refactor: Simplify entry points. Suppress unneeded stdout and font errors

Author: Linked-Liszt

compose.yaml

@@ -10,7 +10,7 @@ services:
       - ./configs:/configs:ro
       - ./src/downloader:/app:ro
     user: "${UID}:${GID}"
-    command: python /app/main.py --config /configs/download.yaml
+    command: python /app/downloader.py --config /configs/download.yaml
     restart: "no"
 
   simulator:
@@ -23,5 +23,5 @@ services:
     environment:
       - MPLCONFIGDIR=/tmp/matplotlib
     user: "${UID}:${GID}"
-    command: python -m simulator.run --config /app/configs/simulator.yaml
+    command: python -m simulator.diffraction_generator --config /app/configs/simulator.yaml
     restart: "no"

docker/downloader.Dockerfile

@@ -13,4 +13,4 @@ RUN pip install --no-cache-dir -r /tmp/requirements.txt
 # Code and configs are mounted at runtime via docker-compose volumes:
 #   - ./src/downloader -> /app
 #   - ./configs -> /configs
-CMD ["python", "/app/main.py", "--config", "/configs/download.yaml"]
+CMD ["python", "/app/downloader.py", "--config", "/configs/download.yaml"]

docker/simulator.Dockerfile

@@ -9,12 +9,16 @@ SHELL ["bash", "-lc"]
 ENV MAMBA_DEFAULT_ENV=sim
 ENV PATH="/opt/conda/envs/sim/bin:${PATH}"
 ENV GSAS_II_PATH=/opt/conda/envs/sim/GSAS-II
+ENV XDG_CACHE_HOME=/app/.cache
+ENV MPLCONFIGDIR=/app/.cache/matplotlib
 
 
 
 # App layout
 WORKDIR /app
+# Ensure writable cache directories for fontconfig and matplotlib
+RUN mkdir -p /app/.cache/fontconfig /app/.cache/matplotlib && chmod -R 777 /app/.cache
 COPY src/simulator /app/simulator
 
 # One-shot job entrypoint
-CMD ["python", "-m", "simulator.run", "--config", "/app/configs/simulator.yaml"]
+CMD ["python", "-m", "simulator.diffraction_generator", "--config", "/app/configs/simulator.yaml"]

src/downloader/main.py src/downloader/downloader.pysrc/downloader/main.py renamed to src/downloader/downloader.py

@@ -6,6 +6,9 @@
 import time
 import shutil
 from tqdm import tqdm
+import argparse
+import yaml
+from typing import Dict, Any
 
 # Container: insert parent GSAS-II directory and import package
 GSAS_II_PARENT = os.environ.get("GSAS_II_PATH", "/opt/conda/envs/sim/GSAS-II")
@@ -101,7 +104,7 @@ def run(self, n_sims_per_file, master_seed=12345, cleanup_worker_dirs=True, **kw
         simulation_start_time = time.time()
         success_count = 0
 
-        with Pool(self.n_parallel_sims, maxtasksperchild=1000) as p:
+        with Pool(self.n_parallel_sims, maxtasksperchild=1000, initializer=simulation_worker.suppress_worker_stdout) as p:
             results = tqdm(p.imap_unordered(simulation_worker.run_single_simulation, tasks), total=total_jobs_to_run)
             for result in results:
                 if result: success_count += 1
@@ -129,3 +132,91 @@ def run(self, n_sims_per_file, master_seed=12345, cleanup_worker_dirs=True, **kw
         print(f"Throughput:            {sims_per_sec:.2f} simulations/sec")
         print(f"Avg. Time/Simulation:  {avg_time_per_sim:.3f} seconds")
         print("---------------------------")
+
+
+def load_config(path: Path) -> Dict[str, Any]:
+    if not path.exists():
+        raise FileNotFoundError(f"Config file not found: {path}")
+    with open(path, "r") as f:
+        cfg = yaml.safe_load(f) or {}
+    return cfg
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description="AlphaDiffract simulator runner")
+    parser.add_argument("--config", type=str, default="/app/configs/simulator.yaml", help="Path to YAML config file")
+    return parser.parse_args()
+
+
+def main() -> None:
+    args = parse_args()
+    cfg = load_config(Path(args.config))
+
+    # Paths
+    input_directory = str(Path(cfg["input_directory"]))
+    output_directory = str(Path(cfg["output_directory"]))
+    instprm_file = str(Path(cfg["instprm_file"]))
+    error_directory = str(Path(cfg["error_directory"]))
+    worker_base_dir = str(Path(cfg["worker_base_dir"]))
+
+    # Execution controls
+    parallel_jobs = int(cfg["parallel_jobs"])
+    sims_per_file = int(cfg["sims_per_file"])
+    master_seed = int(cfg["master_seed"])
+    cleanup_worker_dirs = bool(cfg["cleanup_worker_dirs"])
+
+    # Parameter ranges
+    ranges = {
+        "strain_range": tuple(map(float, cfg["strain_range"])),
+        "size_range": tuple(map(float, cfg["size_range"])),
+        "U_range": tuple(map(float, cfg["U_range"])),
+        "V_range": tuple(map(float, cfg["V_range"])),
+        "W_range": tuple(map(float, cfg["W_range"])),
+        "st_range": tuple(map(float, cfg["st_range"])),
+        "en_range": tuple(map(float, cfg["en_range"])),
+        "Npoints_range": tuple(map(int, cfg["Npoints_range"])),
+        "scaler_range": tuple(map(float, cfg["scaler_range"])),
+        "wl_range": tuple(map(float, cfg["wl_range"])),
+        "proportional_noise_range": tuple(map(float, cfg["proportional_noise_range"])),
+        "constant_noise_range": tuple(map(float, cfg["constant_noise_range"])),
+    }
+
+    # Log planned operation
+    print("\n--- Simulator configuration ---")
+    print(f"Input directory:            {input_directory}")
+    print(f"Output directory:           {output_directory}")
+    print(f"Error directory:            {error_directory}")
+    print(f"Worker base dir:            {worker_base_dir}")
+    print(f"Instrument file:            {instprm_file}")
+    print(f"Parallel jobs:              {parallel_jobs}")
+    print(f"Sims per file:              {sims_per_file}")
+    print(f"Master seed:                {master_seed}")
+    print(f"Cleanup worker dirs:        {cleanup_worker_dirs}")
+    print("Parameter ranges:")
+    for k, v in ranges.items():
+        print(f"  {k}: {v}")
+    print("--------------------------------\n")
+
+    # Run simulations
+    try:
+        gen = DiffractionGenerator(
+            input_dir=input_directory,
+            output_dir=output_directory,
+            instprm_file=instprm_file,
+            n_parallel_sims=parallel_jobs,
+            error_dir=error_directory,
+            worker_base_dir=worker_base_dir,
+        )
+        gen.run(
+            n_sims_per_file=sims_per_file,
+            master_seed=master_seed,
+            cleanup_worker_dirs=cleanup_worker_dirs,
+            **ranges,
+        )
+    except Exception as e:
+        print(f"FATAL: Simulator run failed: {e}", file=sys.stderr)
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()

src/simulator/diffraction_generator.py

@@ -17,6 +17,24 @@
 except ImportError:
     sys.exit(f"FATAL ERROR: Could not import GSASII.GSASIIscriptable. Checked parent '{GSAS_II_PARENT}'. Ensure gsas2pkg is installed or set GSAS_II_PATH correctly.")
 
+def suppress_worker_stdout():
+    """
+    Redirect stdout to /dev/null in worker processes without affecting stderr.
+    This silences Python-level prints and C-extension writes to fd=1.
+    """
+    # Redirect Python-level prints
+    try:
+        sys.stdout = open(os.devnull, 'w')
+    except Exception:
+        pass
+    # Redirect OS-level fd 1 (stdout) to /dev/null for any C-level writes
+    try:
+        devnull_fd = os.open(os.devnull, os.O_WRONLY)
+        os.dup2(devnull_fd, 1)
+        os.close(devnull_fd)
+    except Exception:
+        pass
+
 # --- START: Metadata Parsing Logic ---
 # This section contains all functions needed to parse metadata directly from
 # the source .cif and .dif/.txt files, which is more reliable than using