When compiling a single expression, the parser now explicitly consumes the TOKEN_EOF token after parsing the expression. This ensures that any trailing garbage or unexpected tokens after the expression are caught and reported as errors, rather than being silently ignored. The comment for the non-expression path is also updated to clarify that it checks for end of file rather than end of block.
Replace strncpy with memcpy in readBuiltInModule to avoid potential string termination issues
when copying module source code, and increase ERROR_MESSAGE_SIZE from 60 to 80 to provide
more room for compiler error messages with long variable names.
Migrate the internal `String` struct to use the heap-allocated `ObjString*` throughout the compiler, debugger, VM, and value system. This unifies string representation, enables proper garbage collection of symbol table entries via new `wrenBlackenSymbolTable` function, and removes manual memory management for symbol names.
Add handling for UTF-8 byte order mark (BOM) at the start of source code
in wrenCompile function by checking for the three-byte sequence \xEF\xBB\xBF
and advancing the source pointer past it. Include a regression test file
(test/regression/520.wren) that contains a BOM to verify the fix works
correctly.
Add a MAX_PARAMETERS bound to the loop in signatureParameterList to prevent
buffer overflow when a method signature has an excessive number of parameters.
A compile error is already reported in such cases, but the unbounded loop could
still write past the end of the name buffer. Also add regression test 494.wren
to cover this scenario.
The compiler now always reserves slot zero for either the closure or method
receiver, eliminating a corner case where top-level REPL code lacked the
pre-allocated slot that function/method bodies expect. This ensures local
variables declared in REPL loops (e.g. `for (i in 1..2)`) get correct slot
indexes.
Additionally, inlines `wrenResetFiber` into `wrenNewFiber` and stores the
imported module's closure on the stack before invoking it, preventing a GC
from collecting the closure during import.
Fix two documentation errors: change `Foo.bar.printFromStatic()` to
`Foo.printFromStatic()` in classes.markdown, and replace `map` with `each`
in concurrency.markdown to match the actual Wren API.
The previous file used a non-standard preamble ("Wren uses the MIT License:")
and included an extra parenthetical clarification about binary distributions.
This commit rewrites the license body to exactly follow the standard MIT
License template published at https://choosealicense.com/licenses/mit/,
removes the custom header, and normalizes line wrapping so that GitHub and
other license detection tools correctly identify the project as MIT-licensed.
The example code in the embedding documentation for bindForeignMethod() incorrectly used `!isStatic` when checking for a static method `Math.add(_,_)`. Changed the condition to `isStatic` so the example correctly demonstrates how to dispatch static foreign methods.
The assertion for the `name` parameter in `wrenGetVariable` was incorrectly checking `module != NULL` instead of `name != NULL`. This fix ensures the variable name parameter is properly validated, preventing potential null pointer dereferences when retrieving variables.
Additionally, adds Marshall Bowers to the AUTHORS file.
The getNumArguments function had CODE_IMPORT_MODULE incorrectly returning 1 instead of 2, and CODE_IMPORT_VARIABLE returning 2 instead of 4. Additionally, wrenBindMethodCode had two bugs: it was incrementing ip before reading the instruction, causing incorrect opcode dispatch, and it was using ip++ instead of ip+1 for field offset adjustment, corrupting subsequent bytecode. The super instruction handling also incorrectly skipped over the symbol bytes before reading the constant index.
The error callback function signature in the embedding documentation was missing the required WrenVM* first parameter, which would cause compilation errors for users following the example. The fix adds the vm parameter to the error function declaration in the configuring-the-vm markdown file.
The compileInModule function now returns a closure directly instead of wrapping it in a fiber, simplifying the compilation pipeline and removing unnecessary fiber overhead for module imports. The wrenImportModule function is moved into wren_vm.c and rewritten to execute the imported module's closure immediately rather than returning a fiber for deferred execution. This change also adds a test for yielding from an imported module to verify correct behavior with the new execution model.
Remove redundant NULL-check branches for module parameter by merging the
wrenPopRoot call into a single conditional after compileInModule, eliminating
the early return path and simplifying control flow in the compilation entry point.
Replace the bundled GYP build system with a newer upstream version that adds z/OS platform support, improves MSVC SDK version detection with compatible SDK fallback, fixes host LDFLAGS handling in ninja generator, and includes various bug fixes and maintainer updates across multiple generator backends.
The switch statement in getNumArguments covers all opcode cases, but the compiler
cannot prove exhaustiveness. Adding UNREACHABLE() followed by return 0 eliminates
the "control reaches end of non-void function" warning on strict compiler settings.
The documentation examples for the `from`, `to`, and `max` properties of the Range class were incorrectly using `.min` in their code samples. Updated the examples in `doc/site/modules/core/range.markdown` to call the correct methods: `.from` for the `from` property, `.to` for the `to` property, and `.max` for the `max` property, ensuring the printed output values match the documented behavior.
Replace the previous approach of desugaring import statements into calls to System.importModule() and System.getModuleVariable() with two new bytecode instructions: CODE_IMPORT_MODULE and CODE_IMPORT_VARIABLE. This eliminates the need for the corresponding primitive methods in the core library, adds proper debug dump support for the new opcodes, and updates the compiler to emit these instructions directly. The getNumArguments function is also updated to handle the new opcodes and the unreachable default case is removed.
The link target for the embedded API reference was pointing to
`embedding-api.html` but the actual generated page uses the path
`embedding` without the `.html` extension. This change fixes the
link so it resolves correctly on the documentation site.
When a fiber is started for the first time via call() or transfer(), the value argument is now bound to the fiber's function parameter if it takes one. Previously, the first value was always ignored. Also adds runtime error for fibers created with functions taking more than one parameter, and updates documentation and tests accordingly.
- Replace string concatenation with interpolation in class method examples
- Remove backticks from Fiber class references in concurrency docs
- Add guidelines for choosing between getters and empty-parentheses methods
- Clarify block syntax for single-expression functions and chaining
- Delete the empty "Application Lifecycle" page (was just a TODO placeholder)
- Expand "Storing C Data" from stub to full documentation covering foreign classes, initialization, access, and finalization
- Add note that VM is not re-entrant; warn against calling wrenCall/wrenInterpret from foreign methods
- Fix broken cross-reference links (configuring-the-vm.html, wren.hpp include)
- Clarify slot API usage: remove confusing sentence about implicit return of receiver
- Fix grammar: "each VM them differently" → "each VM differently", "is only be called" → "is only called"
- Improve phrasing throughout for clarity and consistency
Replace `Meta.eval()` to compile source into a fiber instead of a function, fixing issue #456 where top-level variable declarations failed in the REPL. Add `Stdout` import and flush stdout before reading stdin since `System.print()` no longer auto-flushes. Refactor token detection to use `lexFirst()` instead of full `lex()`, and add new `Meta.compile()` method returning a fiber. Internally rename `wrenNewString` to `wrenNewStringLength` with explicit length parameter, updating all call sites across compiler, core, and meta modules.
Add a new `servedocs` Makefile target that runs `generate_docs.py --serve`, enabling continuous doc generation and local web serving. Extend the Python script with `RootedHTTPServer` and `RootedHTTPRequestHandler` classes to serve generated HTML files from the output directory, refreshing content on each request via `format_files(True)`. Also update the shebang to `python3` and change the conversion print message from "converted" to "Built".
The comment in the WrenForeignClassMethods struct incorrectly referenced
`wrenAllocateForeign` instead of the actual function `wrenSetSlotNewForeign()`.
This fixes the documentation to match the correct API function name that must
be called exactly once inside the allocate callback body.
Remove the automatic fflush(stdout) call from the write() function in vm.c, which was causing gratuitous performance overhead on every System.write() invocation. Introduce a new Stdout class with a foreign static flush() method that delegates to fflush(stdout), giving users explicit control over when output is flushed. Update the io module registration in modules.c to include the Stdout class and bump MAX_CLASSES_PER_MODULE from 5 to 6. Add documentation for Stdout in doc/site/modules/io/stdout.markdown, update the io index and template to list the new class, and add flush() usage notes to the Stdin readByte() and readLine() docs. Implement the stdoutFlush() foreign function in io.c.
Rename the internal `ClassCompiler` struct typedef to `ClassInfo` to better reflect its purpose as a descriptor for the enclosing class context during compilation. Update all variable declarations, function signatures, and references throughout `wren_compiler.c` to use the new type name, including in `getEnclosingClass`, `field`, `super_`, and `declareMethod` functions. This resolves issue #469 by eliminating the misleading "Compiler" suffix from a type that tracks class metadata rather than compilation state.
These test cases cover malformed class declarations, invalid string literals,
corrupted identifiers, and various edge cases that previously caused crashes
but now produce expected parse errors.
The compiler previously relied on `match(TOKEN_RIGHT_BRACE)` in the while loop condition, which consumed the brace and made it impossible to detect a missing one. Changed to peek-based loop termination and added explicit `consume` call for the closing brace, ensuring proper error reporting when '}' is absent. Added regression test for issue #429.
Add parentheses around the `count` parameter in both `ALLOCATE_FLEX` and
`ALLOCATE_ARRAY` macro definitions to prevent operator precedence bugs when
the argument is an expression. Without this fix, expressions like
`sizeof(type) * n + 1` would be evaluated incorrectly due to `*` having
higher precedence than `+`.
When the fiber's stack is reallocated, pointer subtraction between old and new stack
addresses is undefined behavior. Instead of computing an offset and adding it to each
pointer, recalculate each pointer relative to the new stack base using well-defined
arithmetic within the single array. This affects the apiStack, call frame stackStart
pointers, open upvalue values, and the stackTop pointer.
Implement the `round` method on the Num class in the Wren VM, which rounds a number to the nearest integer using standard rounding rules (away from zero for .5). The implementation adds a new `DEF_NUM_FN(round, round)` macro invocation in `wren_core.c`, registers the corresponding primitive `num_round`, and includes comprehensive documentation in the core module reference. A new test file `test/core/number/round.wren` validates behavior for positive, negative, zero, and fractional values. Also adds Kyle Charters to the AUTHORS file.
In runtimeError, walk the entire fiber call chain and set each fiber's error field to the same error value, unhooking callers as we go. Previously only the immediate caller was considered, leaving intermediate fibers in an inconsistent state. This ensures that when a fiber aborts, every fiber in the chain up to (but not including) the one that catches the error via try() is properly marked as errored.
Also remove the old nested fiber tests from try.wren and add a new comprehensive test (try_through_call.wren) that verifies the percolation behavior through multiple levels of fiber calls and confirms that intermediate fibers carry the error while the catching fiber and its callers remain clean.
Move the aborted fiber check in runFiber to execute before the "already called" check, ensuring a clear error message for aborted fibers. In runtimeError, replace the single caller unhook with a loop that walks the fiber chain, preserving the callerIsTrying flag and only unhooking fibers until a try-catch boundary is found, enabling proper error propagation through nested fiber calls. Add test cases for one and two levels of nested fiber try calls.
The off-by-one error caused the arch argument to be read from args[2] instead of args[1], which would either read an out-of-bounds index or pick the wrong argument when exactly two arguments were provided. This fix ensures the optional architecture parameter is correctly retrieved from the second command-line argument.
Remove dynamic dependency download during build by checking in libuv source and GYP build tooling. Update .gitignore to exclude intermediate build artifacts from the vendored deps directory, adjust Makefile to remove the now-unnecessary cleanall target and deps folder cleanup, and add the full libuv project tree including its own .gitignore, .mailmap, AUTHORS, CONTRIBUTING.md, ChangeLog, and LICENSE files.
The embedding API reference link was pointing to the old `embedding-api.html` path, which would result in a 404. Updated it to the correct `embedding/` path to ensure users can access the documentation.
The deleted test file `test/core/range/floating_point.wren` was a direct copy of the integer range test file, testing `.from` on integer ranges rather than actual floating-point ranges. This removal resolves issue #438 by eliminating the redundant test that provided no coverage for floating-point range behavior.
Delete the entire util/msvc2013 directory containing Visual Studio 2013 solution and project files. Add new util/vs2017 directory with updated solution targeting Visual Studio 2017 (v141 toolset) and supporting both Win32 and x64 platforms. Update .gitignore to ignore .vs/ and obj/ directories. Modify build_libuv_windows in util/libuv.py to accept an arch parameter and pass x86 or x64 to vcbuild.bat.
When a module is loaded via the host's loadModuleFn callback, the source string is dynamically allocated and ownership is transferred to the VM. Previously, this memory was never freed after compilation, causing a leak. This change introduces an `allocatedSource` flag that tracks whether the source was dynamically allocated (host-provided) or a static constant (built-in modules like "random"). After `loadModule` completes, the source is freed if it was dynamically allocated, resolving the memory leak described in issue #430.
When the lexer encounters an invalid character, it now explicitly sets the
current token's type to TOKEN_ERROR and its length to 0. Previously, the
token retained the previous token's type, which could cause the compiler
to get stuck in an infinite loop in certain code paths.
Also improves the error message for non-ASCII invalid bytes by showing
the raw hex value instead of attempting to display the character, since
the lexer operates on raw bytes without UTF-8 decoding.
Adds regression test for issue #428 that was crashing the compiler with
an out-of-bounds memory access due to this loop.
Add a new test suite for user data get/set operations on WrenVM, including
a Wren script that validates default NULL userData, retrieval after VM
creation, and mutation via wrenSetUserData. Also reorder the vm parameter
in WrenErrorFn callback signature and all its call sites to match the
convention used by other callback types, updating the typedef, the
reportError implementation, and all invocations in the compiler and
debug modules. Include the new source and header files in the Xcode
project build configuration.
Extend WrenConfiguration with a void* userData field for attaching arbitrary state to a VM instance. Add wrenGetUserData and wrenSetUserData accessor functions. Modify the WrenErrorFn typedef and all call sites (compiler, debug, VM) to pass the WrenVM pointer as the final argument, enabling error handlers to access VM context.
Add type and range checks to Sequence.skip() and Sequence.take() methods, aborting with "Count must be a non-negative integer." when count is not a non-negative integer. Rename split() parameter from 'delim' to 'delimiter' and update its error message from "Argument must be a non-empty string." to "Delimiter must be a non-empty string." for consistency. Remove old behavior allowing negative counts and add new test files for skip/take validation errors.
Implement `skip(count)` and `take(count)` on the Sequence class, backed by new SkipSequence and TakeSequence wrapper classes that lazily skip or limit iteration. Includes documentation in the sequence module markdown, core source definitions, and unit tests for edge cases (zero, negative, and overflow counts).
Fix three occurrences of the misspelled word "seperator" in the string module
documentation, updating both the function signature header and the descriptive
text in the split() method section.
Implement String.split(delim) and String.replace(from, to) methods in the core Wren VM. split returns a list of substrings separated by the given delimiter, while replace returns a new string with all occurrences of the old substring replaced by the new one. Both methods validate arguments (must be non-empty strings) and abort the fiber on invalid input. Include comprehensive test suites covering basic usage, multiple matches, non-ASCII characters, 8-bit clean data, and error cases. Update the string.markdown documentation with usage examples for both methods.
Reworded opening paragraph to improve flow and remove redundant phrasing.
Updated description of wrenInitConfiguration to emphasize prevention of uninitialized configuration.
Changed "the VM" to "Wren" in loadModuleFn description for consistent terminology.
Replace the stub "TODO: Write these docs." in configuring-the-vm.markdown with comprehensive documentation covering all WrenConfiguration fields including binding callbacks, memory management, and error handling. Also fix the typo "phyisically" to "physically" in the wren.h header comment and remove trailing whitespace on several lines.
Adds `log` method to Num class via DEF_NUM_FN macro, and `pow(_)` method via a new DEF_PRIMITIVE that calls C's `pow()`. Registers both primitives in wrenInitializeCore. Includes new test files `test/core/number/log.wren` and `test/core/number/pow.wren` covering basic cases and edge values (negative input for log, zero exponent for pow). Also adds Sven Bergström to AUTHORS.
The previous implementation used strtol() for parsing hex literals, which only supports up to 32-bit values on 32-bit architectures. This caused hex literals larger than 0x7FFFFFFF to overflow or fail. Changed to strtoll() which guarantees 64-bit parsing regardless of platform word size.
Also updated the error message to include the actual sizeof(long int) for debugging, uncommented the previously disabled 64-bit hex literal examples in syntax.wren, and added a test case for 0xdeadbeef to verify the fix works correctly.
Remove the two debug print statements for WREN_APP and TEST_APP paths that were cluttering test runner output. Update the status line format to include the relative path of the test application being run, making it clearer which binary's results are being reported when running multiple test suites.
The fpclassify() macro is not available in C++98, causing build failures on older compilers. This commit replaces the switch on fpclassify with explicit isnan() and isinf() checks, handling NaN and infinity edge cases directly. The logic for signed infinity is preserved by comparing value > 0.0 instead of using signbit(), which is also not guaranteed in C++98. This aligns with the existing usage of isnan/isinf elsewhere in the Wren codebase.
The bitwise operator tests used decimal literals exceeding u32 range, causing undefined behavior when converting double to u32 on some compilers. This replaces those values with hex equivalents that fit within u32, removes tests for values past max u32, and adds TODO comments for future handling. Also simplifies the `num_bitwiseNot` primitive by removing the intermediate `wideVal` variable, and fixes minor formatting in `util/test.py`.
The previous formula `bytesAllocated * (100 + growthPercent) / 100` incorrectly
treated `heapGrowthPercent` as a multiplier on total heap size relative to
in-use memory, rather than as an additive percentage of current allocation.
The new formula `bytesAllocated + (bytesAllocated * growthPercent / 100)`
correctly computes the next GC threshold as the current allocation plus a
configured percentage of that allocation, matching the documented intent of
the configuration parameter.
Add explanation that single-expression blocks cannot contain statements
(like class, for, if, import, return, var, while) since they don't
produce values, and show how to use a newline to create a block with
a single statement instead. Fixes#390.
Refine phrasing in the embedding guide for better readability: change "generate a runtime error" to "generate runtime errors", reword "What that gives you in return" to "What you get in return", and replace "get to the public header" with "find the public header" for more natural expression.
The test for chained assignment was incorrectly printing variable `a` twice instead of printing `b` for the second assertion. This fix changes the second `System.print(a)` to `System.print(b)` so the test correctly verifies that both `a` and `b` hold the value "chain" after the chained assignment `a = b = "chain"`.
When invoked with exactly two arguments where the second is "--version", the
main function now prints the version string (e.g. "wren 0.4.0") and exits
successfully before any other initialization occurs.
Change the C implementation of `num_smallest` primitive to return `DBL_MIN` instead of `DBL_EPSILON`, correcting the value from the machine epsilon (difference between 1.0 and next representable value) to the smallest positive representable double. Update the documentation in `num.markdown` to describe `Num.smallest` as "the smallest positive representable numeric value" and `Num.largest` as "the largest representable numeric value". Add test files `largest.wren` and `smallest.wren` verifying the expected output values.
Add two new static properties to the Num class: `Num.largest` returns `DBL_MAX`
and `Num.smallest` returns `DBL_EPSILON`. Include corresponding documentation
in the core module markdown and register the primitives in the VM core
initialization.
The function signature in both the declaration (wren_utils.h) and definition
(wren_utils.c) is updated to accept a const SymbolTable* instead of a
non-const pointer, enabling callers to pass const-qualified symbol tables
without casting away constness.
Replace the linear scan of the function's constant table in `addConstant()` with a hash map lookup using a new `ObjMap* constants` field on the `Compiler` struct. This avoids O(n) performance degradation when compiling functions with many constants, such as large literal-heavy functions. The map stores constant values as keys and their index as numeric values, lazily initialized on first insertion. Also adds OBJ_FN hashing support in `hashObject()` to allow bare function objects as map keys internally, and includes a regression test with 65,550 constant entries to validate the fix.
Add linear scan in addConstant() to check for duplicate constant values before appending. This reduces memory usage for repeated constants and enables support for functions with large numbers of repeated entries. Includes TODO noting O(n) complexity for future optimization.
Remove the List.new(_) primitive that initialized a list with null elements, merging its functionality into List.filled(_,_) which now accepts both size and default value. Add validation for negative size with a clear error message, and introduce comprehensive test coverage for filled list creation including edge cases for negative size, non-integer size, and non-numeric size arguments.
Add two new list constructors to the Wren VM: `List.new(size)` creates a list of given size with null elements, and `List.filled(size, value)` creates a list with all elements initialized to a specified default value. Includes corresponding C primitives `list_newWithSize` and `list_newWithSizeDefault`, registration in `wrenInitializeCore`, and a test file verifying both constructors produce correct counts and element values. Also adds Max Ferguson to AUTHORS.
Simplify arithmetic in wrenStringFind by removing startIndex offset from loop range and using start directly as the initial index. Rename internal primitives to string_indexOf1 and string_indexOf2 for clarity. Add validateIndex helper to clamp negative start values and reject out-of-bounds indices. Update documentation to describe the new start parameter behavior. Add comprehensive test suite for indexOf with start, including negative offsets, edge cases, and error conditions.
Add a new `indexOf(_,_)` primitive that accepts a start index for searching within strings, and modify the existing `wrenStringFind` function to accept and use a start index parameter. The existing `indexOf(_)` and `contains(_)` methods now pass 0 as the start index to maintain backward compatibility. The Boyer-Moore-Horspool search algorithm is updated to offset comparisons by the start index and return the absolute position in the original string. New test cases verify correct behavior with various start index values including edge cases where the start index exceeds string length.
The `DEBUG_TRACE` macro was using `fn->bytecode` instead of `fn->code.data` to compute the instruction offset for `wrenDumpInstruction`. This caused incorrect offset values when dumping bytecode during interpreter tracing, as the bytecode pointer is stored in the `code.data` field of the function object.
Bob Nystrom