185 lines
8.2 KiB
Python
185 lines
8.2 KiB
Python
|
||
"""
|
||
toc_parse.py (OOP版)
|
||
-------------------
|
||
功能：
|
||
- 解析EPUB电子书的toc.ncx目录文件，递归构建章节树结构。
|
||
- 支持通过ref和filepos查找完整label路径。
|
||
- 支持通过selectedtext在html文件中定位章节标题。
|
||
- 兼容多种EPUB格式，支持批量测试。
|
||
|
||
依赖：config.py 统一管理路径和配置项。
|
||
主要接口：TOCParser
|
||
- parse_navpoints(navpoints)：递归解析navPoint节点，返回章节树结构。
|
||
- find_label_path(node, ref, filepos, path)：查找指定ref和filepos的章节label路径。
|
||
- find_section_by_selectedtext(html_path, selectedtext)：通过选中文本定位章节标题。
|
||
- parse_html_title(html_path)：解析html文件标题。
|
||
依赖：BeautifulSoup4, pprint, os, typing
|
||
"""
|
||
import config
|
||
from bs4 import BeautifulSoup
|
||
import os
|
||
|
||
class TOCParser:
|
||
def __init__(self):
|
||
pass
|
||
|
||
@staticmethod
|
||
def find_section_by_selectedtext(html_path, selectedtext):
|
||
try:
|
||
with open(html_path, 'r', encoding='utf-8') as f:
|
||
soup = BeautifulSoup(f, 'html.parser')
|
||
for elem in soup.find_all(string=True):
|
||
if selectedtext and selectedtext.strip() and selectedtext.strip() in elem:
|
||
parent = elem.parent
|
||
while parent:
|
||
prev = parent.previous_sibling
|
||
while prev:
|
||
if prev.name and prev.name.lower() in ['h1','h2','h3','h4','h5','h6']:
|
||
return prev.get_text(strip=True)
|
||
prev = prev.previous_sibling
|
||
parent = parent.parent
|
||
for tag in ['h1','h2','h3','h4','h5','h6']:
|
||
h = soup.find(tag)
|
||
if h and h.get_text(strip=True):
|
||
return h.get_text(strip=True)
|
||
except Exception:
|
||
pass
|
||
return None
|
||
|
||
@staticmethod
|
||
def parse_html_title(html_path):
|
||
try:
|
||
with open(html_path, 'r', encoding='utf-8') as f:
|
||
soup = BeautifulSoup(f, 'html.parser')
|
||
if soup.title and soup.title.string:
|
||
return soup.title.string.strip()
|
||
for tag in ['h1','h2','h3','h4','h5','h6']:
|
||
h = soup.find(tag)
|
||
if h and h.get_text(strip=True):
|
||
return h.get_text(strip=True)
|
||
except Exception:
|
||
pass
|
||
return None
|
||
|
||
@staticmethod
|
||
def parse_navpoints(navpoints):
|
||
result = {}
|
||
for navpoint in navpoints:
|
||
label = navpoint.navLabel.text.strip().strip('"“”')
|
||
src = navpoint.content["src"]
|
||
if "#" in src:
|
||
ref, filepos = src.split("#", 1)
|
||
else:
|
||
ref, filepos = src, None
|
||
entry = {
|
||
"label": label,
|
||
"ref": ref,
|
||
"filepos": filepos,
|
||
"children": TOCParser.parse_navpoints(navpoint.find_all("navPoint", recursive=False))
|
||
}
|
||
result[navpoint.get("id")] = entry
|
||
return result
|
||
|
||
@staticmethod
|
||
def find_label_path(node, ref, filepos=None, path=None):
|
||
if path is None:
|
||
path = []
|
||
if isinstance(node, dict):
|
||
nodes = node.values() if "label" not in node else [node]
|
||
for v in nodes:
|
||
if "label" in v:
|
||
new_path = path + [v["label"]]
|
||
if v["ref"] == ref and (filepos is None or v["filepos"] == filepos):
|
||
title = " / ".join(new_path)
|
||
return title
|
||
title = TOCParser.find_label_path(v["children"], ref, filepos, new_path)
|
||
if title:
|
||
return title
|
||
if filepos is not None:
|
||
for v in nodes:
|
||
if "label" in v:
|
||
new_path = path + [v["label"]]
|
||
if v["ref"].split("#", 1)[0] == ref.split("#", 1)[0]:
|
||
title = " / ".join(new_path)
|
||
return title
|
||
title = TOCParser.find_label_path(v["children"], ref, None, new_path)
|
||
if title:
|
||
return title
|
||
if path == [] and ref and ref.endswith('.html'):
|
||
caller_dir = os.path.dirname(os.path.abspath(__file__))
|
||
search_dirs = [caller_dir, os.getcwd()]
|
||
for d in search_dirs:
|
||
html_path = os.path.join(d, ref)
|
||
if os.path.isfile(html_path):
|
||
title = TOCParser.parse_html_title(html_path)
|
||
if title:
|
||
return title
|
||
for d in search_dirs:
|
||
for root, _, files in os.walk(d):
|
||
if ref in files:
|
||
html_path = os.path.join(root, ref)
|
||
title = TOCParser.parse_html_title(html_path)
|
||
if title:
|
||
return title
|
||
return None
|
||
|
||
if __name__ == "__main__":
|
||
# ==== 批量测试指定toc/html/filepos列表 ====
|
||
test_cases = [
|
||
[config.EXAMPLES_DIR + "/epub_format_1", "index_split_015.html", "filepos684970"],
|
||
[config.EXAMPLES_DIR + "/epub_format_2", "Text/8c7276f38ead4738ee19249418898c18_split_006.html", "sigil_toc_id_12"],
|
||
[config.EXAMPLES_DIR + "/epub_format_3", "Text/011.xhtml", ""],
|
||
[config.EXAMPLES_DIR + "/epub_format_4", "xhtml/p-006.xhtml", ""],
|
||
[config.EXAMPLES_DIR + "/变宋", "text/part0005.html", ""],
|
||
[config.EXAMPLES_DIR + "/变宋", "text/part0002_split_003.html", ""],
|
||
[config.EXAMPLES_DIR + "/规训与惩罚", "index_split_006.html", ""],
|
||
[config.EXAMPLES_DIR + "/政治哲學的12堂Podcast", "ch1.xhtml#_idParaDest-4", ""],
|
||
]
|
||
for epub_dir, html_file, filepos in test_cases:
|
||
toc_path = None
|
||
for root, _, files in os.walk(epub_dir):
|
||
for f in files:
|
||
if f.lower() == "toc.ncx":
|
||
toc_path = os.path.join(root, f)
|
||
break
|
||
if toc_path:
|
||
break
|
||
print(f"\n==== 测试 epub: {epub_dir} html: {html_file} filepos: {filepos} ====")
|
||
if not toc_path:
|
||
print(f"未找到toc.ncx: {epub_dir}")
|
||
continue
|
||
try:
|
||
with open(toc_path, "r", encoding="utf-8") as f:
|
||
soup = BeautifulSoup(f, "xml")
|
||
nav_map = soup.find("navMap")
|
||
toc_tree = TOCParser.parse_navpoints(nav_map.find_all("navPoint", recursive=False))
|
||
label_path = TOCParser.find_label_path(toc_tree, html_file, filepos)
|
||
print(f"find_label_path: {label_path if label_path else '未找到章节/标题'}")
|
||
html_path = os.path.join(epub_dir, html_file.split('#')[0])
|
||
if os.path.exists(html_path):
|
||
title = TOCParser.parse_html_title(html_path)
|
||
print(f"解析html标题: {html_path} => {title if title else '未找到标题'}")
|
||
selectedtext = '从变法思想看，王安石变法最大的魅力是“民不加赋而国用足”：老百姓上缴的税率不增，国库的总收入仍可以'
|
||
section = TOCParser.find_section_by_selectedtext(html_path, selectedtext)
|
||
print(f"selectedtext定位到的章节标题: {section if section else '未找到相关标题'}")
|
||
else:
|
||
print(f"未找到html文件: {html_path}")
|
||
except Exception as e:
|
||
print(f"测试失败: {e}")
|
||
|
||
print("\n==== 测试: 变宋笔记章节定位和html标题解析 ====")
|
||
note_idref = 'text/part0002_split_003.html'
|
||
note_filepos = None
|
||
bian_song_toc = config.EXAMPLES_DIR + "/变宋/toc.ncx"
|
||
if os.path.exists(bian_song_toc):
|
||
with open(bian_song_toc, "r", encoding="utf-8") as f:
|
||
soup = BeautifulSoup(f, "xml")
|
||
nav_map = soup.find("navMap")
|
||
toc_tree = TOCParser.parse_navpoints(nav_map.find_all("navPoint", recursive=False))
|
||
label_path = TOCParser.find_label_path(toc_tree, note_idref, note_filepos)
|
||
print(f"查找 {note_idref}: ", label_path if label_path else "未找到章节，尝试解析html标题")
|
||
else:
|
||
print(f"未找到toc.ncx: {bian_song_toc}")
|
||
|