Two new languages in preview (PHP and C#), a stricter diff, a sturdier napi bridge, and 18
fixes from the first outside contributor.
- 18 pull requests from @rudycelekli, merged with their history: HTTP responses that end early
or carry informational headers (#7, #8), empty recordings no longer pass (#9), float tolerance
keeps key order (#10), own__proto__keys survive (#11), curl examples with a query right
after the host (#12), TS projects without a manifest and.mts/.cts(#13),setupkeeps
custom MCP server options (#14), replay deadlines cover child processes (#15), colliding case
names (#16), binary CLI cases keep both streams exactly (#17), the Rust 1.97 minimum is declared
and checked (#18), a missing port source failsgate(#19), Ruby recordings are published
only when complete (#20), JNI names with$and non-ASCII (#21), Go internal packages and
versioned modules (#22), generated Go names that clash with parameters (#23), and invalid Java
replay arguments (#24). - PHP, in preview: detected from
composer.jsonor.phpfiles.analyzereads Composer
packages and PSR-4 roots, resolvesrequireanduse, finds functions and public static
methods with native and phpdoc types, flags PHP's risks (loose==,/returning floats, byte
strings againstmb_*, PCRE features theregexcrate lacks,eval, magic methods) and maps
common Composer packages to crates.analyze --whole-appfinds CLIs ($argv, composerbin)
and servers (front controllers,$_SERVER).app record,app replayandapp front-door
work on PHP CLIs and on servers run withphp -S. Bridging PHP units into Rust (PHP FFI over a
C ABI) comes next;bridgesays so for now. - Fix (#3):
.gitignorefiles in folders
above the repository (a*in a home-folder.gitignore, for one) no longer hide every source
file. Ignore files are read the way git reads them: from the repository root down, and not from
parent folders at all outside a repository. When ignore rules still exclude every source file,
analyzesays so instead of returning an empty plan. - PHP's built-in server echoes a
Hostheader with the port it ran on; HTTP recordings now skip
it, likeDate. - Fix (#4): Python units bridged on macOS
failed to link (Undefined symbols ... _Py_IsInitialized). The PyO3 crate now gets the same
build.rsthe Ruby bridge has, which lets the interpreter provide Python's symbols at load time.
rust-anything refresh <crate>adds it to crates bridged before this fix. - Fix: a Rust panic no longer passes
diffwhere the original also threw. An unfinished port
(unimplemented!) used to "match" every recorded error case in TS, Ruby (a panic arrives as
RuntimeError), Java (RuntimeException) and Go (a panic where the original panicked). Python's
driver crashed on the firstPanicExceptioninstead of reporting it, andshadowlet it reach
the app. Each language now has a test that an unported crate passes no case. - The napi bridge, after a study of napi-rs:
- errors keep their JS class: a port's
Err("TypeError: ...")throws a realTypeError, a
wrong argument type is aTypeError, and a panic carriescode: 'RUST_PANIC'; - async exports reject their promise instead of throwing at the call;
Uint8Array,Buffer,Map,Set,Dateandbigintcross the bridge with the markers
recordings use, instead of turning into plain objects or throwing;- a string with a lone surrogate is refused with a
TypeErrorinstead of silently becoming
U+FFFD in a RustString; buildreplaces the.node(and the PyO3 module) by renaming, anddiffrefuses to run when
the built module is older than the crate's sources.
- errors keep their JS class: a port's
- From Reddit feedback:
analyzerefuses a compiled app (APK, JAR, .exe and the like) with an explanation instead of
an empty plan, and warns when a folder holds only compiled files.- Platform-dependent C (inline assembly,
volatile, bitfields, packed layouts, compiler
extensions, integers cast to addresses) is a newplatform-specificrisk, and its unit is
kept in C: a recording on this machine says nothing about the target. analyze --whole-apprecognizes Electron apps and plans the main process, keeping the web UI.gatehas asizecheck: the port fails above 3x the original's lines (plus 50).- README FAQ: APKs, battle-tested C, Electron, C#, and code growth.
- C#, in preview: detected from a
.csprojor.sln.analyzereads projects and their NuGet
packages (mapped to crates, test projects left out), resolvesusingdirectives to folders,
findspublic staticmethods with their types, and flags C#'s risks (decimal,dynamic,
LINQ queries,Task,lock,unsafe, reflection, regex features theregexcrate lacks).
analyze --whole-appfinds CLIs (<OutputType>Exe) and ASP.NET servers (the web SDK,
WebApplication.CreateBuilder).app record,app replayandapp front-doorrun built C#
apps (dotnet out/app.dll). Bridging C# units (P/Invoke over a C ABI) comes next;bridgesays
so for now. - Fix (#25): a bridge crate in its default
place inside the project (rust/<crate>) was read as source, so Javarecordcompiled its
generated wrappers next to the originals ("duplicate class"). Every source walk now skips
folders holding arust-anything.json. - Fix (#26): on Windows, paths handed to
java -cp,-Dlibrary paths and the other runtimes carried the\\?\verbatim prefix, so
Javarecord,diffandbenchcould not find their classes. Paths are now made absolute
without it everywhere. - Fixes found by gating every demo port (#27
reported the first): the generated binary wire tripped the new
chunks_exact_to_as_chunkslint (every Go and Java port failedgate), and on macOS the
Python and Ruby extension's test binary failed to link; the generatedbuild.rsnow covers it,
andrefreshadds it to Ruby crates too. CI now gates all demo ports on macOS. - CI also runs every test on macOS, and a new
tests/native.rsbuilds a napi addon and a PyO3
extension for real, loads them from Node and Python and replays the recording, so link problems
like #4 fail CI instead of reaching users. - A
.github/FUNDING.ymlturns on the repository's Sponsor button.