coding-nim

Purpose

This skill equips the AI to handle Nim 2.x programming tasks, focusing on advanced features like macros, templates, compile-time execution, memory management (ARC/ORC), FFI for interoperability, Nimble for package handling, and systems programming. Use it to generate, debug, and optimize Nim code efficiently.

When to Use

Apply this skill for systems-level programming needing low-level control, such as embedded systems, performance-critical apps, or when integrating with C/C++ via FFI. Use it for projects requiring compile-time metaprogramming (e.g., via macros) or automatic memory management with ARC/ORC to avoid manual garbage collection.

Key Capabilities

• Macros for code generation: Define reusable code transformations at compile-time.

• Templates for type-safe string-based metaprogramming.

• Compile-time execution: Run code during compilation using static blocks.

• Memory management: Switch between ARC (automatic reference counting) and ORC (optional reference counting) via compiler flags.

• FFI: Call external libraries (e.g., C functions) without wrappers.

• Nimble: Manage dependencies and build projects like a package manager.

• Systems programming: Direct hardware access, concurrency, and cross-platform compilation.

Usage Patterns

To accomplish tasks, structure Nim code with modules and use the compiler for builds. For macros, define them in a separate proc and invoke at compile-time. When writing FFI code, use the importc pragma for C functions. For memory management, specify --gc:arc or --gc:orc flags during compilation. Always test code with nim check before full builds to catch errors early. Integrate templates for generic functions to reduce boilerplate.

Common Commands/API

Use the Nim compiler (nim ) for core operations. Compile a file: nim c --verbosity:0 -r main.nim (flags: -r for run, --verbosity:0 for minimal output). For Nimble, install packages: nimble install somepkg . Define a macro:

macro doubleIt(x: expr): stmt =
result = quote do: x * 2

Call it as echo doubleIt(5) . For FFI, import C: proc printf(format: cstring; args: varargs[pointer]) {.importc: "printf", header: "<stdio.h>".} . Config format: Use nim.cfg for settings, e.g., gcc.exe = "gcc" to specify compiler. Environment variables: Set $NIMBLE_DIR for package cache.

Integration Notes

Integrate Nim with other languages via FFI; for C++, use importcpp . To embed in a project, compile Nim code to a shared library: nim c --app:lib -d:release mylib.nim . For CI/CD, use GitHub Actions with: run: nim c --opt:speed file.nim . If auth is needed (e.g., for Nimble registry), set env vars like $NIMBLE_TOKEN for private repos. Link against external libs: Add --passL:-lssl for OpenSSL. Ensure path configurations match, e.g., add path = "/path/to/headers" in nim.cfg .

Error Handling

In Nim, use try-except blocks for runtime errors:

try:
raise newException(ValueError, "Invalid input")
except ValueError:
echo "Handled error: ", getCurrentExceptionMsg()

For compile-time errors, enable detailed output with nim c --verbosity:2 file.nim . Check for memory issues by switching to ORC: nim c --gc:orc file.nim . Parse compiler output for specifics; common flags include --warnings:on to catch undeclared vars. Log errors with echo or a logging library like chronicles .

Concrete Usage Examples

• Define and use a macro for code generation: To create a loop macro, write:

macro forEach(items: seq, body: stmt): stmt =
result = quote do: for item in items: body

Use it as: forEach(@[1, 2, 3], echo item) . This generates efficient loops at compile-time.

• Use FFI to call a C function: Import and call a C printf:

proc cPrintf(format: cstring) {.importc: "printf", header: "<stdio.h>".}
cPrintf("%s\n", "Hello from Nim")

Compile with nim c --passC:-I/usr/include file.nim to link C headers, enabling seamless integration.

Graph Relationships

• Related to cluster: coding

• Connected via tags: nim (direct link to language-specific skills), systems (links to low-level programming tools), coding (broad connections to other coding skills like coding-python or coding-c)