tocsmith 0.1.0__tar.gz → 0.2.0__tar.gz

{tocsmith-0.1.0 → tocsmith-0.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tocsmith
-Version: 0.1.0
+Version: 0.2.0
 Summary: Create PDF bookmarks automatically using heuristics, with CLI and async tkinter GUI
 Author-email: Wesley Yang <yxnian@outlook.com>
 Project-URL: Homepage, https://github.com/wesleyel/pdf-bookmark
@@ -25,8 +25,12 @@ Requires-Dist: mypy>=1.10.0; extra == "dev"
 
 ## 功能概览
 - 手动粘贴目录文本（每行以书中页码结尾），自动解析标题、页码与层级（1..6）
+- 支持两种层级解析方式：
+  - **按序号**：`1 标题 1` / `1.1 子标题 2` / `1.1.1 子子标题 3`，可通过 `keep_numbering` 控制序号是否写入书签标题
+  - **按缩进**：通过行首空格/Tab 缩进表示层级，标题不含序号
+  - 默认 `auto` 自动识别；也可通过 `--toc-mode` 或配置 `toc_mode` 显式指定
 - 支持页码偏移（实际页码 - 书籍页码），用于扫描件/前置页差异
-- 编号前缀会被保留到标题中：如 `第1章`、`1.1` 将出现在最终书签标题里
+- 按序号模式下，默认保留编号前缀到标题中（如 `第1章`、`1.1`）；设置 `keep_numbering = false` 或 `--no-keep-numbering` 可仅用于推断层级
 - 支持行首星号标记：允许输入 `*1.1 Title` 或 `* 1.1 Title`，输出统一为 `*1.1 Title`
 - 将条目以父子层级写入 PDF 书签
 - 提供 CLI 与 GUI；亦可通过 Python API 使用
@@ -82,6 +86,10 @@ tocsmith --help
 page_offset = 0
 # global minimum length
 min_len = 3
+# TOC hierarchy mode: auto | numbering | indent
+toc_mode = "auto"
+# keep numbering prefix in bookmark titles (numbering mode only)
+keep_numbering = true
 
 # input folder
 input_prefix = "input"
@@ -100,6 +108,26 @@ toc = """
 """
 page_offset = 10
 min_len = 2
+toc_mode = "numbering"
+keep_numbering = false
+```
+
+### 目录文本格式
+
+**按序号**（`toc_mode = "numbering"` 或自动识别）：
+
+```
+1  我是标题  1
+1.1  我是子标题  2
+1.1.1  我是子子标题  3
+```
+
+**按缩进**（`toc_mode = "indent"` 或自动识别）：
+
+```
+我是标题  1
+    我是子标题  2
+        我是子子标题  3
 ```
 
 运行：
@@ -109,7 +137,7 @@ tocsmith --config config.toml
 ```
 
 说明：
-- `defaults` 中的 `page_offset`、`min_len` 可被每个任务覆盖。
+- `defaults` 中的 `page_offset`、`min_len`、`toc_mode`、`keep_numbering` 可被每个任务覆盖。
 - `input_prefix` 用于解析任务中的 `input_file`；`output_prefix` 为输出目录根。
 - 输出文件名为 `{stem}{output_suffix}`，其中 `stem` 来源于 `input_file`。
 - 任务可直接内联 `toc` 文本；也兼容 `toc_file` 指定外部文件。
@@ -125,6 +153,8 @@ uv run python -m tocsmith.gui
 - 选择输入 PDF
 - 可选：修改输出路径
 - 在 “TOC text” 中粘贴目录文本；在 “Page Offset” 填写偏移（实际 - 书籍）
+- 选择 “TOC Mode”：`auto`（自动识别）、`numbering`（按序号）、`indent`（按缩进）
+- 勾选 “Keep numbering” 控制按序号模式下是否保留标题中的序号（默认保留）
 - 点击 “Parse TOC Text” 查看解析结果
 - 点击 “Generate” 生成带书签的 PDF
 

{tocsmith-0.1.0 → tocsmith-0.2.0}/README.md

@@ -8,8 +8,12 @@
 
 ## 功能概览
 - 手动粘贴目录文本（每行以书中页码结尾），自动解析标题、页码与层级（1..6）
+- 支持两种层级解析方式：
+  - **按序号**：`1 标题 1` / `1.1 子标题 2` / `1.1.1 子子标题 3`，可通过 `keep_numbering` 控制序号是否写入书签标题
+  - **按缩进**：通过行首空格/Tab 缩进表示层级，标题不含序号
+  - 默认 `auto` 自动识别；也可通过 `--toc-mode` 或配置 `toc_mode` 显式指定
 - 支持页码偏移（实际页码 - 书籍页码），用于扫描件/前置页差异
-- 编号前缀会被保留到标题中：如 `第1章`、`1.1` 将出现在最终书签标题里
+- 按序号模式下，默认保留编号前缀到标题中（如 `第1章`、`1.1`）；设置 `keep_numbering = false` 或 `--no-keep-numbering` 可仅用于推断层级
 - 支持行首星号标记：允许输入 `*1.1 Title` 或 `* 1.1 Title`，输出统一为 `*1.1 Title`
 - 将条目以父子层级写入 PDF 书签
 - 提供 CLI 与 GUI；亦可通过 Python API 使用
@@ -65,6 +69,10 @@ tocsmith --help
 page_offset = 0
 # global minimum length
 min_len = 3
+# TOC hierarchy mode: auto | numbering | indent
+toc_mode = "auto"
+# keep numbering prefix in bookmark titles (numbering mode only)
+keep_numbering = true
 
 # input folder
 input_prefix = "input"
@@ -83,6 +91,26 @@ toc = """
 """
 page_offset = 10
 min_len = 2
+toc_mode = "numbering"
+keep_numbering = false
+```
+
+### 目录文本格式
+
+**按序号**（`toc_mode = "numbering"` 或自动识别）：
+
+```
+1  我是标题  1
+1.1  我是子标题  2
+1.1.1  我是子子标题  3
+```
+
+**按缩进**（`toc_mode = "indent"` 或自动识别）：
+
+```
+我是标题  1
+    我是子标题  2
+        我是子子标题  3
 ```
 
 运行：
@@ -92,7 +120,7 @@ tocsmith --config config.toml
 ```
 
 说明：
-- `defaults` 中的 `page_offset`、`min_len` 可被每个任务覆盖。
+- `defaults` 中的 `page_offset`、`min_len`、`toc_mode`、`keep_numbering` 可被每个任务覆盖。
 - `input_prefix` 用于解析任务中的 `input_file`；`output_prefix` 为输出目录根。
 - 输出文件名为 `{stem}{output_suffix}`，其中 `stem` 来源于 `input_file`。
 - 任务可直接内联 `toc` 文本；也兼容 `toc_file` 指定外部文件。
@@ -108,6 +136,8 @@ uv run python -m tocsmith.gui
 - 选择输入 PDF
 - 可选：修改输出路径
 - 在 “TOC text” 中粘贴目录文本；在 “Page Offset” 填写偏移（实际 - 书籍）
+- 选择 “TOC Mode”：`auto`（自动识别）、`numbering`（按序号）、`indent`（按缩进）
+- 勾选 “Keep numbering” 控制按序号模式下是否保留标题中的序号（默认保留）
 - 点击 “Parse TOC Text” 查看解析结果
 - 点击 “Generate” 生成带书签的 PDF
 

{tocsmith-0.1.0 → tocsmith-0.2.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "tocsmith"
-version = "0.1.0"
+version = "0.2.0"
 description = "Create PDF bookmarks automatically using heuristics, with CLI and async tkinter GUI"
 authors = [{ name = "Wesley Yang", email = "yxnian@outlook.com" }]
 urls = { Homepage = "https://github.com/wesleyel/pdf-bookmark", Source = "https://github.com/wesleyel/pdf-bookmark" }

{tocsmith-0.1.0 → tocsmith-0.2.0}/tocsmith/cli.py

@@ -5,7 +5,7 @@ from pathlib import Path
 from typing import Any, Dict, List, Optional
 import sys
 
-from .core import generate_bookmarks, parse_toc_lines
+from .core import TocMode, generate_bookmarks, parse_toc_lines
 
 try:  # Python 3.11+
     import tomllib  # type: ignore[attr-defined]
@@ -23,6 +23,18 @@ def parse_args(argv: List[str] | None = None) -> argparse.Namespace:
     p.add_argument("--min-len", type=int, default=3, help="Minimum heading text length")
     p.add_argument("--page-offset", type=int, default=0, help="Page offset: actual - book page")
     p.add_argument("--toc-file", help="Path to a text file containing TOC lines")
+    p.add_argument(
+        "--toc-mode",
+        choices=["auto", "numbering", "indent"],
+        default="auto",
+        help="TOC hierarchy mode: numbering (1/1.1), indent (spaces), or auto-detect",
+    )
+    p.add_argument(
+        "--keep-numbering",
+        action=argparse.BooleanOptionalAction,
+        default=True,
+        help="Keep numbering prefix in bookmark titles (numbering mode only)",
+    )
     p.add_argument(
         "-c",
         "--config",
@@ -46,6 +58,8 @@ def _run_single(
     page_offset: int,
     min_len: int,
     toc_text: Optional[str] = None,
+    toc_mode: TocMode = "auto",
+    keep_numbering: bool = True,
 ) -> int:
     """Run a single task and return process exit code."""
     if not src.exists():
@@ -55,10 +69,22 @@ def _run_single(
 
     headings = []
     if toc_text is not None and toc_text.strip():
-        headings = parse_toc_lines(toc_text, page_offset=page_offset, min_len=min_len)
+        headings = parse_toc_lines(
+            toc_text,
+            page_offset=page_offset,
+            min_len=min_len,
+            mode=toc_mode,
+            keep_numbering=keep_numbering,
+        )
     elif toc_file:
         file_text = Path(toc_file).read_text(encoding="utf-8")
-        headings = parse_toc_lines(file_text, page_offset=page_offset, min_len=min_len)
+        headings = parse_toc_lines(
+            file_text,
+            page_offset=page_offset,
+            min_len=min_len,
+            mode=toc_mode,
+            keep_numbering=keep_numbering,
+        )
     else:
         print("No TOC source provided (use --toc-file). Producing a copy without outline.")
         headings = []
@@ -86,6 +112,8 @@ def _run_batch(config_path: Path) -> int:
     # Alternatively: toc_file = "toc.txt"
     page_offset = 10                     # optional overrides default
     min_len = 2                          # optional overrides default
+    toc_mode = "auto"                    # optional: auto | numbering | indent
+    keep_numbering = true                # optional: keep numbering in bookmark titles
     '''
     if tomllib is None:
         print("Error: TOML support not available. Please install 'tomli' for Python < 3.11.")
@@ -107,6 +135,11 @@ def _run_batch(config_path: Path) -> int:
 
     default_page_offset = int(defaults.get("page_offset", 0) or 0)
     default_min_len = int(defaults.get("min_len", 3) or 3)
+    default_toc_mode = str(defaults.get("toc_mode", "auto") or "auto").strip() or "auto"
+    if default_toc_mode not in ("auto", "numbering", "indent"):
+        print(f"Invalid defaults.toc_mode: {default_toc_mode!r}")
+        return 2
+    default_keep_numbering = bool(defaults.get("keep_numbering", True))
     input_prefix = str(defaults.get("input_prefix", "")).strip() or ""
     output_prefix = str(defaults.get("output_prefix", "")).strip() or ""
     output_suffix = (
@@ -139,11 +172,18 @@ def _run_batch(config_path: Path) -> int:
         toc_file = _resolve_relative(base_dir, t.get("toc_file"))
         page_offset = int(t.get("page_offset", default_page_offset) or default_page_offset)
         min_len = int(t.get("min_len", default_min_len) or default_min_len)
+        toc_mode = str(t.get("toc_mode", default_toc_mode) or default_toc_mode).strip() or "auto"
+        if toc_mode not in ("auto", "numbering", "indent"):
+            print(f"[Task {idx}] Skipped: invalid toc_mode {toc_mode!r}")
+            failures += 1
+            continue
+        keep_numbering = bool(t.get("keep_numbering", default_keep_numbering))
 
         print(
             f"[Task {idx}] Running: src={src} out={out} "
             f"toc={'inline' if (toc_inline and toc_inline.strip()) else (toc_file or '<none>')} "
-            f"offset={page_offset} min_len={min_len}"
+            f"offset={page_offset} min_len={min_len} toc_mode={toc_mode} "
+            f"keep_numbering={keep_numbering}"
         )
         try:
             # Ensure output directory exists
@@ -155,6 +195,8 @@ def _run_batch(config_path: Path) -> int:
                 page_offset=page_offset,
                 min_len=min_len,
                 toc_text=toc_inline,
+                toc_mode=toc_mode,  # type: ignore[arg-type]
+                keep_numbering=keep_numbering,
             )
             if code != 0:
                 failures += 1
@@ -186,6 +228,8 @@ def main(argv: List[str] | None = None) -> int:
         toc_file=Path(ns.toc_file) if ns.toc_file else None,
         page_offset=ns.page_offset,
         min_len=ns.min_len,
+        toc_mode=ns.toc_mode,
+        keep_numbering=ns.keep_numbering,
     )
 
 

tocsmith-0.2.0/tocsmith/core.py

@@ -0,0 +1,223 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+import re
+from functools import reduce
+from math import gcd
+from typing import Iterable, List, Literal, Tuple, Optional
+
+TocMode = Literal["numbering", "indent", "auto"]
+
+from pypdf import PdfReader, PdfWriter
+
+
+@dataclass
+class Heading:
+    title: str
+    page: int  # 1-based
+    level: int  # 1..6
+
+
+def generate_bookmarks(src_pdf: str, out_pdf: str, headings: Iterable[Heading]) -> None:
+    """Write given headings into a new PDF file as outline/bookmarks."""
+    reader = PdfReader(src_pdf)
+    writer = PdfWriter()
+    for page in reader.pages:
+        writer.add_page(page)
+
+    # Build hierarchical outlines using a simple stack by levels
+    stack: List[Tuple[int, object]] = []  # (level, parent_ref)
+
+    for h in headings:
+        page_index = max(0, min(len(reader.pages) - 1, h.page - 1))
+        while stack and stack[-1][0] >= h.level:
+            stack.pop()
+        parent = stack[-1][1] if stack else None
+        dest = writer.add_outline_item(h.title, page_index, parent=parent)
+        stack.append((h.level, dest))
+
+    with open(out_pdf, "wb") as f:
+        writer.write(f)
+
+
+# -------------------- TOC parsing utilities --------------------
+
+_NUM_PREFIX_RE = re.compile(
+    r"^\s*(?P<num>(第\s*\d+[一二三四五六七八九十百千]*[章节节部分编]?)|((\d+\.)+\d+)|\d+)?\s*"
+)
+_TRAILING_PAGE_RE = re.compile(r"(?P<page>\d{1,5})\s*$")
+
+
+def _infer_level_from_numbering(num: Optional[str]) -> int:
+    if not num:
+        return 1
+    num = num.strip()
+    if num.startswith("第"):
+        # "第1章" style => top-level
+        return 1
+    if "." in num:
+        # "1.2.3" => level = segments + 1 (so 1.2 is level 2)
+        return min(6, max(1, num.count(".") + 1))
+    # Simple leading integer like "1" => level 1
+    return 1
+
+
+def _leading_indent_width(raw_line: str) -> int:
+    width = 0
+    for ch in raw_line:
+        if ch == " ":
+            width += 1
+        elif ch == "\t":
+            width += 4
+        else:
+            break
+    return width
+
+
+def _detect_indent_unit(indents: Iterable[int]) -> int:
+    non_zero = sorted({i for i in indents if i > 0})
+    if not non_zero:
+        return 4
+    unit = non_zero[0]
+    if all(i % unit == 0 for i in indents):
+        return max(1, unit)
+    return max(1, reduce(gcd, non_zero))
+
+
+def _infer_level_from_indent(indent: int, unit: int) -> int:
+    if indent <= 0:
+        return 1
+    return min(6, max(1, indent // unit + 1))
+
+
+def _strip_star_prefix(line: str) -> Tuple[str, str]:
+    star_prefix = ""
+    m_star = re.match(r"^\*+\s*", line)
+    if m_star:
+        star_prefix = "*" * m_star.group(0).count("*")
+        line = line[m_star.end() :].lstrip()
+    return star_prefix, line
+
+
+def _detect_toc_mode(toc_text: str, min_len: int = 1) -> TocMode:
+    """Auto-detect whether TOC hierarchy is expressed by numbering or indentation."""
+    indent_signals = 0
+    numbering_signals = 0
+    for raw_line in toc_text.splitlines():
+        if len(raw_line.strip()) < min_len:
+            continue
+        line = raw_line.lstrip()
+        _, line = _strip_star_prefix(line)
+        page_m = _TRAILING_PAGE_RE.search(line)
+        if not page_m:
+            continue
+        line_wo_page = line[: page_m.start()].rstrip()
+        indent = _leading_indent_width(raw_line)
+        num_m = _NUM_PREFIX_RE.match(line_wo_page)
+        has_numbering = bool(num_m and num_m.group("num"))
+        if has_numbering:
+            numbering_signals += 1
+        elif indent > 0:
+            indent_signals += 1
+    return "indent" if indent_signals > numbering_signals else "numbering"
+
+
+def _parse_toc_lines_numbering(
+    toc_text: str,
+    page_offset: int = 0,
+    min_len: int = 1,
+    keep_numbering: bool = True,
+) -> List[Heading]:
+    headings: List[Heading] = []
+    for raw_line in toc_text.splitlines():
+        line = raw_line.strip()
+        if len(line) < min_len:
+            continue
+        star_prefix, line = _strip_star_prefix(line)
+
+        page_m = _TRAILING_PAGE_RE.search(line)
+        if not page_m:
+            continue
+        page_num = int(page_m.group("page"))
+        line_wo_page = line[: page_m.start()].rstrip()
+        num_m = _NUM_PREFIX_RE.match(line_wo_page)
+        numbering = None
+        title_part = line_wo_page
+        if num_m:
+            numbering = num_m.group("num")
+            title_part = line_wo_page[num_m.end() :].strip()
+        if numbering and keep_numbering:
+            combined = f"{numbering.strip()} {title_part}".strip()
+        else:
+            combined = title_part
+        title = re.sub(r"\s+", " ", combined)
+        if not title:
+            title = line_wo_page.strip() if keep_numbering else title_part.strip()
+        if star_prefix:
+            title = f"{star_prefix}{title}".strip()
+        level = _infer_level_from_numbering(numbering)
+        pdf_page = max(1, page_num + page_offset)
+        headings.append(Heading(title=title, page=pdf_page, level=level))
+
+    headings.sort(key=lambda h: (h.page, h.level, h.title.lower()))
+    return headings
+
+
+def _parse_toc_lines_indent(toc_text: str, page_offset: int = 0, min_len: int = 1) -> List[Heading]:
+    lines_data: List[Tuple[int, str, int]] = []
+    indents: List[int] = []
+    for raw_line in toc_text.splitlines():
+        if len(raw_line.strip()) < min_len:
+            continue
+        indent = _leading_indent_width(raw_line)
+        line = raw_line.lstrip()
+        star_prefix, line = _strip_star_prefix(line)
+
+        page_m = _TRAILING_PAGE_RE.search(line)
+        if not page_m:
+            continue
+        page_num = int(page_m.group("page"))
+        title = re.sub(r"\s+", " ", line[: page_m.start()].rstrip())
+        if star_prefix:
+            title = f"{star_prefix}{title}".strip()
+        indents.append(indent)
+        lines_data.append((indent, title, page_num))
+
+    unit = _detect_indent_unit(indents)
+    headings: List[Heading] = []
+    for indent, title, page_num in lines_data:
+        level = _infer_level_from_indent(indent, unit)
+        pdf_page = max(1, page_num + page_offset)
+        headings.append(Heading(title=title, page=pdf_page, level=level))
+
+    headings.sort(key=lambda h: (h.page, h.level, h.title.lower()))
+    return headings
+
+
+def parse_toc_lines(
+    toc_text: str,
+    page_offset: int = 0,
+    min_len: int = 1,
+    mode: TocMode = "auto",
+    keep_numbering: bool = True,
+) -> List[Heading]:
+    """
+    Parse a pasted TOC text into Heading entries.
+    - Each line should end with the book page number (digits)
+    - mode="numbering": hierarchy from leading numbers like "1", "1.1", "第1章"
+    - mode="indent": hierarchy from leading spaces/tabs
+    - mode="auto": detect numbering vs indent automatically
+    - keep_numbering: when True (default), numbering prefix is kept in bookmark titles
+    - page_offset is added to the parsed page number to map to PDF actual pages
+    """
+    resolved_mode = _detect_toc_mode(toc_text, min_len) if mode == "auto" else mode
+    if resolved_mode == "indent":
+        return _parse_toc_lines_indent(toc_text, page_offset=page_offset, min_len=min_len)
+    return _parse_toc_lines_numbering(
+        toc_text, page_offset=page_offset, min_len=min_len, keep_numbering=keep_numbering
+    )
+
+
+## URL/website TOC fetching intentionally removed; only manual text input is supported.
+
+

{tocsmith-0.1.0 → tocsmith-0.2.0}/tocsmith/gui.py

@@ -68,6 +68,24 @@ class App:
         self.offset_entry = ttk.Entry(ctrl, textvariable=self.offset_var, width=6)
         self.offset_entry.pack(side=tk.LEFT, padx=(4, 12))
 
+        ttk.Label(ctrl, text="TOC Mode:").pack(side=tk.LEFT)
+        self.toc_mode_var = tk.StringVar(value="auto")
+        self.toc_mode_combo = ttk.Combobox(
+            ctrl,
+            textvariable=self.toc_mode_var,
+            values=["auto", "numbering", "indent"],
+            state="readonly",
+            width=10,
+        )
+        self.toc_mode_combo.pack(side=tk.LEFT, padx=(4, 12))
+
+        self.keep_numbering_var = tk.BooleanVar(value=True)
+        ttk.Checkbutton(
+            ctrl,
+            text="Keep numbering",
+            variable=self.keep_numbering_var,
+        ).pack(side=tk.LEFT)
+
         # TOC input
         toc_row = ttk.Frame(frm)
         toc_row.pack(fill=tk.BOTH, expand=True)
@@ -162,6 +180,20 @@ class App:
 
     # Auto analysis removed
 
+    def _get_parse_kwargs(self) -> dict:
+        try:
+            offset = int(self.offset_var.get() or 0)
+        except ValueError:
+            offset = 0
+        mode = self.toc_mode_var.get() or "auto"
+        if mode not in ("auto", "numbering", "indent"):
+            mode = "auto"
+        return {
+            "page_offset": offset,
+            "mode": mode,
+            "keep_numbering": self.keep_numbering_var.get(),
+        }
+
     def _on_generate(self) -> None:
         if not self.in_var.get():
             messagebox.showwarning("Missing", "Please choose an input PDF")
@@ -176,11 +208,7 @@ class App:
             text = self.toc_text.get("1.0", tk.END).strip()
             hs = []
             if text:
-                try:
-                    offset = int(self.offset_var.get() or 0)
-                except ValueError:
-                    offset = 0
-                hs = await run_in_thread(parse_toc_lines, text, offset)
+                hs = await run_in_thread(parse_toc_lines, text, **self._get_parse_kwargs())
             else:
                 hs = []
             await run_in_thread(generate_bookmarks, self.in_var.get(), self.out_var.get(), hs)
@@ -194,14 +222,9 @@ class App:
         if not text:
             messagebox.showwarning("Empty", "Please paste TOC text or URL first")
             return
-        try:
-            offset = int(self.offset_var.get() or 0)
-        except ValueError:
-            offset = 0
-
         async def task():
             self._set_status("Parsing TOC…")
-            hs = await run_in_thread(parse_toc_lines, text, offset)
+            hs = await run_in_thread(parse_toc_lines, text, **self._get_parse_kwargs())
             self._populate_tree(hs)
             self._set_status(f"Parsed {len(hs)} entries")
 

{tocsmith-0.1.0 → tocsmith-0.2.0}/tocsmith/tests/test_core.py

@@ -163,3 +163,82 @@ def test_parse_toc_lines_preserve_numbering_with_asterisk():
     assert any(t.startswith("*2 ") and "星标章节" in t for t in titles)
 
 
+def test_parse_toc_lines_numbering_mode_example():
+    toc = "\n".join([
+        "1  我是标题  1",
+        "1.1  我是子标题  2",
+        "1.1.1  我是子子标题  3",
+    ])
+    hs = parse_toc_lines(toc, page_offset=0, mode="numbering")
+    assert len(hs) == 3
+    assert hs[0].title == "1 我是标题"
+    assert hs[0].level == 1
+    assert hs[1].title == "1.1 我是子标题"
+    assert hs[1].level == 2
+    assert hs[2].title == "1.1.1 我是子子标题"
+    assert hs[2].level == 3
+
+
+def test_parse_toc_lines_indent_mode_example():
+    toc = "\n".join([
+        "我是标题  1",
+        "    我是子标题  2",
+        "        我是子子标题  3",
+    ])
+    hs = parse_toc_lines(toc, page_offset=0, mode="indent")
+    assert len(hs) == 3
+    assert hs[0].title == "我是标题"
+    assert hs[0].level == 1
+    assert hs[1].title == "我是子标题"
+    assert hs[1].level == 2
+    assert hs[2].title == "我是子子标题"
+    assert hs[2].level == 3
+
+
+def test_parse_toc_lines_auto_detect_indent():
+    toc = "\n".join([
+        "Chapter A 1",
+        "    Section B 2",
+        "        Subsection C 3",
+    ])
+    hs = parse_toc_lines(toc, page_offset=0, mode="auto")
+    assert [h.level for h in hs] == [1, 2, 3]
+
+
+def test_parse_toc_lines_strip_numbering_when_disabled():
+    toc = "\n".join([
+        "第1章 计算机系统概述 1",
+        "1.1 操作系统的基本概念 2",
+        "2 其他章节 10",
+    ])
+    hs = parse_toc_lines(toc, page_offset=0, keep_numbering=False)
+    titles = [h.title for h in hs]
+    assert any(t == "计算机系统概述" for t in titles)
+    assert any(t == "操作系统的基本概念" for t in titles)
+    assert any(t == "其他章节" for t in titles)
+    assert [h.level for h in hs] == [1, 2, 1]
+
+
+def test_parse_toc_lines_strip_numbering_with_asterisk():
+    toc = "\n".join([
+        "*1.1 星标小节 12",
+        "* 2 星标章节 13",
+    ])
+    hs = parse_toc_lines(toc, page_offset=0, keep_numbering=False)
+    titles = [h.title for h in hs]
+    assert any(t == "*星标小节" for t in titles)
+    assert any(t == "*星标章节" for t in titles)
+
+
+def test_parse_toc_lines_auto_detect_numbering():
+    toc = "\n".join([
+        "第1章 基础 1",
+        "1.1 小节 2",
+    ])
+    hs = parse_toc_lines(toc, page_offset=0, mode="auto")
+    assert hs[0].level == 1
+    assert hs[1].level == 2
+    assert hs[0].title.startswith("第1章")
+    assert hs[1].title.startswith("1.1")
+
+

{tocsmith-0.1.0 → tocsmith-0.2.0}/tocsmith.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tocsmith
-Version: 0.1.0
+Version: 0.2.0
 Summary: Create PDF bookmarks automatically using heuristics, with CLI and async tkinter GUI
 Author-email: Wesley Yang <yxnian@outlook.com>
 Project-URL: Homepage, https://github.com/wesleyel/pdf-bookmark
@@ -25,8 +25,12 @@ Requires-Dist: mypy>=1.10.0; extra == "dev"
 
 ## 功能概览
 - 手动粘贴目录文本（每行以书中页码结尾），自动解析标题、页码与层级（1..6）
+- 支持两种层级解析方式：
+  - **按序号**：`1 标题 1` / `1.1 子标题 2` / `1.1.1 子子标题 3`，可通过 `keep_numbering` 控制序号是否写入书签标题
+  - **按缩进**：通过行首空格/Tab 缩进表示层级，标题不含序号
+  - 默认 `auto` 自动识别；也可通过 `--toc-mode` 或配置 `toc_mode` 显式指定
 - 支持页码偏移（实际页码 - 书籍页码），用于扫描件/前置页差异
-- 编号前缀会被保留到标题中：如 `第1章`、`1.1` 将出现在最终书签标题里
+- 按序号模式下，默认保留编号前缀到标题中（如 `第1章`、`1.1`）；设置 `keep_numbering = false` 或 `--no-keep-numbering` 可仅用于推断层级
 - 支持行首星号标记：允许输入 `*1.1 Title` 或 `* 1.1 Title`，输出统一为 `*1.1 Title`
 - 将条目以父子层级写入 PDF 书签
 - 提供 CLI 与 GUI；亦可通过 Python API 使用
@@ -82,6 +86,10 @@ tocsmith --help
 page_offset = 0
 # global minimum length
 min_len = 3
+# TOC hierarchy mode: auto | numbering | indent
+toc_mode = "auto"
+# keep numbering prefix in bookmark titles (numbering mode only)
+keep_numbering = true
 
 # input folder
 input_prefix = "input"
@@ -100,6 +108,26 @@ toc = """
 """
 page_offset = 10
 min_len = 2
+toc_mode = "numbering"
+keep_numbering = false
+```
+
+### 目录文本格式
+
+**按序号**（`toc_mode = "numbering"` 或自动识别）：
+
+```
+1  我是标题  1
+1.1  我是子标题  2
+1.1.1  我是子子标题  3
+```
+
+**按缩进**（`toc_mode = "indent"` 或自动识别）：
+
+```
+我是标题  1
+    我是子标题  2
+        我是子子标题  3
 ```
 
 运行：
@@ -109,7 +137,7 @@ tocsmith --config config.toml
 ```
 
 说明：
-- `defaults` 中的 `page_offset`、`min_len` 可被每个任务覆盖。
+- `defaults` 中的 `page_offset`、`min_len`、`toc_mode`、`keep_numbering` 可被每个任务覆盖。
 - `input_prefix` 用于解析任务中的 `input_file`；`output_prefix` 为输出目录根。
 - 输出文件名为 `{stem}{output_suffix}`，其中 `stem` 来源于 `input_file`。
 - 任务可直接内联 `toc` 文本；也兼容 `toc_file` 指定外部文件。
@@ -125,6 +153,8 @@ uv run python -m tocsmith.gui
 - 选择输入 PDF
 - 可选：修改输出路径
 - 在 “TOC text” 中粘贴目录文本；在 “Page Offset” 填写偏移（实际 - 书籍）
+- 选择 “TOC Mode”：`auto`（自动识别）、`numbering`（按序号）、`indent`（按缩进）
+- 勾选 “Keep numbering” 控制按序号模式下是否保留标题中的序号（默认保留）
 - 点击 “Parse TOC Text” 查看解析结果
 - 点击 “Generate” 生成带书签的 PDF
 

tocsmith-0.1.0/tocsmith/core.py

@@ -1,122 +0,0 @@
-from __future__ import annotations
-
-from dataclasses import dataclass
-import re
-from typing import Iterable, List, Tuple, Optional
-
-from pypdf import PdfReader, PdfWriter
-
-
-@dataclass
-class Heading:
-    title: str
-    page: int  # 1-based
-    level: int  # 1..6
-
-
-def generate_bookmarks(src_pdf: str, out_pdf: str, headings: Iterable[Heading]) -> None:
-    """Write given headings into a new PDF file as outline/bookmarks."""
-    reader = PdfReader(src_pdf)
-    writer = PdfWriter()
-    for page in reader.pages:
-        writer.add_page(page)
-
-    # Build hierarchical outlines using a simple stack by levels
-    stack: List[Tuple[int, object]] = []  # (level, parent_ref)
-
-    for h in headings:
-        page_index = max(0, min(len(reader.pages) - 1, h.page - 1))
-        while stack and stack[-1][0] >= h.level:
-            stack.pop()
-        parent = stack[-1][1] if stack else None
-        dest = writer.add_outline_item(h.title, page_index, parent=parent)
-        stack.append((h.level, dest))
-
-    with open(out_pdf, "wb") as f:
-        writer.write(f)
-
-
-# -------------------- TOC parsing utilities --------------------
-
-_NUM_PREFIX_RE = re.compile(
-    r"^\s*(?P<num>(第\s*\d+[一二三四五六七八九十百千]*[章节节部分编]?)|((\d+\.)+\d+)|\d+)?\s*"
-)
-_TRAILING_PAGE_RE = re.compile(r"(?P<page>\d{1,5})\s*$")
-
-
-def _infer_level_from_numbering(num: Optional[str]) -> int:
-    if not num:
-        return 1
-    num = num.strip()
-    if num.startswith("第"):
-        # "第1章" style => top-level
-        return 1
-    if "." in num:
-        # "1.2.3" => level = segments + 1 (so 1.2 is level 2)
-        return min(6, max(1, num.count(".") + 1))
-    # Simple leading integer like "1" => level 1
-    return 1
-
-
-def parse_toc_lines(toc_text: str, page_offset: int = 0, min_len: int = 1) -> List[Heading]:
-    """
-    Parse a pasted TOC text into Heading entries.
-    - Each line should end with the book page number (digits)
-    - Leading numbering like "第1章" or "1.2" is used to infer the level
-    - page_offset is added to the parsed page number to map to PDF actual pages
-    """
-    headings: List[Heading] = []
-    for raw_line in toc_text.splitlines():
-        line = raw_line.strip()
-        if len(line) < min_len:
-            continue
-        # Detect and temporarily strip leading asterisk marker(s)
-        star_prefix = ""
-        m_star = re.match(r"^\*+\s*", line)
-        if m_star:
-            stars = m_star.group(0)
-            star_count = stars.count("*")
-            # Preserve star(s) without trailing space; spacing will be normalized later
-            star_prefix = ("*" * star_count)
-            line = line[m_star.end() :].lstrip()
-
-        # Extract trailing page digits
-        page_m = _TRAILING_PAGE_RE.search(line)
-        if not page_m:
-            continue
-        page_num = int(page_m.group("page"))
-        # Remove trailing page from the line
-        line_wo_page = line[: page_m.start()].rstrip()
-        # Extract leading numbering if exists
-        num_m = _NUM_PREFIX_RE.match(line_wo_page)
-        numbering = None
-        title_part = line_wo_page
-        if num_m:
-            numbering = num_m.group("num")
-            title_part = line_wo_page[num_m.end() :].strip()
-        # Build title while preserving numbering prefix (e.g., "第1章" or "1.1")
-        if numbering:
-            combined = f"{numbering.strip()} {title_part}".strip()
-        else:
-            combined = title_part
-        # Cleanup whitespace
-        title = re.sub(r"\s+", " ", combined)
-        if not title:
-            # fallback to raw without numbering
-            title = line_wo_page.strip()
-        # Restore asterisk prefix if present
-        if star_prefix:
-            # No space between star(s) and numbering/title
-            title = f"{star_prefix}{title}".strip()
-        level = _infer_level_from_numbering(numbering)
-        pdf_page = max(1, page_num + page_offset)
-        headings.append(Heading(title=title, page=pdf_page, level=level))
-
-    # Sort by page then by inferred level
-    headings.sort(key=lambda h: (h.page, h.level, h.title.lower()))
-    return headings
-
-
-## URL/website TOC fetching intentionally removed; only manual text input is supported.
-
-