Buckets:

download
raw
2.39 kB
#!/usr/bin/env python3
from dataclasses import dataclass, asdict
from typing import List, Dict, Optional
import json
from pathlib import Path
from pydbml import PyDBML # ✅ use PyDBML, not DBML
BASE_DIR = Path(__file__).resolve().parent
DBML_PATH = BASE_DIR / "schema.dbml"
OUT_PATH = BASE_DIR / "schema.json"
@dataclass
class ColumnSchema:
name: str
type: str
not_null: bool
pk: bool
unique: bool
default: Optional[str]
note: Optional[str]
@dataclass
class TableSchema:
name: str
note: Optional[str]
columns: List[ColumnSchema]
indexes: List[str]
relations: List[str]
def load_dbml(path: Path):
"""
Load the DBML using PyDBML.
PyDBML accepts:
- Path
- file-like object
- or source string
"""
print(f"[parse_dbml] Loading DBML → {path}")
# You can pass the Path directly:
return PyDBML(path)
def extract_schema(dbml: PyDBML) -> Dict[str, TableSchema]:
tables: Dict[str, TableSchema] = {}
for t in dbml.tables:
columns: List[ColumnSchema] = []
for c in t.columns:
columns.append(
ColumnSchema(
name=c.name,
type=str(c.type), # ensure it's a string
not_null=c.not_null, # ✅ correct attribute name
pk=c.pk, # primary key flag
unique=c.unique, # unique constraint
default=str(c.default) if c.default is not None else None,
note=(c.note.text if c.note else None),
)
)
indexes = [idx.name or str(idx) for idx in t.indexes]
tables[t.name] = TableSchema(
name=t.name,
note=t.note.text if t.note else None,
columns=columns,
indexes=indexes,
relations=[], # we can wire actual relations later if you want
)
return tables
def main():
if not DBML_PATH.exists():
raise SystemExit(f"[parse_dbml] ERROR: {DBML_PATH} does not exist")
dbml = load_dbml(DBML_PATH)
schema = extract_schema(dbml)
out = {tbl.name: asdict(tbl) for tbl in schema.values()}
OUT_PATH.write_text(json.dumps(out, indent=2))
print(f"[parse_dbml] ✔ wrote → {OUT_PATH}")
if __name__ == "__main__":
main()

Xet Storage Details

Size:
2.39 kB
·
Xet hash:
e76300eb09a88b094798a254b586d4bef9d540e0e67314062e6d1e8d75adca11

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.