v0.5.3

- Мелкие исправления увеличивающие работу конвертера
2026-09-10 10:50:15 +03:00
parent 658036f920
commit f7136bd8bd
95 changed files with 7165 additions and 8771 deletions
@@ -3,7 +3,6 @@
 from __future__ import annotations

 import os
-import re
 import shutil
 from dataclasses import dataclass, field
 from pathlib import Path
@@ -11,12 +10,28 @@ from pathlib import Path
 from md2gost.extended_markdown import markdown
 from md2gost.diagram_renderer import DIAGRAM_LANGS
 from md2gost.profiles import preprocess_markdown
+from mdexport import (
+    BlockVisitor,
+    InlineRenderer,
+    is_contents_heading,
+    parse_image_caption,
+    reference_full_name,
+    render_inline,
+    walk_blocks,
+)

 from .escape import escape_text, escape_verbatim_for_listing, latex_label

+_REF_PREFIXES = {
+    "рисунок": "fig",
+    "таблица": "tab",
+    "листинг": "lst",
+    "формула": "eq",
+}
+

@dataclass
-class EmitState:
+class EmitState(BlockVisitor):
    lines: list[str] = field(default_factory=list)
    pending_caption: tuple[str | None, str | None, bool] | None = None  # id, text, +listing
    biblio: list[tuple[str, str]] = field(default_factory=list)
@@ -27,57 +42,75 @@ class EmitState:
    def w(self, s: str = "") -> None:
        self.lines.append(s)

+    def on_heading(self, element) -> None:
+        _emit_heading(self, element)
+
+    def on_table(self, element, caption_id: str | None, caption_text: str | None) -> None:
+        _emit_longtable(self, element, caption_id, caption_text)
+
+    def on_code(self, element, lang: str) -> None:
+        _emit_code(self, element, lang)
+
+    def on_list(self, element) -> None:
+        _emit_list(self, element)
+
+    def on_equation(self, element) -> None:
+        _emit_equation(self, element)
+
+    def on_image(self, image) -> None:
+        _emit_image(self, image)
+
+    def on_body_text(self, children) -> None:
+        text = _inline_text(children).strip()
+        if text:
+            self.w(text)
+            self.w()
+
+    def on_quote(self, element) -> None:
+        self.w(r"\begin{quote}")
+        for child in element.children:
+            if type(child).__name__ == "Paragraph":
+                self.w(_inline_text(child.children))
+        self.w(r"\end{quote}")
+        self.w()
+
+    def on_other(self, element) -> None:
+        self.w(_inline_text(element.children))
+        self.w()
+
+
+class _LatexInline(InlineRenderer):
+    def escape(self, text: str) -> str:
+        return escape_text(text)
+
+    def code_span(self, node) -> str:
+        return f"\\texttt{{{escape_text(node.children)}}}"
+
+    def emphasis(self, inner: str) -> str:
+        return f"\\textit{{{inner}}}"
+
+    def strong_emphasis(self, inner: str) -> str:
+        return f"\\textbf{{{inner}}}"
+
+    def link(self, node, inner: str) -> str:
+        return f"\\href{{{escape_text(node.dest)}}}{{{inner}}}"
+
+    def reference(self, node) -> str:
+        prefix = _REF_PREFIXES.get(node.type.lower(), "obj")
+        return f"{reference_full_name(node.type)}~\\ref{{{latex_label(node.name, prefix)}}}"
+
+    def inline_equation(self, node) -> str:
+        return f"${node.latex_equation}$"
+
+    def line_break(self) -> str:
+        return " "
+
+
+_INLINE = _LatexInline()
+

 def _inline_text(children) -> str:
-    parts: list[str] = []
-    for child in children or []:
-        t = type(child).__name__
-        if t in ("RawText", "Literal"):
-            parts.append(escape_text(getattr(child, "children", "") or ""))
-        elif t == "CodeSpan":
-            parts.append(f"\\texttt{{{escape_text(child.children)}}}")
-        elif t == "Emphasis":
-            parts.append(f"\\textit{{{_inline_text(child.children)}}}")
-        elif t == "StrongEmphasis":
-            parts.append(f"\\textbf{{{_inline_text(child.children)}}}")
-        elif t == "Link":
-            parts.append(
-                f"\\href{{{escape_text(child.dest)}}}{{{_inline_text(child.children)}}}"
-            )
-        elif t == "Reference":
-            # @Рисунок:id → \ref{fig:id} with word
-            cat = {
-                "рисунок": "рис.",
-                "таблица": "табл.",
-                "листинг": "лист.",
-                "формула": "форм.",
-            }.get(child.type.lower(), child.type)
-            # Prefer Russian full names as in ГОСТ
-            full = {
-                "рисунок": "Рисунок",
-                "таблица": "Таблица",
-                "листинг": "Листинг",
-                "формула": "формула",
-            }.get(child.type.lower(), child.type)
-            prefix = {
-                "рисунок": "fig",
-                "таблица": "tab",
-                "листинг": "lst",
-                "формула": "eq",
-            }.get(child.type.lower(), "obj")
-            parts.append(f"{full}~\\ref{{{latex_label(child.name, prefix)}}}")
-        elif t == "InlineEquation":
-            parts.append(f"${child.latex_equation}$")
-        elif t == "Image":
-            # handled at block level usually
-            parts.append("")
-        elif t == "LineBreak":
-            parts.append(" ")
-        elif hasattr(child, "children"):
-            parts.append(_inline_text(child.children))
-        else:
-            parts.append(escape_text(str(child)))
-    return "".join(parts)
+    return render_inline(children, _INLINE)


 def _cell_plain(cell) -> str:
@@ -140,7 +173,7 @@ def _emit_heading(st: EmitState, heading) -> None:
    level = heading.level
    numbered = getattr(heading, "numbered", True)

-    if upper == "СОДЕРЖАНИЕ" or upper.startswith("СОДЕРЖАНИЕ"):
+    if is_contents_heading(text):
        st.skip_toc_heading = True
        return  # TOC already in tocpage.tex

@@ -158,48 +191,9 @@ def _emit_heading(st: EmitState, heading) -> None:
    st.w()


-def _emit_paragraph(st: EmitState, para) -> None:
-    # detect bibliography lines [n]: ...
-    raw = _inline_text(para.children).strip()
-    # Also scan children for plain biblio from original — marko may wrap
-    text_probe = ""
-    for ch in para.children or []:
-        if type(ch).__name__ in ("RawText", "Literal"):
-            text_probe += ch.children or ""
-    m = re.match(r"^\[([^\]]+)\]:\s*(.+)$", text_probe.strip())
-    if m:
-        st.biblio.append((m.group(1), m.group(2).strip()))
-        return
-
-    # inline images in paragraph
-    has_image = any(type(c).__name__ == "Image" for c in (para.children or []))
-    if has_image:
-        for c in para.children:
-            if type(c).__name__ == "Image":
-                _emit_image(st, c)
-        # remaining text
-        others = [c for c in para.children if type(c).__name__ != "Image"]
-        t = _inline_text(others).strip()
-        if t:
-            st.w(t)
-            st.w()
-        return
-
-    if raw:
-        st.w(raw)
-        st.w()
-
-
 def _emit_image(st: EmitState, img) -> None:
    dest = img.dest
-    title = getattr(img, "title", None) or ""
-    # title may be "%id Caption"
-    caption_id = getattr(img, "unique_name", None)
-    caption_text = title
-    if title.startswith("%"):
-        parts = title[1:].split(None, 1)
-        caption_id = parts[0]
-        caption_text = parts[1] if len(parts) > 1 else ""
+    caption_id, caption_text = parse_image_caption(img, getattr(img, "unique_name", None))

    # copy image into Images/
    src = dest
@@ -230,10 +224,7 @@ def _emit_image(st: EmitState, img) -> None:

 def _emit_code(st: EmitState, block, lang: str) -> None:
    source = block.children[0].children if block.children else ""
-    cap_id, cap_text, with_listing = (None, None, False)
-    if st.pending_caption:
-        cap_id, cap_text, with_listing = st.pending_caption
-        st.pending_caption = None
+    cap_id, cap_text, _with_listing = st.take_caption()

    lang_l = (lang or "").strip().lower()
    if lang_l in DIAGRAM_LANGS:
@@ -277,10 +268,7 @@ def _emit_list(st: EmitState, lst) -> None:


 def _emit_equation(st: EmitState, eq) -> None:
-    cap_id = None
-    if st.pending_caption:
-        cap_id = st.pending_caption[0]
-        st.pending_caption = None
+    cap_id = st.take_caption()[0]
    st.w(r"\begin{equation}")
    if cap_id:
        st.w(rf"\label{{{latex_label(cap_id, 'eq')}}}")
@@ -294,54 +282,7 @@ def emit_document(md_text: str, *, work_dir: Path, images_dir: Path) -> str:
    pre = preprocess_markdown(md_text, emdash_to_hyphen=False)
    doc = markdown.parse(pre)

-    for el in doc.children:
-        name = type(el).__name__
-        if name == "BlankLine":
-            continue
-        if name == "Caption":
-            st.pending_caption = (
-                el.unique_name,
-                el.text,
-                getattr(el, "with_listing", False),
-            )
-            continue
-        if name == "Heading":
-            _emit_heading(st, el)
-            continue
-        if name == "Paragraph":
-            _emit_paragraph(st, el)
-            continue
-        if name == "Table":
-            cid = ct = None
-            if st.pending_caption:
-                cid, ct, _ = st.pending_caption
-                st.pending_caption = None
-            _emit_longtable(st, el, cid, ct)
-            continue
-        if name in ("FencedCode", "CodeBlock"):
-            lang = getattr(el, "lang", "") or ""
-            _emit_code(st, el, lang)
-            continue
-        if name == "List":
-            _emit_list(st, el)
-            continue
-        if name == "Equation":
-            _emit_equation(st, el)
-            continue
-        if name == "TOC":
-            continue  # tocpage.tex
-        if name == "Quote":
-            st.w(r"\begin{quote}")
-            for ch in el.children:
-                if type(ch).__name__ == "Paragraph":
-                    st.w(_inline_text(ch.children))
-            st.w(r"\end{quote}")
-            st.w()
-            continue
-        # fallback
-        if hasattr(el, "children"):
-            st.w(_inline_text(el.children))
-            st.w()
+    walk_blocks(doc.children, st)

    if st.biblio:
        st.w(r"\begin{thebibliography}{99}")
@@ -2,7 +2,7 @@

 from __future__ import annotations

-import re
+from mdexport.labels import sanitize_label

 _SPECIAL = {
    "\\": r"\textbackslash{}",
@@ -32,11 +32,5 @@ def escape_verbatim_for_listing(s: str) -> str:
    return s.replace("\\end{lstlisting}", "\\end\\{lstlisting}")


-_LABEL_RE = re.compile(r"[^\w\-]+", re.UNICODE)
-
-
 def latex_label(name: str | None, prefix: str = "obj") -> str:
-    if not name:
-        return prefix
-    cleaned = _LABEL_RE.sub("-", name.strip()).strip("-")
-    return f"{prefix}:{cleaned or 'x'}"
+    return sanitize_label(name, prefix, ":")