cdxml-toolkit 0.5.0__py3-none-any.whl

cdxml_toolkit/render/schema.py

@@ -0,0 +1,90 @@
+"""
+schema.py — Dataclasses for the scheme DSL descriptor.
+
+Represents the parsed content of a YAML scheme file. The LLM specifies
+chemistry content and topology; the renderer handles all spatial layout.
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from typing import Optional
+
+
+@dataclass
+class StructureRef:
+    """Reference to a chemical structure — resolved later by the renderer."""
+    id: str                          # user-assigned key (e.g. "ArBr")
+    smiles: Optional[str] = None     # SMILES string
+    name: Optional[str] = None       # compound name (for resolution)
+    file: Optional[str] = None       # path to CDXML file
+    cdxml_id: Optional[int] = None   # existing fragment ID
+    label: Optional[str] = None      # compound number displayed below (e.g. "1")
+
+
+@dataclass
+class ArrowContent:
+    """Content placed above or below an arrow."""
+    structures: list[str] = field(default_factory=list)  # refs to StructureRef ids
+    text: list[str] = field(default_factory=list)        # condition text lines
+
+
+@dataclass
+class StepDescriptor:
+    """A single reaction step."""
+    substrates: list[str]            # refs to StructureRef ids
+    products: list[str]              # refs to StructureRef ids
+    above_arrow: Optional[ArrowContent] = None
+    below_arrow: Optional[ArrowContent] = None
+    yield_: Optional[str] = None
+    number: Optional[int] = None     # for numbered steps
+    id: Optional[str] = None
+    arrow_style: str = "solid"       # "solid", "dashed", "failed" (X on arrow)
+
+
+@dataclass
+class RunArrowEntry:
+    """A single run (one scale) of a reaction step."""
+    input_label: str         # e.g. "2.15 g"
+    output_label: str        # e.g. "1.60 g, 72% yield"
+    note: Optional[str] = None  # per-run annotation, e.g. "HATU (1.2 eq)"
+
+
+@dataclass
+class StepRunArrows:
+    """Run arrows for a specific step (may have multiple scales)."""
+    step: int                # 1-indexed step number
+    runs: list[RunArrowEntry] = field(default_factory=list)
+
+
+VALID_LAYOUTS = frozenset({
+    "linear", "sequential", "divergent", "stacked-rows",
+    "numbered-parallel", "convergent",
+})
+
+VALID_WRAPS = frozenset({"repeat", "serpentine", "none"})
+
+VALID_ARROW_STYLES = frozenset({"solid", "dashed", "failed"})
+
+
+@dataclass
+class SectionDescriptor:
+    """A section in a stacked-rows layout."""
+    label: Optional[str] = None          # "(i)", "(a)", etc.
+    steps: list[StepDescriptor] = field(default_factory=list)
+    layout: str = "linear"               # each section's internal layout
+
+
+@dataclass
+class SchemeDescriptor:
+    """Complete scheme description."""
+    source: Optional[str] = None     # path to reaction_parser JSON file
+    structures: dict[str, StructureRef] = field(default_factory=dict)
+    steps: list[StepDescriptor] = field(default_factory=list)
+    layout: str = "linear"           # layout pattern keyword
+    wrap: str = "repeat"             # "repeat", "serpentine", "none"
+    steps_per_row: Optional[int] = None  # auto-computed if omitted
+    title: Optional[str] = None
+    run_arrows: list[StepRunArrows] = field(default_factory=list)
+    condition_key: Optional[dict[str, str]] = None  # letter conditions: {"a": "..."}
+    sections: list[SectionDescriptor] = field(default_factory=list)