Port chatbot from C to Zig with comprehensive benchmarking

.github/workflows/zig.yml

@@ -0,0 +1,39 @@
+name: Build & Tests
+
+on:
+  push:
+    branches: [ main, zig-port ]
+  pull_request:
+    branches: [ main, zig-port ]
+
+jobs:
+  c-build:
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Build C version
+      run: cd c && make clean && make
+
+    - name: Test C executable runs
+      run: cd c && echo -e "hi\npython\nexit" | timeout 5 ./chat || true
+
+  zig-build-and-test:
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v4
+
+    - uses: goto-bus-stop/setup-zig@v2
+      with:
+        version: 0.15.2
+
+    - name: Build Zig chatbot
+      run: cd zig && mkdir -p zig-out/bin && zig build-exe src/main.zig -femit-bin=zig-out/bin/chat
+
+    - name: Run Zig tests
+      run: cd zig && zig test src/chatbot.zig
+
+    - name: Test Zig executable runs
+      run: cd zig && echo -e "hi\npython\nexit" | timeout 5 ./zig-out/bin/chat || true

.gitignore

@@ -23,3 +23,7 @@ chat
 *.i*86
 *.x86_64
 *.hex
+
+# Zig build artifacts
+.zig-cache/
+zig-out/

BENCHMARK.md

@@ -0,0 +1,48 @@
+# Chatbot Build Benchmark Results
+
+Benchmark comparing C and Zig implementations of the chatbot.
+
+**System:** macOS arm64  
+**Date:** 2026-01-12
+
+## Results
+
+| Metric | C | Zig | Difference |
+|--------|---|-----|-----------|
+| Build Time | 93ms | 710ms | Zig is 7.63x slower |
+| Executable Size | 33K | 1.3M | Zig is 39x larger |
+
+## Analysis
+
+### Build Time
+- **C (93ms)**: Fast compilation using GCC with minimal optimization
+- **Zig (710ms)**: Longer compilation time due to Zig's more comprehensive compiler
+
+The C version compiles significantly faster due to:
+- Simpler compilation pipeline
+- No build system overhead (direct gcc command)
+- Minimal type checking and analysis
+
+### Executable Size
+- **C (33K)**: Small, minimal runtime
+- **Zig (1.3M)**: Larger due to Zig's standard library and runtime
+
+The Zig executable is larger because:
+- Zig stdlib is embedded in the binary
+- More comprehensive runtime features
+- GeneralPurposeAllocator adds overhead
+
+## Notes
+
+- Both versions are unoptimized builds
+- C build uses `-std=c11 -Wall -Wextra -pedantic`
+- Zig build uses default optimization level
+- Times are from cold builds (no cache)
+
+## Running the Benchmark
+
+```bash
+./benchmark.sh
+```
+
+This will clean, rebuild both versions, and compare build times and sizes.

benchmark.sh

@@ -0,0 +1,69 @@
+#!/bin/bash
+
+set -e
+
+echo "=========================================="
+echo "Chatbot Build Benchmark"
+echo "=========================================="
+echo ""
+
+# C Version Benchmark
+echo "C Version Benchmark"
+echo "------------------------------------------"
+
+cd c
+echo "Cleaning..."
+make clean > /dev/null 2>&1
+
+echo "Building C version..."
+C_START=$(date +%s%N)
+make > /dev/null 2>&1
+C_END=$(date +%s%N)
+C_TIME=$(( (C_END - C_START) / 1000000 ))  # Convert to milliseconds
+
+C_SIZE=$(ls -lh chat | awk '{print $5}')
+
+echo "Build time: ${C_TIME}ms"
+echo "Executable size: ${C_SIZE}"
+
+cd ..
+echo ""
+
+# Zig Version Benchmark
+echo "Zig Version Benchmark"
+echo "------------------------------------------"
+
+cd zig
+echo "Cleaning..."
+rm -rf zig-out .zig-cache > /dev/null 2>&1
+mkdir -p zig-out/bin
+
+echo "Building Zig version..."
+ZIG_START=$(date +%s%N)
+zig build-exe src/main.zig -femit-bin=zig-out/bin/chat > /dev/null 2>&1
+ZIG_END=$(date +%s%N)
+ZIG_TIME=$(( (ZIG_END - ZIG_START) / 1000000 ))  # Convert to milliseconds
+
+ZIG_SIZE=$(ls -lh zig-out/bin/chat | awk '{print $5}')
+
+echo "Build time: ${ZIG_TIME}ms"
+echo "Executable size: ${ZIG_SIZE}"
+
+cd ..
+echo ""
+
+# Comparison
+echo "Comparison"
+echo "------------------------------------------"
+if [ $C_TIME -lt $ZIG_TIME ]; then
+  RATIO=$(echo "scale=2; $ZIG_TIME / $C_TIME" | bc)
+  echo "C version is ${RATIO}x faster"
+else
+  RATIO=$(echo "scale=2; $C_TIME / $ZIG_TIME" | bc)
+  echo "Zig version is ${RATIO}x faster"
+fi
+
+echo ""
+echo "Summary:"
+echo "  C:   ${C_TIME}ms, ${C_SIZE}"
+echo "  Zig: ${ZIG_TIME}ms, ${ZIG_SIZE}"

Makefile → c/Makefile

@@ -2,4 +2,4 @@ all:
 	gcc -std=c11 -Wall -Wextra -pedantic src/chatbot.c -o chat
 
 clean:
-	rm -rf *o chat
+	rm -rf *.o chat

readme.md

@@ -1,46 +1,81 @@
 # Chatbot
 
-[![Build Status](https://travis-ci.org/vinitkumar/chatbot.svg?branch=master)](https://travis-ci.org/vinitkumar/chatbot)
+A simple keyword-based chatbot with implementations in both C and Zig.
 
-Chatbot is a simple bot written in C. The base algorithm is that our conversation is based on keywords. The bot seaches for this keyword and present a suitable response on basis of that keyword.
+## Benchmarks
 
-## Installation
+### Build Time & Size
 
-Installation is very easy via `MakeFile` :
+See [BENCHMARK.md](BENCHMARK.md) for detailed build metrics.
 
-Just run:
+| Metric | C | Zig | Difference |
+|--------|---|-----|-----------|
+| Build Time | 93ms | 710ms | Zig is 7.63x slower |
+| Executable Size | 33K | 1.3M | Zig is 39x larger |
 
+Run build benchmark:
+```bash
+./benchmark.sh
 ```
-chatbot ➤ make
-gcc src/chatbot.c -o chat
+
+### Runtime Performance
+
+Both versions produce identical output. Runtime measurements with test inputs:
+
+| Metric | C | Zig | Difference |
+|--------|---|-----|-----------|
+| Execution Time | 286ms | 526ms | C is 1.83x faster |
+
+Run runtime benchmark:
+```bash
+./run_benchmarks.sh
 ```
 
-## Usage
+## Building
+
+### C Version
 
-After running `make all` you get a executable named as `chat`. Now run `./chat` and start chatting with the bot.
+```bash
+cd c
+make
+./chat
+```
+
+See `c/Makefile` for more details.
 
-## Demo
+### Zig Version
 
+```bash
+cd zig
+zig build-exe src/main.zig -femit-bin=zig-out/bin/chat
+./zig-out/bin/chat
 ```
-chatbot ➤ ./chat
-$ Chatbot v1.0.0!
 
-$ (user) hi
+Or use the Zig build system:
 
-$ (chatbot) hello
+```bash
+cd zig
+zig build run
+```
 
-$ (user) python
+See `zig/README.md` for more details.
 
-$ (chatbot) Yo, I love Python
+## Testing
 
-$ (user) What
+### C Version
 
-$ (chatbot) It is clear, ain't it?
+No automated tests (original implementation).
 
-$ (user) exit
+### Zig Version
+
+```bash
+cd zig
+zig test src/chatbot.zig
 ```
 
-## Issues and Pull request.
+Or:
 
-Feel free to create an issue if you notice a bug. Pull request are really
-welcome with good fixes.
+```bash
+cd zig
+zig build test
+```

run_benchmarks.sh

@@ -0,0 +1,102 @@
+#!/bin/bash
+
+set -e
+
+echo "=========================================="
+echo "Chatbot Runtime Benchmark & Testing"
+echo "=========================================="
+echo ""
+
+# Build both versions
+echo "Building both versions..."
+echo ""
+
+cd c
+make clean > /dev/null 2>&1
+make > /dev/null 2>&1
+cd ..
+
+cd zig
+rm -rf zig-out .zig-cache > /dev/null 2>&1
+mkdir -p zig-out/bin
+zig build-exe src/main.zig -femit-bin=zig-out/bin/chat > /dev/null 2>&1
+cd ..
+
+echo "Build complete!"
+echo ""
+
+# Test C version
+echo "=========================================="
+echo "C Version Testing & Benchmarking"
+echo "=========================================="
+echo ""
+echo "Running with test inputs..."
+
+C_START=$(date +%s%N)
+c/./chat < test_inputs.txt > /tmp/c_output.txt 2>&1
+C_END=$(date +%s%N)
+C_TIME=$(( (C_END - C_START) / 1000000 ))  # milliseconds
+
+echo "Output:"
+cat /tmp/c_output.txt
+echo ""
+echo "Execution time: ${C_TIME}ms"
+echo ""
+
+# Test Zig version
+echo "=========================================="
+echo "Zig Version Testing & Benchmarking"
+echo "=========================================="
+echo ""
+echo "Running with test inputs..."
+
+ZIG_START=$(date +%s%N)
+zig/zig-out/bin/./chat < test_inputs.txt > /tmp/zig_output.txt 2>&1
+ZIG_END=$(date +%s%N)
+ZIG_TIME=$(( (ZIG_END - ZIG_START) / 1000000 ))  # milliseconds
+
+echo "Output:"
+cat /tmp/zig_output.txt
+echo ""
+echo "Execution time: ${ZIG_TIME}ms"
+echo ""
+
+# Compare outputs
+echo "=========================================="
+echo "Output Comparison"
+echo "=========================================="
+echo ""
+
+if diff -q /tmp/c_output.txt /tmp/zig_output.txt > /dev/null 2>&1; then
+  echo "✓ Outputs are identical"
+else
+  echo "✗ Outputs differ"
+  echo ""
+  echo "C output:"
+  cat /tmp/c_output.txt
+  echo ""
+  echo "Zig output:"
+  cat /tmp/zig_output.txt
+fi
+
+echo ""
+
+# Runtime comparison
+echo "=========================================="
+echo "Runtime Comparison"
+echo "=========================================="
+echo ""
+
+if [ $C_TIME -lt $ZIG_TIME ]; then
+  RATIO=$(echo "scale=2; $ZIG_TIME / $C_TIME" | bc)
+  echo "C version is ${RATIO}x faster"
+else
+  RATIO=$(echo "scale=2; $C_TIME / $ZIG_TIME" | bc)
+  echo "Zig version is ${RATIO}x faster"
+fi
+
+echo ""
+echo "Summary:"
+echo "  C:   ${C_TIME}ms"
+echo "  Zig: ${ZIG_TIME}ms"
+echo ""

test_inputs.txt

@@ -0,0 +1,11 @@
+hi
+hey
+python
+light
+what
+unknown_keyword
+hear
+hello
+python
+light
+exit