lanyuanxiaoyao
2b81dd49fe
将所有 HTML Parser 的函数签名从接收 HTML 字符串改为接收文件路径， 与其他 Reader（PDF、DOCX 等）保持一致。 主要变更： - 修改 PARSERS 列表，移除 lambda 表达式，直接传递函数引用 - 在 HtmlReader.parse() 中统一管理临时文件（UTF-8 编码） - 每个 Parser 使用独立的临时文件副本，用完即清理 - 移除 download_and_parse() 方法，逻辑合并到 parse() 中 - 更新相关测试，改为直接传递文件路径 受影响的 Parser： - trafilatura.parse(html_content) -> parse(file_path) - domscribe.parse(html_content) -> parse(file_path) - markitdown.parse(html_content, temp_file_path) -> parse(file_path) - html2text.parse(html_content) -> parse(file_path)
79 lines
2.5 KiB
Python
79 lines
2.5 KiB
Python
"""HTML/URL 文件阅读器，支持多种解析方法。"""
|
||
|
||
import os
|
||
import tempfile
|
||
from typing import List, Optional, Tuple
|
||
|
||
from scripts.readers.base import BaseReader
|
||
from scripts.utils import is_url
|
||
from scripts.utils import encoding_detection
|
||
|
||
from . import cleaner
|
||
from . import downloader
|
||
from . import trafilatura
|
||
from . import domscribe
|
||
from . import markitdown
|
||
from . import html2text
|
||
|
||
|
||
PARSERS = [
|
||
("trafilatura", trafilatura.parse),
|
||
("domscribe", domscribe.parse),
|
||
("MarkItDown", markitdown.parse),
|
||
("html2text", html2text.parse),
|
||
]
|
||
|
||
|
||
class HtmlReader(BaseReader):
|
||
"""HTML/URL 文件阅读器"""
|
||
|
||
def supports(self, file_path: str) -> bool:
|
||
return is_url(file_path) or file_path.lower().endswith(('.html', '.htm'))
|
||
|
||
def parse(self, file_path: str) -> Tuple[Optional[str], List[str]]:
|
||
"""解析 HTML 文件或 URL"""
|
||
all_failures = []
|
||
|
||
# 步骤 1: 获取 HTML 内容
|
||
if is_url(file_path):
|
||
# URL 路径: 下载 HTML
|
||
html_content, download_failures = downloader.download_html(file_path)
|
||
all_failures.extend(download_failures)
|
||
if html_content is None:
|
||
return None, all_failures
|
||
else:
|
||
# 本地文件路径: 读取文件
|
||
if not os.path.exists(file_path):
|
||
return None, ["文件不存在"]
|
||
html_content, error = encoding_detection.read_text_file(file_path)
|
||
if error:
|
||
return None, [f"- {error}"]
|
||
|
||
# 步骤 2: 清理 HTML 内容
|
||
html_content = cleaner.clean_html_content(html_content)
|
||
|
||
# 步骤 3: 对每个 Parser 创建独立的临时文件并尝试解析
|
||
for parser_name, parser_func in PARSERS:
|
||
# 创建临时文件
|
||
fd, temp_file_path = tempfile.mkstemp(suffix='.html', text=True)
|
||
try:
|
||
# 写入清理后的 HTML 内容（UTF-8 编码）
|
||
with os.fdopen(fd, 'w', encoding='utf-8') as f:
|
||
f.write(html_content)
|
||
|
||
# 调用 Parser 解析
|
||
content, error = parser_func(temp_file_path)
|
||
if content is not None:
|
||
return content, all_failures
|
||
else:
|
||
all_failures.append(f"- {parser_name}: {error}")
|
||
finally:
|
||
# 清理临时文件
|
||
try:
|
||
os.unlink(temp_file_path)
|
||
except Exception:
|
||
pass
|
||
|
||
# 所有 Parser 都失败
|
||
return None, all_failures
|