refactor: Rename benchmark dir, improve rsync, fix problem stats

benchmark/clone-exercism.sh

@@ -1,10 +1,10 @@
 #!/bin/bash
 
 # Create directories if they don't exist
-mkdir -p tmp.benchmark/exercism
+mkdir -p tmp.benchmarks/exercism
 
 # Change to the exercism directory
-cd tmp.benchmark/exercism
+cd tmp.benchmarks/exercism
 
 # List of languages to clone
 languages=("cpp" "go" "java" "javascript" "python" "rust")

benchmark/install-docker-ubuntu.sh

@@ -49,11 +49,6 @@ fi
 
 # Print success message and instructions
 echo "Docker installation completed successfully!"
-echo -e "\n*** IMPORTANT: Group changes require a complete session refresh."
-echo "Please run these commands in sequence:"
-echo "1. sudo systemctl restart docker"
-echo "2. exit"
-echo -e "Then log back in and docker commands should work without sudo\n"
 
 # Start Docker service
 echo "Starting Docker service..."
@@ -66,8 +61,3 @@ docker --version
 
 echo "Docker Compose version:"
 docker compose version
-
-echo "Docker service status:"
-sudo systemctl status docker --no-pager
-
-EOF

benchmark/problem_stats.py

@@ -23,9 +23,12 @@ def get_dirs_from_leaderboard():
 def load_results(dirname):
     """Load all result files from a benchmark directory"""
     dirname = Path(dirname)
-    benchmark_dir = Path("tmp.benchmarks") / dirname
+
+    benchmark_dir = dirname
     if not benchmark_dir.exists():
-        return None
+        benchmark_dir = Path("tmp.benchmarks") / dirname
+        if not benchmark_dir.exists():
+            return None
 
     all_results = []
     parse_errors = []  # Track which exercises had parse errors for this model
@@ -70,8 +73,11 @@ def analyze_exercise_solutions(dirs=None, topn=None, copy_hard_set=False):
     valid_entries = []
     parse_errors_by_model = {}  # Track which exercises had parse errors for each model
 
+    dump(dir_entries)
+
     for dirname, model in dir_entries:
         results_data = load_results(dirname)
+
         if results_data:
             results, model_parse_errors = results_data
             parse_errors_by_model[model] = set(model_parse_errors)
@@ -299,7 +305,7 @@ def analyze_exercise_solutions(dirs=None, topn=None, copy_hard_set=False):
     if copy_hard_set:
         # Create hard set directory
         src_dir = Path("tmp.benchmarks/exercism")
-        dst_dir = Path("tmp.benchmarks/exercism-hard-set")
+        dst_dir = Path("tmp.benchmarks/exercism-polyglot")
 
         if dst_dir.exists():
             print(f"\nError: Destination directory {dst_dir} already exists")
@@ -340,7 +346,7 @@ if __name__ == "__main__":
     parser.add_argument(
         "--copy-hard-set",
         action="store_true",
-        help="Copy hard set problems to tmp.benchmarks/exercism-hard-set",
+        help="Copy hard set problems to tmp.benchmarks/exercism-polygot",
     )
     args = parser.parse_args()
 

benchmark/rsync.sh

@@ -1,5 +1,7 @@
 #!/bin/bash
 
+set -e
+
 if [ $# -ne 1 ]; then
     echo "Usage: $0 user@host"
     exit 1
@@ -19,10 +21,13 @@ ssh "$DEST" "mkdir -p ~/aider"
 
 # Sync the repository
 rsync -avz --delete \
-    --exclude='.git/' \
     --exclude-from="$EXCLUDE_FILE" \
     "$REPO_ROOT/" \
     "$DEST:~/aider/"
 
+rsync -a .env .bash_history .gitignore "$DEST:~/aider/."
+
+rsync -a ~/dotfiles/screenrc "$DEST:.screenrc"
+
 # Clean up
 rm "$EXCLUDE_FILE"