hackmd2cyberblog.py

[bliutech]

Summary

A utility script to automatically take a HackMD markdown file and prepare it for the Cyber Blog.

Test Plan

Tested locally. Here is a sample run. Note that it gracefully errors with an invalid image capture of the URL (not a markdown image).

$ python3 data/scripts/hackmd2cyberblog.py data/blog/2024-12-12-dummy-blog.md
[*] Downloading images...
[*] Downloading image: https://hackmd.io/_uploads/B1-LPRDXJx.png -> public/images/blog/2024-12-12-dummy-blog/B1-LPRDXJx.png
[*] Downloading image: https://hackmd.io/_uploads/SkWtd0wXJx.png -> public/images/blog/2024-12-12-dummy-blog/SkWtd0wXJx.png
[*] Downloading image: https://hackmd.io/_uploads/Skl6O0PXyx.png -> public/images/blog/2024-12-12-dummy-blog/Skl6O0PXyx.png
[*] Downloading image: https://hackmd.io/_uploads/Hk_zFRwm1g.png -> public/images/blog/2024-12-12-dummy-blog/Hk_zFRwm1g.png
[*] Downloading image: https://hackmd.io/_uploads/BJvmOXnfke.png -> public/images/blog/2024-12-12-dummy-blog/BJvmOXnfke.png
[*] Downloading image: https://hackmd.io/_uploads/rJguJav7Jg.png -> public/images/blog/2024-12-12-dummy-blog/rJguJav7Jg.png
[*] Downloading image: https://hackmd.io/_uploads/BkCeSQnG1x.png -> public/images/blog/2024-12-12-dummy-blog/BkCeSQnG1x.png
[*] Failed to download image: https://hackmd.io/_uploads/BJe_rX2fke.png" width="50%">

Example of a XML file in our seed corpus 
We performed coverage-guided fuzzing via writing a harness with Honggfuzz.

```cpp
#include <cstdint>
#include <string>
#include <tinyxml2.h>

extern "C" int LLVMFuzzerTestOneInput(const std::uint8_t *data, std::size_t size
[*] Downloading image: https://hackmd.io/_uploads/Sy6U1NnM1e.png -> public/images/blog/2024-12-12-dummy-blog/Sy6U1NnM1e.png
[*] Downloading image: https://hackmd.io/_uploads/H1-ehX2Mkx.jpg -> public/images/blog/2024-12-12-dummy-blog/H1-ehX2Mkx.jpg
[*] Downloading image: https://hackmd.io/_uploads/Hyix0mhzye.png -> public/images/blog/2024-12-12-dummy-blog/Hyix0mhzye.png
[*] Downloading image: https://hackmd.io/_uploads/BJ9mFWamJx.png -> public/images/blog/2024-12-12-dummy-blog/BJ9mFWamJx.png
[*] Downloading image: https://hackmd.io/_uploads/rk4R67hz1e.png -> public/images/blog/2024-12-12-dummy-blog/rk4R67hz1e.png
[*] Downloading image: https://hackmd.io/_uploads/HJuiJVnzkl.png -> public/images/blog/2024-12-12-dummy-blog/HJuiJVnzkl.png
[*] Downloading image: https://hackmd.io/_uploads/rJ0aJ4nfyl.png -> public/images/blog/2024-12-12-dummy-blog/rJ0aJ4nfyl.png
[*] Downloading image: https://hackmd.io/_uploads/rkBqjr3z1x.png -> public/images/blog/2024-12-12-dummy-blog/rkBqjr3z1x.png
[*] Done!

[bliutech]

Added a quick fix to allow failed image downloads to have atomic failures (we shouldn't replace image URLs in the markdown that failed to download) and errors gracefully.