摘要
近期有读论文的需求,就需要去了解一下论文到底要怎么读,同一个系列之间的论文如何作整理和归纳,之前也有了解过市面上有成熟的论文阅读工具,但是对于学生党来讲没什么性价比,在B站上看到沐神有讲解他的思路Typora作为工作中的md生产工具,我有一点浅显的认识希望和大家交流学习。Typora可以作为编辑工具,之前有被同事安利过,但是那个时候md格式还并不了解,今天重拾起,简单来讲我的做法就是Typora+gitee形成云端存储的一套方案,配套使用我自己的一些脚本,可以很好的实现论文阅读的功能。
一:Typora的安装
Typora 是一个所见即所得的 Markdown 跨平台写作工具,目前已经发布正式版,并且更改为付费模式,0.11.18_beta 是最后一个免费的测试版,有需要的可以选择下载。
Windows 用户
下载地址: [https://github.com/iuxt/src/releases/download/2.0/typora-0-11-18.exe](https://github.com/iuxt/src/releases/download/2.0/typora-0-11-18.exe)
0.11.18 现在被远程施法了,会提示过期无法使用,可以使用 0.9.96 版
下载地址:https://github.com/iuxt/src/releases/download/2.0/typora-setup-x64_0.9.96.exe
Mac 用户
下载地址: https://github.com/iuxt/src/releases/download/2.0/typora-0-11-18.dmg
Ubuntu 用户
下载地址:https://github.com/iuxt/src/releases/download/2.0/Typora_Linux_0.11.18_amd64.deb
安装方法
使用 apt 安装:
sudo apt install ./Typora_Linux_0.11.18_amd64.deb
如此你就完成了笔记编辑器的安装。
二:Gitee的配置
https://gitee.com在这里去作账户注册和登陆,新建仓库
在这里插入图片描述在本地新建立一个文件夹,在该文件下开命令行
#Git 全局设置:git config --global user.name "YourName"
git config --global user.email "YourInfo@user.noreply.gitee.com"#创建 git 仓库:mkdir paper
cd paper
git init
touch README.md
git add README.md
git commit -m "first commit"
git remote add origin https://gitee.com/YourName/paper.git
git push -u origin "master"#已有仓库?cd existing_git_repo
git remote add origin https://gitee.com/YourName/paper.git
git push -u origin "master"
如此每次更改后可以配合gitee去作同步。
三:脚本
脚本一:通过arxiv自动下载论文,提取论文标题,作者,日期,索引数等
arxiv_2_md.py
#!/usr/bin/env python3
import os
import re
import requests
import arxiv
from urllib.parse import urlparse, quotedef extract_arxiv_id(url: str) -> str:"""从arXiv链接中提取arXiv ID。形如:https://arxiv.org/abs/1605.08386则返回:1605.08386"""parsed = urlparse(url)if 'arxiv.org' not in parsed.netloc:raise ValueError("这不是一个有效的arXiv链接。")match = re.search(r'/abs/([0-9]+\.[0-9]+)', parsed.path)if not match:match = re.search(r'/pdf/([0-9]+\.[0-9]+)', parsed.path)if not match:raise ValueError("未能从链接中提取到arXiv ID。")return match.group(1)def fetch_arxiv_metadata(arxiv_id: str):"""使用arxiv Python包从arxiv获取元数据返回字典包含:title, authors, year, journal_ref, pdf_url"""search = arxiv.Search(id_list=[arxiv_id])paper = next(search.results(), None)if paper is None:raise ValueError("未能在arXiv找到对应论文信息。")journal_ref = paper.journal_ref if paper.journal_ref else "N/A"authors = [au.name for au in paper.authors]year = paper.published.yearreturn {"title": paper.title.strip(),"authors": authors,"year": year,"journal": journal_ref,"pdf_url": paper.pdf_url}def download_pdf(pdf_url: str, save_dir: str = "./pdfs") -> str:"""下载pdf文件到本地save_dir中,并返回本地文件相对路径。"""if not os.path.exists(save_dir):os.makedirs(save_dir)# 尝试从pdf_url中提取文件名basename = os.path.basename(pdf_url)if not basename.endswith(".pdf"):basename += ".pdf"local_filename = os.path.join(save_dir, basename)r = requests.get(pdf_url)r.raise_for_status()with open(local_filename, 'wb') as f:f.write(r.content)return local_filenamedef fetch_citation_count_by_arxiv_id(arxiv_id: str) -> int:"""调用 Semantic Scholar API 使用 ArXiv:<arxiv_id> 获取引用数。"""url = f"https://api.semanticscholar.org/graph/v1/paper/ArXiv:{arxiv_id}?fields=citationCount"r = requests.get(url)if r.status_code == 200:data = r.json()return data.get("citationCount", 0)return 0def fetch_citation_count_by_title(title: str) -> int:"""如果直接使用ArXiv ID获取不到合适引用数,则通过标题在 Semantic Scholar 搜索。取搜索结果中匹配度最高(即第一个结果)的citationCount作为参考。"""query = quote(title)url = f"https://api.semanticscholar.org/graph/v1/paper/search?query={query}&fields=title,citationCount"r = requests.get(url)if r.status_code == 200:data = r.json()papers = data.get("data", [])if papers:best_match = papers[0]if best_match["title"].lower().strip() == title.lower().strip():return best_match.get("citationCount", 0)return 0def fetch_citation_count(arxiv_id: str, title: str) -> int:"""尝试通过arxiv_id获取citationCount,如果为0则尝试通过标题获取。"""count = fetch_citation_count_by_arxiv_id(arxiv_id)if count == 0:# 如果通过arxiv_id获取不到或为0,尝试通过标题搜索count = fetch_citation_count_by_title(title)return countdef generate_markdown(md_filename: str, title: str, authors: list, journal: str, year: int, local_pdf_path: str, citation_count: int, arxiv_url: str):"""生成Markdown文件:包含标题、作者、期刊/会议信息、年份、本地PDF链接、原始arxiv链接和引用次数。"""authors_str = ", ".join(authors)rel_pdf_path = os.path.relpath(local_pdf_path)with open(md_filename, 'w', encoding='utf-8') as f:f.write(f"# {title}\n\n")f.write(f"- **Authors:** {authors_str}\n")f.write(f"- **Venue/Journal:** {journal}\n")f.write(f"- **Year:** {year}\n")f.write(f"- **Local PDF:** [{rel_pdf_path}]({rel_pdf_path})\n")f.write(f"- **ArXiv Link:** [{arxiv_url}]({arxiv_url})\n\n")f.write(f"**Citations:** {citation_count}\n")def main():# 输入 arxiv 链接arxiv_url = input("请输入arXiv链接:").strip()arxiv_id = extract_arxiv_id(arxiv_url)# 获取arxiv元数据meta = fetch_arxiv_metadata(arxiv_id)# 下载PDFlocal_pdf = download_pdf(meta["pdf_url"])# 获取引用数citation_count = fetch_citation_count(arxiv_id, meta["title"])# 询问用户md文件名default_md_name = f"{arxiv_id}.md"md_name_input = input(f"请输入要保存的Markdown文件名(不需扩展名,留空则使用 {default_md_name[:-3]}): ").strip()if md_name_input == "":md_filename = default_md_nameelse:md_filename = f"{md_name_input}.md"# 生成markdown文件generate_markdown(md_filename,meta["title"],meta["authors"],meta["journal"],meta["year"],local_pdf,citation_count,arxiv_url)print(f"Markdown文件已生成:{md_filename}")if __name__ == "__main__":main()
运行:
python ./arxiv_2_md.py
如果在输入arXiv后报错:
python ./arxiv_to_md_1.2.py
请输入arXiv链接:https://arxiv.org/abs/2410.24207
/home/crist/WorkSpace/3D-reconstruction-paper/./arxiv_to_md_1.2.py:30: DeprecationWarning: The 'Search.results' method is deprecated, use 'Client.results' insteadpaper = next(search.results(), None)
Traceback (most recent call last):File "/home/crist/WorkSpace/3D-reconstruction-paper/./arxiv_to_md_1.2.py", line 150, in <module>main()File "/home/crist/WorkSpace/3D-reconstruction-paper/./arxiv_to_md_1.2.py", line 125, in mainlocal_pdf = download_pdf(meta["pdf_url"])^^^^^^^^^^^^^^^^^^^^^^^^^^^^^File "/home/crist/WorkSpace/3D-reconstruction-paper/./arxiv_to_md_1.2.py", line 56, in download_pdfr = requests.get(pdf_url)^^^^^^^^^^^^^^^^^^^^^File "/home/crist/miniconda3/lib/python3.12/site-packages/requests/api.py", line 73, in getreturn request("get", url, params=params, **kwargs)^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^File "/home/crist/miniconda3/lib/python3.12/site-packages/requests/api.py", line 59, in requestreturn session.request(method=method, url=url, **kwargs)^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^File "/home/crist/miniconda3/lib/python3.12/site-packages/requests/sessions.py", line 589, in requestresp = self.send(prep, **send_kwargs)^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^File "/home/crist/miniconda3/lib/python3.12/site-packages/requests/sessions.py", line 703, in sendr = adapter.send(request, **kwargs)^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^File "/home/crist/miniconda3/lib/python3.12/site-packages/requests/adapters.py", line 633, in sendconn = self.get_connection_with_tls_context(^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^File "/home/crist/miniconda3/lib/python3.12/site-packages/requests/adapters.py", line 483, in get_connection_with_tls_contextproxy_manager = self.proxy_manager_for(proxy)^^^^^^^^^^^^^^^^^^^^^^^^^^^^^File "/home/crist/miniconda3/lib/python3.12/site-packages/requests/adapters.py", line 282, in proxy_manager_formanager = self.proxy_manager[proxy] = SOCKSProxyManager(^^^^^^^^^^^^^^^^^^File "/home/crist/miniconda3/lib/python3.12/site-packages/urllib3/contrib/socks.py", line 212, in __init__raise ValueError(f"Unable to determine SOCKS version from {proxy_url}")
ValueError: Unable to determine SOCKS version from socks://127.0.0.1:7890/
解决办法:
export ALL_PROXY=socks5://127.0.0.1:7890
export HTTP_PROXY=socks5://127.0.0.1:7890
export HTTPS_PROXY=socks5://127.0.0.1:7890
脚本二:
提取PDF中的图片,将我的脚本和pdf文件放到一起:
#!/usr/bin/env python3
import os
import subprocess
import tkinter as tk
from tkinter import messagebox, fontdef list_pdfs(directory="."):"""列出指定目录中的所有PDF文件并返回列表。"""pdfs = [f for f in os.listdir(directory) if f.lower().endswith('.pdf')]return pdfsdef extract_images(pdf_path, output_dir="images"):"""使用pdfimages从指定PDF中提取图片。"""if not os.path.exists(output_dir):os.makedirs(output_dir)base_name = os.path.splitext(os.path.basename(pdf_path))[0]output_prefix = os.path.join(output_dir, base_name)cmd = ["pdfimages", "-j", pdf_path, output_prefix]try:subprocess.run(cmd, check=True)return True, f"图片已提取到 {output_dir} 目录中,以 {base_name}-xxx 的形式命名。"except subprocess.CalledProcessError:return False, "提取图片失败,请确保已安装pdfimages工具。"def on_extract():selection = listbox.curselection()if not selection:messagebox.showwarning("警告", "请先选择一个PDF文件")returnindex = selection[0]pdf_file = pdfs[index]success, msg = extract_images(pdf_file)if success:messagebox.showinfo("提取完成", msg)else:messagebox.showerror("错误", msg)root = tk.Tk()
root.title("PDF图片提取器")# 设置全局字体
root.option_add("*Font", "Helvetica 12")pdfs = list_pdfs(".")frame = tk.Frame(root)
frame.pack(padx=10, pady=10, fill="both", expand=True)label = tk.Label(frame, text="请选择一个PDF文件:", font=("Helvetica", 12, "bold"))
label.pack(anchor="w")listbox = tk.Listbox(frame, height=10)
listbox.pack(fill="both", expand=True)for pdf in pdfs:listbox.insert(tk.END, pdf)if not pdfs:listbox.insert(tk.END, "当前目录未找到PDF文件")btn_frame = tk.Frame(root)
btn_frame.pack(pady=5)
extract_btn = tk.Button(btn_frame, text="提取图片", font=("Helvetica", 12))
extract_btn.config(command=on_extract)
extract_btn.pack()root.mainloop()
这样就可以把图片保存到img文件夹下了