docvortex 0.2.1__py3-none-any.whl

docvortex/analyzers/native/office/doc/styles.py

@@ -0,0 +1,215 @@
+"""解析 Word STSH/STD 样式表并解析继承链。"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+import re
+
+from loguru import logger
+
+from ..legacy.binary import bounded_slice, get_u16
+from .models import DocCharStyle
+from .records import DocBudget
+from .sprm import PapDelta, apply_character_sprms, apply_paragraph_sprms
+
+ISTD_NIL = 0x0FFF
+
+
+@dataclass(frozen=True, slots=True)
+class ResolvedStyle:
+    """一个样式继承完成后的可见属性。"""
+
+    name: str = ""
+    character: DocCharStyle = DocCharStyle()
+    paragraph: PapDelta = PapDelta()
+    heading_level: int | None = None
+    is_title: bool = False
+    toc_level: int | None = None
+    is_code: bool = False
+
+
+@dataclass(frozen=True, slots=True)
+class _RawStyle:
+    """STSH 内尚未解析继承的 STD。"""
+
+    sti: int
+    name: str
+    base: int
+    paragraph_grpprl: bytes
+    character_grpprl: bytes
+    paragraph_style: bool
+
+
+class Stylesheet:
+    """按 istd 提供解析后样式的只读集合。"""
+
+    def __init__(self, styles: dict[int, ResolvedStyle] | None = None) -> None:
+        """保存已解析样式并准备默认样式。"""
+
+        self._styles = styles or {}
+        self._default = ResolvedStyle()
+
+    def get(self, style_id: int) -> ResolvedStyle:
+        """返回指定样式，不存在时使用空默认值。"""
+
+        return self._styles.get(style_id, self._default)
+
+
+def _style_semantics(sti: int, name: str) -> tuple[int | None, bool, int | None, bool]:
+    """从内建 ID 与样式名识别标题、目录和代码语义。"""
+
+    normalized = re.sub(r"\s+", " ", name.strip()).casefold()
+    heading_level = sti if 1 <= sti <= 9 else None
+    if heading_level is None:
+        match = re.match(r"^(?:heading|标题)\s*([1-9])$", normalized)
+        if match:
+            heading_level = int(match.group(1))
+    is_title = normalized in {"title", "标题", "文档标题"} or sti == 15
+    toc_level: int | None = None
+    match = re.match(r"^(?:toc|目录)\s*([1-9])$", normalized)
+    if match:
+        toc_level = int(match.group(1)) - 1
+    is_code = any(token in normalized for token in ("code", "preformatted", "source code", "代码"))
+    return heading_level, is_title, toc_level, is_code
+
+
+def _parse_std(record: bytes, base_size: int) -> _RawStyle | None:
+    """解析一条 STD 的名称、继承和 UPX 内容。"""
+
+    if len(record) < max(base_size, 10):
+        return None
+    first = get_u16(record, 0)
+    second = get_u16(record, 2)
+    third = get_u16(record, 4)
+    if first is None or second is None or third is None:
+        return None
+    sti = first & 0x0FFF
+    kind = second & 0x000F
+    base = (second >> 4) & 0x0FFF
+    upx_count = third & 0x000F
+    name_offset = max(base_size, 10)
+    name_length = get_u16(record, name_offset)
+    if name_length is None:
+        return None
+    name_bytes = name_length * 2
+    raw_name = bounded_slice(record, name_offset + 2, name_bytes)
+    if raw_name is None:
+        return None
+    name = raw_name.decode("utf-16le", errors="replace")
+    cursor = name_offset + 2 + name_bytes + 2
+    upx: list[bytes] = []
+    for _ in range(upx_count):
+        if cursor % 2:
+            cursor += 1
+        length = get_u16(record, cursor)
+        if length is None:
+            break
+        payload = bounded_slice(record, cursor + 2, length)
+        if payload is None:
+            break
+        upx.append(payload)
+        cursor += 2 + length
+    paragraph_style = kind == 1
+    if paragraph_style:
+        paragraph_grpprl = upx[0][2:] if upx and len(upx[0]) >= 2 else b""
+        character_grpprl = upx[1] if len(upx) > 1 else b""
+    elif kind == 2:
+        paragraph_grpprl = b""
+        character_grpprl = upx[0] if upx else b""
+    else:
+        paragraph_grpprl = b""
+        character_grpprl = b""
+    return _RawStyle(
+        sti=sti,
+        name=name,
+        base=base,
+        paragraph_grpprl=paragraph_grpprl,
+        character_grpprl=character_grpprl,
+        paragraph_style=paragraph_style,
+    )
+
+
+def parse_stylesheet(
+    table_stream: bytes,
+    *,
+    offset: int,
+    size: int,
+    budget: DocBudget,
+) -> Stylesheet:
+    """解析 STSH，并在循环样式链处安全截断继承。"""
+
+    payload = bounded_slice(table_stream, offset, size)
+    if payload is None or len(payload) < 8:
+        return Stylesheet()
+    header_size = get_u16(payload, 0)
+    style_count = get_u16(payload, 2)
+    base_size = get_u16(payload, 4)
+    if header_size is None or style_count is None or base_size is None:
+        return Stylesheet()
+    cursor = 2 + header_size
+    raw_styles: dict[int, _RawStyle] = {}
+    for style_id in range(style_count):
+        length = get_u16(payload, cursor)
+        if length is None:
+            break
+        cursor += 2
+        if length == 0:
+            continue
+        record = bounded_slice(payload, cursor, length)
+        if record is None:
+            break
+        cursor += length
+        budget.charge()
+        parsed = _parse_std(record, base_size)
+        if parsed is not None:
+            raw_styles[style_id] = parsed
+
+    memo: dict[int, ResolvedStyle] = {}
+
+    def resolve(style_id: int) -> ResolvedStyle:
+        """迭代解析一个样式的 based-on 链。"""
+
+        if style_id in memo:
+            return memo[style_id]
+        chain: list[int] = []
+        visiting: set[int] = set()
+        cursor_id: int | None = style_id
+        base = ResolvedStyle()
+        while cursor_id is not None:
+            if cursor_id in memo:
+                base = memo[cursor_id]
+                break
+            if cursor_id in visiting:
+                logger.warning(f"DOC style inheritance cycle at istd={cursor_id}")
+                break
+            raw = raw_styles.get(cursor_id)
+            if raw is None:
+                break
+            visiting.add(cursor_id)
+            chain.append(cursor_id)
+            cursor_id = raw.base if raw.base not in {ISTD_NIL, cursor_id} else None
+        for current_id in reversed(chain):
+            raw = raw_styles[current_id]
+            paragraph = apply_paragraph_sprms(raw.paragraph_grpprl, b"", base.paragraph, budget=budget)
+            character = apply_character_sprms(
+                raw.character_grpprl,
+                base.character,
+                base.character,
+                budget=budget,
+            )
+            heading, is_title, toc_level, is_code = _style_semantics(raw.sti, raw.name)
+            base = ResolvedStyle(
+                name=raw.name or base.name,
+                character=character,
+                paragraph=paragraph,
+                heading_level=heading if heading is not None else base.heading_level,
+                is_title=is_title or base.is_title,
+                toc_level=toc_level if toc_level is not None else base.toc_level,
+                is_code=is_code or base.is_code,
+            )
+            memo[current_id] = base
+        return base
+
+    for style_id in raw_styles:
+        resolve(style_id)
+    return Stylesheet(memo)

docvortex/analyzers/native/office/docx/__init__.py

@@ -0,0 +1,3 @@
+"""Flash DOCX 转换实现。"""
+
+__all__: list[str] = []

docvortex/analyzers/native/office/docx/context.py

@@ -0,0 +1,71 @@
+"""DOCX 转换共享的 XML 常量和字段栈契约。"""
+
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Final, Optional, TypeAlias, Union
+from pydantic import AnyUrl
+from .formatting_types import Formatting
+from docvortex.foundation.type_identity import preserve_type_module
+
+_ParagraphHyperlink: TypeAlias = Optional[Union[AnyUrl, Path, str]]
+_ParagraphElement: TypeAlias = tuple[str, Optional[Formatting], _ParagraphHyperlink]
+
+
+@dataclass(slots=True)
+class _DocxComplexFieldFrame:
+    """保存一个 DOCX 复杂字段的指令、阶段与已解析结果元素。"""
+
+    instruction_parts: list[str] = field(default_factory=list)
+    phase: str = "instr"
+    result_elements: list[_ParagraphElement] = field(default_factory=list)
+
+
+class _DocxConstants:
+    """保存跨职责使用的固定命名空间与公式载体优先级。"""
+
+    _BLIP_NAMESPACES: Final = {
+        "a": "http://schemas.openxmlformats.org/drawingml/2006/main",
+        "r": "http://schemas.openxmlformats.org/officeDocument/2006/relationships",
+        "w": "http://schemas.openxmlformats.org/wordprocessingml/2006/main",
+        "wp": "http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing",
+        "mc": "http://schemas.openxmlformats.org/markup-compatibility/2006",
+        "o": "urn:schemas-microsoft-com:office:office",
+        "v": "urn:schemas-microsoft-com:vml",
+        "wps": "http://schemas.microsoft.com/office/word/2010/wordprocessingShape",
+        "w10": "urn:schemas-microsoft-com:office:word",
+        "a14": "http://schemas.microsoft.com/office/drawing/2010/main",
+        "c": "http://schemas.openxmlformats.org/drawingml/2006/chart",
+    }
+    _PARAGRAPH_TRANSPARENT_INLINE_CONTAINERS: Final = {
+        "bdo",
+        "customXml",
+        "dir",
+        "fldSimple",
+        "ins",
+        "moveTo",
+        "smartTag",
+    }
+    _FORMULA_TOKEN_KINDS: Final = frozenset({"omml", "equationxml", "mtef", "image_mtef"})
+    _FORMULA_SOURCE_PRIORITY: Final = (
+        "omml",
+        "equationxml",
+        "mtef",
+        "image_mtef",
+    )
+    """
+    Word 文档中使用的 XML 命名空间映射。
+
+    这些命名空间用于解析 DOCX 文件中的各种元素，包括：
+    - a: DrawingML 主命名空间
+    - r: Office 文档关系命名空间
+    - w: WordprocessingML 主命名空间
+    - wp: Wordprocessing Drawing 命名空间
+    - mc: 标记兼容性命名空间
+    - v: VML (Vector Markup Language) 命名空间
+    - wps: Wordprocessing Shape 命名空间
+    - w10: Office Word 命名空间
+    - a14: Office 2010 Drawing 命名空间
+    """
+
+
+preserve_type_module(_DocxComplexFieldFrame, "docvortex.analyzers.native.office.docx.docx_converter")