Files
llm/plugins/1c/parser/dbnames.py
T

122 lines
4.3 KiB
Python

"""Parser for Params/DBNames files."""
from __future__ import annotations
from dataclasses import asdict, dataclass
from pathlib import Path
import re
from typing import Any
from .payload import parse_brace_text, payload_to_text, scalar
@dataclass(frozen=True)
class DBNamesRecord:
guid: str
storage_role: str
sql_number: int
index: int
source: str
def to_dict(self) -> dict[str, Any]:
return asdict(self)
def _unwrap_bom_sequence(value: Any) -> Any:
if (
isinstance(value, dict)
and value.get("type") == "sequence"
and len(value.get("items") or []) == 2
and isinstance(value["items"][0], dict)
and value["items"][0].get("type") == "atom"
and str(value["items"][0].get("value") or "").strip("\ufeff") == ""
):
return value["items"][1]
return value
def parse_dbnames_bytes(data: bytes, *, source: str = "DBNames") -> dict[str, Any]:
decoded = payload_to_text(data)
text = decoded.get("text")
if text is None:
raise ValueError(f"{source}: cannot decode DBNames text")
parsed = _unwrap_bom_sequence(parse_brace_text(text))
if not (isinstance(parsed, dict) and parsed.get("type") == "list"):
raise ValueError(f"{source}: expected root list")
items = parsed.get("items") or []
if len(items) != 2:
raise ValueError(f"{source}: expected 2 root items, got {len(items)}")
records_node = items[1]
if not (isinstance(records_node, dict) and records_node.get("type") == "list"):
raise ValueError(f"{source}: expected records list")
record_items = records_node.get("items") or []
records: list[DBNamesRecord] = []
for index, node in enumerate(record_items[1:], start=1):
if not (isinstance(node, dict) and node.get("type") == "list"):
continue
fields = node.get("items") or []
if len(fields) != 3:
continue
records.append(
DBNamesRecord(
guid=scalar(fields[0]).lower(),
storage_role=scalar(fields[1]),
sql_number=int(scalar(fields[2])),
index=index,
source=source,
)
)
return {
"source": source,
"compression": decoded["compression"],
"encoding": decoded["encoding"],
"root_number": int(scalar(items[0])),
"declared_count": int(scalar(record_items[0]) or "0") if record_items else 0,
"records": records,
}
def parse_dbnames_file(path: Path) -> dict[str, Any]:
return parse_dbnames_bytes(path.read_bytes(), source=path.name)
_GUID_RE = re.compile(
r"^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$"
)
def parse_dbnames_version_bytes(data: bytes, *, source: str = "DBNamesVersion") -> dict[str, Any]:
"""Parse the version marker stored separately from DBNames records."""
decoded = payload_to_text(data)
text = decoded.get("text")
if text is None:
raise ValueError(f"{source}: cannot decode DBNamesVersion text")
parsed = _unwrap_bom_sequence(parse_brace_text(text))
if not (isinstance(parsed, dict) and parsed.get("type") == "list"):
raise ValueError(f"{source}: expected root list")
items = parsed.get("items") or []
if len(items) != 2:
raise ValueError(f"{source}: expected 2 root items, got {len(items)}")
if not all(isinstance(item, dict) and item.get("type") == "atom" for item in items):
actual_types = [item.get("type") if isinstance(item, dict) else type(item).__name__ for item in items]
raise ValueError(f"{source}: expected scalar marker and version, got {actual_types}")
marker_text = scalar(items[0]).strip()
try:
marker = int(marker_text)
except (TypeError, ValueError) as exc:
raise ValueError(f"{source}: marker must be an integer, got {marker_text!r}") from exc
version = scalar(items[1]).strip().lower()
if not _GUID_RE.fullmatch(version):
raise ValueError(f"{source}: version must be a GUID string, got {version!r}")
return {
"source": source,
"compression": decoded["compression"],
"encoding": decoded["encoding"],
"marker": marker,
"version": version,
}
def parse_dbnames_version_file(path: Path) -> dict[str, Any]:
return parse_dbnames_version_bytes(path.read_bytes(), source=path.name)