and a linter keeps further literals out of the
sources that reach the binary.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
Output buffering captures whatever the PHP interpreter would echo to
stdout. The shim routes no output through a buffer, so these functions
could never capture anything and stayed todo!(). Record the gap in
docs/known-incompatibilities.md instead.
|
|
The SignalHandler port was a no-op stub, so all four of Composer's abort
paths were dead code: nothing removed a half-created project, reverted
composer.json, or cleaned up half-installed packages.
Composer runs those handlers from pcntl callbacks, which a Rust signal
handler cannot do -- it may touch nothing beyond atomics. SignalSubscription
records the signal instead, and the abort runs from checkpoints on the normal
call stack, where the clean-up can borrow the state it needs. That also
resolves the closure-capture TODO(phase-c)s in RequireCommand and
InstallationManager, and replaces exit_with_last_signal's exit(0) with the
restore-and-re-raise Seld\Signal does.
A subscription is live only inside the four abort regions, so elsewhere the
signals keep their default disposition and kill the process at once. It is
installed without SA_RESTART so a signal interrupts an interactive prompt
rather than resuming the read. A signal reaches only the innermost
subscription, reproducing SignalHandler's single-stack dispatch.
Drop SignalRegistry, SignalableCommandInterface and the Application wiring
for them: nothing in Composer reaches that path, and SignalHandler discards
whatever they register. Signal handling from plugins and scripts is
undefined behavior; see docs/dev/signals.md.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
Composer's Composer\Util\TlsHelper is marked deprecated for removal in
Composer 3.0 and has no caller in composer/composer outside its own
test: PHP's stream layer verifies certificate hostnames itself, and the
one surviving method delegates to composer/ca-bundle.
The Rust port had no caller either, so it, its test, and the
openssl_x509_parse shim it was the sole user of are removed.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
Composer restarts itself with Xdebug unloaded because Xdebug makes PHP
several times slower. Xdebug is never loaded into this process, so what
needs dealing with is the PHP worker: it is spawned with
`-d xdebug.mode=off` and `XDEBUG_MODE=off` (Xdebug reads the environment
variable first and lets it override every ini setting), which makes its
module init return before it installs any executor, compile, error or
opcode hook.
Rewriting the ini files and re-executing, the way xdebug-handler does,
would additionally cover Xdebug 2, which hooks unconditionally and has no
equivalent setting. That is not worth its machinery here: Xdebug 2 caps
out at PHP 7.4, while every PHP version Composer supports can run
Xdebug 3.
What remains of XdebugHandler is small enough to live beside the worker
it governs, so its crate is gone and its callers inline it. isXdebugActive
answers false without asking PHP whenever the worker is switched off, so a
command that needs no PHP does not spawn one just for the Xdebug warning;
diagnose reports what the worker measures instead, which still surfaces an
Xdebug that ignores the setting. PlatformRepository has no unloaded
extension to restore, since switching the mode off leaves it loaded.
COMPOSER_ORIGINAL_INIS is neither written nor read: it exists so a
restarted process can name the ini files it replaced, and IniHelper can
report the worker's own. IniHelperTest injects through that variable, so
none of its cases are ported.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
PluginManager::getPluginCapability hands the plugin itself to the
capability constructor. Only a PHP-implemented plugin had an entity the
child could receive, so a Rust-implemented one bailed out; it now crosses
as a handle to an R-table entity behind Shirabe\RustPluginStub, or its
Capable flavour, since `$plugin instanceof Capable` is what decides
whether Composer asks a plugin for capabilities at all.
This is what the two capability tests of PluginInstallerTest were waiting
on: the mocked Capable plugin is Rust-side, and one of them asserts the
identity of the plugin read back out of $capability->args.
|
|
is unsupported
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
InstallationManager left both PRE_PACKAGE_* and POST_PACKAGE_* as empty
stubs, so a subscriber never ran at all and the difference from upstream was
silent rather than an explicit error.
Operations cross the boundary as R-table entities with generated proxy
stubs. Materializing them the way a Link crosses is not possible: a
materialized value is revived by unserialize() on the child side, so its
properties never pass through the wire decoder and a nested handle
descriptor would not come back as a stub — and an operation always holds a
PackageInterface. execute() now shares one Rc per operation through the
whole batch pipeline, so a plugin sees one object for both the pre- and the
post-event of an operation, as it does in PHP.
POST_PACKAGE_* also moves out of the operation's promise chain into the
post-exec callback list PHP runs after waitOnPromises().
The stub generator materializes non-public class constants verbatim now,
which the operation classes need for their `protected const TYPE`: a
constant has no entity behind it, so a copy in the worker cannot diverge,
and keeping the declared visibility exposes nothing the real class hides.
The E2E fixture added here compares the recorded events against upstream
Composer. It also surfaced that upstream starts an operation's chain where
it is built (a null prepare() becomes an already-fulfilled React promise
whose handlers run through the immediately drained queue) while this port
only drives its futures in wait_on_promises, so the repository state a
pre-event observes differs; that half of the comparison is a separate
`#[ignore]`d test.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
A proxy stub declares none of the real class's instance properties, and the
`__get`/`__set` forwarders were emitted only for classes that declare a public
one. Every other property access therefore got PHP's own answer for an
undeclared property — null on a read, a dynamic property on a write — so plugin
code reading state the entity holds ran on with null and failed somewhere else
entirely, or not at all.
Emit the forwarders on every root stub, add `__isset`/`__unset` alongside them
so `isset()` cannot answer false silently either, and serve all four from one
dispatcher that answers the state the Rust-side entity exposes and rejects every
other name.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
The RuntimeInterface seam asked the worker one question at a time: a round
trip per loaded extension, per ReflectionExtension::info() output and per
constant, so a single `show --platform` cost 70 to 100 of them. A `platform`
dispatch entry now answers all of it as one PHP array, which shirabe-php-rpc
decodes into a OnceLock-cached PlatformInfo, the way the diagnose command
already works.
Composer\Platform\Runtime therefore has no Rust counterpart any more. Its
work belongs to the running interpreter, and invoke()/construct() could only
be ported as a whitelist that panicked on anything unlisted; it is ported as
PHP into the worker instead, and PlatformRepository reads the answers off
PlatformInfo. Accessors panic on a name the payload does not carry, so the
worker and its consumers cannot drift apart unnoticed.
The tests describe the runtime as payload data where they used to mock the
seam, with the datasets unchanged. The one loss is the call-count assertion
of test_inet_pton_regression: the payload reports the result of
`@inet_pton('::')` rather than answering a call.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
A materialized value used to cross as a constructor call: the class name,
the arguments, and any post-construction setter. Describing a real
instance that way needed a ReflectionProperty read for every field the
class exposes no getter for, and state no constructor takes (an unset
pretty string, a Link built without a pretty constraint) had no faithful
call to describe it at all.
The value now crosses as the object record serialize() writes for it,
which unserialize() revives without running a constructor, so both sides
transfer the state itself instead of a recipe for rebuilding it. The PHP
half keeps only the class list (also the allowed_classes list of every
frame payload) and the UTC rebasing of dates; describe(), build() and the
reflection are gone.
The wire codec gains O: records (PluginValue::PhpObject) and r: back
references, whose resolution reproduces PHP's numbering of every value in
a payload; a cyclic object graph and a PHP reference (R:) are rejected.
Two behaviours change with it: a date crosses carrying timezone_type 3
"UTC" rather than a +00:00 offset, which is what ArrayLoader builds a
release date as, and a Link subclass crosses as a P-table entity instead
of being silently downgraded to a plain Link.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
The transport bound an AF_UNIX path under TMPDIR and waited for the child to
connect. That path has to fit in sun_path (108 bytes), so a deep TMPDIR made
every worker spawn fail, and the bind(2) itself is denied under a sandbox.
The parent now keeps one end of a socketpair and installs the other on
descriptor 3 before exec, which the worker opens as php://fd/3. The pair is
connected from the start, so the accept poll and its ten-second deadline are
gone; a child that dies before reading surfaces as EOF on the first call,
where worker_state already attaches its exit status.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
The link getters and setters and the release date accessors were explicit
errors for every non-empty value, because an immutable value has no entity
to point a handle at. They now cross as materialized values: the descriptor
names the real class and the constructor arguments, and each side builds a
genuine instance of its own.
The semver constraint a link holds is encoded structurally rather than
re-parsed from its string form, so the pretty strings and the conjunctive
flag survive the crossing.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
An installer that extends LibraryInstaller reaches for the Composer object
graph in ways the proxy did not answer: the download manager and the config,
the writable repository methods, a React promise as its own return value, and
`new Package(...)` from its supports() path.
* `Composer::getConfig`/`getDownloadManager` are dispatched, and both classes
become generated proxy stubs. Their reads and mutators answer from the Rust
entity; the surfaces needing stubs of their own (ConfigSourceInterface,
DownloaderInterface) stay explicit errors.
* The download manager's futures are driven to completion and handed back as
already-settled React promises, since PHP declares a non-nullable
PromiseInterface there. A promise a plugin returns is drained the same way:
settled yields its value, rejected re-raises, pending is an explicit error.
* Proxy stubs now carry the real class's constructor and ask the Rust side to
allocate the entity; reviving a stub for an existing entity binds the handle
without running it. Classes Rust cannot build name themselves in the error.
* The package proxy covers `Package`'s own setters, `CompletePackage`'s
metadata, `RootPackage`'s root-only state, and `BasePackage::$id`.
Link values and release dates still have no wire image, so the methods
carrying them remain explicit errors.
|
|
A plugin can now hand an InstallerInterface implementation to
InstallationManager::addInstaller across the wire, and a legacy
composer-installer package is loaded as one; both are backed by a
PhpInstallerProxy forwarding the whole installer contract to the entity in
the PHP worker. An installer returning a real promise is an explicit error
until promises can cross the boundary.
InstallationManager takes installers as shared handles instead of boxes, so
the object identity removeInstaller and PluginManager's registeredPlugins
compare against survives registration, and holds them in a RefCell:
Installer::run keeps a shared borrow of the manager for the whole run, and a
plugin activated inside it registers its installer from there. The type
cache keys on the installer itself, like upstream, so re-entrant
registration cannot leave a stale index behind. InstallerInterface::supports
is fallible for the same reason getCapabilities and getCommands are: it
answers over RPC.
Cloning a proxy stub clones the Rust-side entity and rebinds the copy to the
fresh handle. Previously only the classes declaring __clone got a throwing
body, and the rest let two stubs share (and twice release) one handle.
Package entities answer with AnyPackage::dup, which already carries
BasePackage::__clone and the RootAliasPackage override; the others are an
explicit error.
The package proxy covers the whole PackageInterface surface; only the link
maps and the release date still lack a wire image for their value objects.
PluginManager gains a test-only seam for the reported Plugin API version,
and the three PluginInstallerTest cases that need it are ported.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
A same-FQCN Composer\Console\Application, hand-written under the new
php/runtime/ tree, hosts CommandProvider commands inside the PHP worker:
PhpCommandProxy overrides run() and forwards the stringified input, so the
real Symfony machinery binds, validates and executes against the live
command object, while help/list render Rust-side from a definition read
back at construction. Reverse \Shirabe\RustCommandStub rows let a plugin
command invoke built-in commands back in the Rust process, keeping every
command on the side whose helper set it was written for.
Composer\EventDispatcher\Event moves from a generated stub to a dual-mode
runtime class: the real BaseCommand::initialize constructs a
PreCommandRunEvent natively in the worker, which a proxy-only constructor
guard rejected. Its PRE_COMMAND_RUN dispatch reaches a new EventDispatcher
stub whose dispatch supports the observably-no-op no-listener case and
fails explicitly otherwise. The stub generator now accepts runtime-provided
classes as stub bases (never as targets) and cross-checks the Application
handoff property table against the real class.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
Emit the eleven stubs a subscriber plugin's object-graph calls reach
(BasePackage through RootPackage, the installed-repository hierarchy,
RepositoryManager, InstallationManager). The generator learns the member
kinds these classes need: builtin interfaces contribute no closure entries,
public static properties are materialized, public instance properties
forward through __get/__set, __toString forwards like a plain method,
__clone throws (proxy clone semantics are still an open design question),
and a subclass stub may add interfaces to its inherited surface.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
Replace the hand-written proxy stubs under crates/shirabe-php-rpc/php/stubs
with output of scripts/plugin-stub-generator, a deterministic emitter that
derives every stub from the Composer checkout and the classifier report.
Anything it cannot faithfully proxy (by-ref/variadic parameters, magic
methods, public properties, diverging omitted overrides, stale stub files,
a STUB_FILES entry missing in lib.rs) fails generation instead of degrading
silently, so future Composer releases surface new members as explicit
errors rather than silent gaps.
Regenerating the stubs also normalizes the hand-written inconsistencies
(uniform guarded constructors, import-based type spellings) and fixes real
gaps the review of the generated diff uncovered: the BaseIO authentication
methods now carry the real class's untyped signatures, and the previously
missing ConsoleIO::sanitize is materialized together with its private
static helper. A cargo test runs generate-stubs --check to keep the
committed stubs, the generator and the embedded list from drifting apart.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
InstallationManager::execute now takes &self (the mock recorder moved
into a RefCell) and its callers hold only shared borrows: plugin
registration inside a batch re-enters the same manager handle through
Composer::getInstallationManager()->getInstallPath(), which panicked
on the RefCell re-borrow under the &mut shape — the same re-entrancy
the repository side already fixed, unreachable from the ported tests
because they call PluginInstaller::install directly like PHPUnit does.
The worker-side InstalledVersions mirror now matches the full tail of
FilesystemRepository::write: unconditional reload plus the
reflection-based selfDir/installedIsLocalDir restore. The previous
class_exists(false) guard rested on a lazy-load assumption that does
not hold in the worker (its real ClassLoader only knows the Composer
checkout's vendor dir, so a later lazy load would read the checkout's
installed.php, not the project's); the mirror is now skipped only when
the class is not autoloadable at all, i.e. no plugin runtime and hence
no observer code. Boot-time seeding stays TODO(plugin).
Also from the review: registered_plugins entries are removed only
after the deactivate/uninstall loop (PHP unsets last, and a throw must
leave the entry observable); extra.class keeps associative-array
values and fails loudly on non-strings instead of silently dropping
them; the two discarded write() results now propagate (they carry the
reload-push failure); register_package's allow-plugins skip message is
DEBUG like the addPlugin side; the loader-eviction divergence of
REGISTERED_LOADERS and the lossy UTF-8 spots carry searchable
markers; the test-only proxy downcast follows the __ naming rule; the
R-table dispatch clones the entity out instead of holding the table
borrow across the handler; the IO/PartialComposer stubs turn a
plugin-side `new NullIO()` into an explicit error instead of an
ArgumentCountError; and the empty() emulation covers float 0.0.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
Implement the remainder of PluginManager::registerPackage: the plugin
autoload map is built by the ported createLoader/parseAutoloads and
served to the worker over the existing reverse-RPC autoloader, files
entries go through a composerRequire-equivalent glue call, and
already-defined classes take the upstream _composer_tmp rename/eval
path. Instantiation uses the new NewObject/CallPhpMethod lanes backed
by a P table in the worker; PhpPluginProxy adapts the resulting handle
to PluginInterface, with $composer/$io exposed to plugin callbacks via
an R table (unsupported methods stay explicit errors). Hand-written
proxy stubs cover Composer, PartialComposer and the IO hierarchy, and
the stub autoloader is re-prepended after loading the Composer PHP
runtime so its vendor autoloader cannot shadow proxied FQCNs.
FilesystemRepository::write now mirrors InstalledVersions::reload into
a running worker (class_exists-guarded, so an unloaded class keeps its
upstream lazy-load behavior), removing the previously undefined
observation window.
The installer pipeline passes the installed repository as a shared
handle instead of a long-lived `&mut dyn`: plugin registration runs
inside InstallationManager::execute and re-enters the same local
repository through the RepositoryManager, which would panic on the
RefCell re-borrow under the old shape.
PluginInterface lifecycle methods now take an owned ComposerHandle
(plugins retain $composer past the call) and return anyhow::Result
(PHP plugin code may throw); the plugin list uses shared ownership so
the identity comparison of removePlugin survives the dual storage in
registeredPlugins, matching PHP reference semantics.
Ports the activate/upgrade/uninstall tests of PluginInstallerTest,
serialized across the shared worker process whose persistent class
table is exactly what exercises the rename path.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
Replace the name\0arg framing with the plugin wire protocol: tagged
frames with corr_id multiplexing, a MAX_FRAME_LEN bound, a thread-ID
based reentrant SessionLock, and the PluginValue codec (encoder plus
the first recursive decoder, iterative with a 512-level depth cap).
Float formatting is ported from php-src into shirabe-php-src so the
encoder is byte-compatible with serialize() under serialize_precision=-1,
which the spawned worker now pins. The worker gains a standing dispatch
loop, CallRustMethod reentrancy, hand-written Event proxy stubs, and
explicit-error answers for everything not implemented yet. The public
query API (get_php_version and friends) is unchanged and now rides the
new protocol; the codec is verified against real PHP by roundtrip
oracle tests covering floats, non-UTF-8 bytes and deep nesting.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
The Package family settles on pure rust-proxy: it is genuinely mutable
(60+ setters, and Composer itself mutates packages in flight), so the
planned snapshot-with-writeback treatment is dropped rather than filed
as an override. The other open questions remain open with their interim
behavior pinned instead of resolved: proxy-side getLoop() access and the
ConsoleIO table/progress-bar members are explicit errors, and the
InstalledVersions state a plugin observes after a Rust-side dump is
undefined, marked TODO(plugin) at the reload site.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
diagnose used to read hardcoded shim stubs, so it described a fictional
runtime: OPENSSL_VERSION_NUMBER was always 0 and tripped the TLSv1.1/1.2
check, PHP_BINARY and OPENSSL_VERSION_TEXT were empty, and the extension,
function and ini probes answered from a fixed table.
The PHP worker gained a `diagnose` entry that returns every fact the
command needs as one PHP array, cached in a OnceLock so the several call
sites share a single round trip. Reading it back needed array support in
the serialize() parser, which in turn lets get_loaded_extensions and
get_all_ini_files return real lists instead of comma-joined strings.
Also fixes the openssl_version message, which dropped strstr()'s
before_needle argument during the port, and check_connectivity's
allow_url_fopen test, which did not follow PHP string truthiness.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
A plugin can reach ProcessExecutor::executeAsync() through the
rust-proxy stub, which resolves with a real Symfony Process instance
that can't be reconstructed on the Rust side (its state is tied to
whichever process calls proc_open(), and it refuses serialization).
Add execute_async_php() as a todo!() stub, documented as a dual-
instantiation split in plugin-class-classification.md: Rust-internal
callers keep using execute_async(), while the plugin path must forward
spawning to the PHP child once the RPC channel exists.
|
|
Answers how a given class is treated at the plugin boundary, accepting a
PHP source file, a Rust source file, or a (short or fully qualified)
class name. Reads report.json and generates it first when missing. Rust
paths resolve by normalized segment matching because the snake_case
mapping is not reversible for acronyms (io_interface.rs -> IOInterface).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
Decides, for every composer/composer class, how it is treated at the
plugin boundary (rust-proxy / rust-snapshot / contract / two-world /
php-native / unsupported) so that upstream updates re-classify new or
rewritten classes without re-deriving the design by hand. Rules and
category definitions live in docs/dev/plugin-class-classification.md;
the tool (PHP + nikic/PHP-Parser) implements them as a reachability
closure with direction marks, per-method pure/mutator analysis, and a
leaf-first fixed point for unreachable classes, with three small
versioned exception lists.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
|
|
Worker spawn failure, socket I/O errors, and unparseable responses used to
fall back to plausible-looking defaults (empty string, false, None), making
real failures indistinguishable from legitimate PHP-side results. Panicking
is not the final design, but replaces silent data corruption with a loud
failure until proper error propagation is implemented.
|
|
Avoid colliding with an existing Composer installation on the same
machine by defaulting COMPOSER_HOME/CACHE_DIR/DATA_DIR paths to
shirabe/Shirabe instead of composer/Composer, while keeping the env
var names, composer.json/lock, and vendor/composer/ unchanged for
ecosystem compatibility.
|
|
Runtime::hasConstant/getConstant need a real PHP interpreter's defined()/
constant() to answer platform requirement checks (e.g. PHP_ZTS, PHP_INT_SIZE),
which the shim can't provide since Rust constants aren't queryable by string.
Extend shirabe-php-rpc's protocol to carry one string argument and return the
full PHP scalar range, add defined/constant dispatch entries to the worker,
and wire Runtime and get_php_version/get_php_binary onto them.
|
|
Add a minimal shirabe-php-rpc crate that spawns the system PHP as a
child process and asks it for runtime information over a Unix domain
socket, then use it to fill the `--version` PHP line with the real
\PHP_VERSION and \PHP_BINARY instead of fixed placeholder values.
See docs/dev/php-rpc.md for the design and scope.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
|
Rework the environment shim around getenv/putenv on the real environment
and $_ENV/$_SERVER as startup snapshots, all over OsString. Migrate every
caller off the old server()/server_argv() helpers and force the snapshots
in main() before any putenv() runs. Document the porting rules in
docs/dev/env-vars-porting.md.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
|
json_encode/json_encode_ex now return anyhow::Result<String> instead of
Option, so callers no longer need json_last_error() to get the failure
reason.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
|
Document the conventions for porting PCRE patterns to the regex crate
(no PCRE crate, panic on compile failure, dropping performance-only
possessive quantifiers, ad-hoc compatibility comments). Apply the rule
to Platform::expand_path by rewriting its conditional subpattern as an
explicit alternation, which the regex crate can compile.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
