Ensure consistent locale settings and optimize IR compilation

.gitignore

@@ -56,3 +56,4 @@ runtime/dgsh-tee
 .vscode/*
 .idea/*
 .vs/*
+venv/

compiler/pash.py

@@ -17,6 +17,9 @@
 def main():
     ## Parse arguments
     args, shell_name = parse_args()
+
+    check_locale_settings()
+
     ## If it is interactive we need a different execution mode
     ##
     ## The user can also ask for an interactive mode irregardless of whether pash was invoked in interactive mode.
@@ -164,5 +167,38 @@ def execute_script(compiled_script_filename, command, arguments, shell_name):
     return exec_obj.returncode
 
 
+
+def check_locale_settings():
+    """
+    Checks if locale settings are set correctly to avoid execution inconsistencies.
+    """
+
+    locale_vars = {
+        "LC_ALL": os.environ.get("LC_ALL", "Not Set"),
+        "LANG": os.environ.get("LANG", "Not Set"),
+        "LC_CTYPE": os.environ.get("LC_CTYPE", "Not Set"),
+        "LC_COLLATE": os.environ.get("LC_COLLATE", "Not Set"),
+        "LC_NUMERIC": os.environ.get("LC_NUMERIC", "Not Set"),
+    }
+
+    # If LC_ALL is set, it overrides all locale settings
+    if locale_vars["LC_ALL"] not in ["Not Set", "C", "C.UTF-8"]:
+        log(f"WARNING: LC_ALL is set to '{locale_vars['LC_ALL']}', which overrides all other locale settings.")
+        log(f"Consider setting LC_ALL=C for full consistency.")
+        return  # No need to check others if LC_ALL is set
+
+    # Warn if LC_COLLATE is neither C nor C.UTF-8
+    if locale_vars["LC_COLLATE"] not in ["C", "C.UTF-8"]:
+        log(f"WARNING: LC_COLLATE is set to '{locale_vars['LC_COLLATE']}'. This may affect sorting behavior.")
+        log(f"Consider setting LC_COLLATE=C for strict ASCII ordering.")
+
+    # Warn if LC_NUMERIC is neither C nor C.UTF-8
+    if locale_vars["LC_NUMERIC"] not in ["C", "C.UTF-8"]:
+        log(f"WARNING: LC_NUMERIC is set to '{locale_vars['LC_NUMERIC']}'. This may affect numerical operations.")
+        log(f"Consider setting LC_NUMERIC=C for consistent behavior.")
+
+    log("Locale settings checked. To avoid performance inconsistencies, use `LC_ALL=C.UTF-8`.")
+    pid = os.getpid()
+
 if __name__ == "__main__":
     main()

compiler/pash_compilation_server.py

@@ -39,7 +39,19 @@ def parse_args():
 
 
 # Initialize the daemon
-
+initial_env_vars = {}
+
+def store_initial_env():
+    """Stores initial locale environment variables at startup."""
+    global initial_env_vars
+    initial_env_vars = {
+        "LC_ALL": os.environ.get("LC_ALL", "Not Set"),
+        "LANG": os.environ.get("LANG", "Not Set"),
+        "LC_CTYPE": os.environ.get("LC_CTYPE", "Not Set"),
+        "LC_COLLATE": os.environ.get("LC_COLLATE", "Not Set"),
+        "LC_NUMERIC": os.environ.get("LC_NUMERIC", "Not Set")
+    }
+    log(f"Initial locale settings: {initial_env_vars}")
 
 def init():
     ## Set the logging prefix
@@ -53,7 +65,7 @@ def init():
         config.load_config(args.config_path)
 
     pash_compiler.runtime_config = config.config["distr_planner"]
-
+    store_initial_env()
     return args
 
 
@@ -511,13 +523,31 @@ def run(self):
         self.connection_manager.close()
         shutdown()
 
+def check_final_env():
+    """Checks if locale environment variables changed during execution."""
+    final_env_vars = {
+        "LC_ALL": os.environ.get("LC_ALL", "Not Set"),
+        "LANG": os.environ.get("LANG", "Not Set"),
+        "LC_CTYPE": os.environ.get("LC_CTYPE", "Not Set"),
+        "LC_COLLATE": os.environ.get("LC_COLLATE", "Not Set"),
+        "LC_NUMERIC": os.environ.get("LC_NUMERIC", "Not Set")
+    }
+
+    if final_env_vars != initial_env_vars:
+        log("WARNING: Locale environment variables changed during execution!")
+        log(f"Initial: {initial_env_vars}")
+        log(f"Final: {final_env_vars}")
+
 
 def shutdown():
-    ## There may be races since this is called through the signal handling
+    """Shuts down the PaSh daemon."""
+    check_final_env()  
     log("PaSh daemon is shutting down...")
     log("PaSh daemon shut down successfully...")
 
 
+
+
 def main():
     args = init()
     if args.distributed_exec:

compiler/pash_compiler.py

@@ -2,7 +2,6 @@
 import pickle
 import traceback
 from datetime import datetime
-
 from sh_expand import env_vars_util
 from sh_expand.expand import ExpansionError