feat(): new feature

This commit is contained in:
2026-01-29 15:59:08 +01:00
parent e9cca3daeb
commit c365b4fa6b
17 changed files with 617 additions and 83 deletions

193
backend/profile_manager.py Normal file
View File

@@ -0,0 +1,193 @@
import os
import json
import logging
from typing import Dict, List, Tuple, Optional
from profile_cache import get_cache_key
logger = logging.getLogger(__name__)
class ProfileManager:
def __init__(self, profiles_root: str = "profiles"):
# Assuming profiles_root is relative to backend or absolute
if not os.path.isabs(profiles_root):
base_dir = os.path.dirname(os.path.abspath(__file__))
self.profiles_root = os.path.join(base_dir, profiles_root)
else:
self.profiles_root = profiles_root
if not os.path.exists(self.profiles_root):
logger.warning(f"Profiles root not found: {self.profiles_root}")
def get_profiles(self, machine: str, filament: str, process: str) -> Tuple[Dict, Dict, Dict]:
"""
Main entry point to get merged profiles.
Args:
machine: e.g. "Bambu Lab A1 0.4 nozzle"
filament: e.g. "Bambu PLA Basic @BBL A1"
process: e.g. "0.20mm Standard @BBL A1"
"""
# Try cache first (although specific logic is needed if we cache the *result* or the *files*)
# Since we implemented a simple external cache helper, let's use it if we want,
# but for now we will rely on internal logic or the lru_cache decorator on a helper method.
# But wait, the `get_cached_profiles` in profile_cache.py calls `build_merged_profiles` which is logic WE need to implement.
# So we should probably move the implementation here and have the cache wrapper call it,
# OR just implement it here and wrap it.
return self._build_merged_profiles(machine, filament, process)
def _build_merged_profiles(self, machine_name: str, filament_name: str, process_name: str) -> Tuple[Dict, Dict, Dict]:
# We need to find the files.
# The naming convention in OrcaSlicer profiles usually involves the Vendor (e.g. BBL).
# We might need a mapping or search.
# For this implementation, we will assume we know the relative paths or search for them.
# Strategy: Search in all vendor subdirs for the specific JSON files.
# Because names are usually unique enough or we can specify the expected vendor.
# However, to be fast, we can map "machine_name" to a file path.
machine_file = self._find_profile_file(machine_name, "machine")
filament_file = self._find_profile_file(filament_name, "filament")
process_file = self._find_profile_file(process_name, "process")
if not machine_file:
raise FileNotFoundError(f"Machine profile not found: {machine_name}")
if not filament_file:
raise FileNotFoundError(f"Filament profile not found: {filament_name}")
if not process_file:
raise FileNotFoundError(f"Process profile not found: {process_name}")
machine_profile = self._merge_chain(machine_file)
filament_profile = self._merge_chain(filament_file)
process_profile = self._merge_chain(process_file)
# Apply patches
machine_profile = self._apply_patches(machine_profile, "machine")
process_profile = self._apply_patches(process_profile, "process")
return machine_profile, process_profile, filament_profile
def _find_profile_file(self, profile_name: str, profile_type: str) -> Optional[str]:
"""
Searches for a profile file by name in the profiles directory.
The name should match the filename (without .json possibly) or be a precise match.
"""
# Add .json if missing
filename = profile_name if profile_name.endswith(".json") else f"{profile_name}.json"
for root, dirs, files in os.walk(self.profiles_root):
if filename in files:
# Check if it is in the correct type folder (machine, filament, process)
# OrcaSlicer structure: Vendor/process/file.json
# We optionally verify parent dir
if os.path.basename(root) == profile_type or profile_type in root:
return os.path.join(root, filename)
# Fallback: if we simply found it, maybe just return it?
# Some common files might be in root or other places.
# Let's return it if we are fairly sure.
return os.path.join(root, filename)
return None
def _merge_chain(self, final_file_path: str) -> Dict:
"""
Resolves inheritance and merges.
"""
chain = []
current_path = final_file_path
# 1. Build chain
while current_path:
chain.insert(0, current_path) # Prepend
with open(current_path, 'r', encoding='utf-8') as f:
try:
data = json.load(f)
except json.JSONDecodeError as e:
logger.error(f"Failed to decode JSON: {current_path}")
raise e
inherits = data.get("inherits")
if inherits:
# Resolve inherited file
# It is usually in the same directory or relative.
# OrcaSlicer logic: checks same dir, then parent, etc.
# Usually it's in the same directory.
parent_dir = os.path.dirname(current_path)
inherited_path = os.path.join(parent_dir, inherits)
# Special case: if not found, it might be in a common folder?
# But OrcaSlicer usually keeps them local or in specific common dirs.
if not os.path.exists(inherited_path) and not inherits.endswith(".json"):
inherited_path += ".json"
if os.path.exists(inherited_path):
current_path = inherited_path
else:
# Could be a system common file not in the same dir?
# For simplicty, try to look up in the same generic type folder across the vendor?
# Or just fail for now.
# Often "fdm_machine_common.json" is at the Vendor root or similar?
# Let's try searching recursively if not found in place.
found = self._find_profile_file(inherits, "any") # "any" type
if found:
current_path = found
else:
logger.warning(f"Inherited profile '{inherits}' not found for '{current_path}' (Root: {self.profiles_root})")
current_path = None
else:
current_path = None
# 2. Merge
merged = {}
for path in chain:
with open(path, 'r', encoding='utf-8') as f:
data = json.load(f)
# Shallow update
merged.update(data)
# Remove metadata
merged.pop("inherits", None)
return merged
def _apply_patches(self, profile: Dict, profile_type: str) -> Dict:
if profile_type == "machine":
# Patch: G92 E0 to ensure extrusion reference text matches
lcg = profile.get("layer_change_gcode", "")
if "G92 E0" not in lcg:
# Append neatly
if lcg and not lcg.endswith("\n"):
lcg += "\n"
lcg += "G92 E0"
profile["layer_change_gcode"] = lcg
# Patch: ensure printable height is sufficient?
# Only if necessary. For now, trust the profile.
elif profile_type == "process":
# Optional: Disable skirt/brim if we want a "clean" print estimation?
# Actually, for accurate cost, we SHOULD include skirt/brim if the profile has it.
pass
return profile
def list_machines(self) -> List[str]:
# Simple helper to list available machine JSONs
return self._list_profiles_by_type("machine")
def list_filaments(self) -> List[str]:
return self._list_profiles_by_type("filament")
def list_processes(self) -> List[str]:
return self._list_profiles_by_type("process")
def _list_profiles_by_type(self, ptype: str) -> List[str]:
results = []
for root, dirs, files in os.walk(self.profiles_root):
if os.path.basename(root) == ptype:
for f in files:
if f.endswith(".json") and "common" not in f:
results.append(f.replace(".json", ""))
return sorted(results)