How Github's Code Search works?

[@TheMrSnuSnu]

I love how well researched your videos are

[@core_dump]

Thanks! ❤️

[@Sibearian_]

I am very excited for this feature. Thanks for explaining the concept!

[@core_dump]

Thanks for watching!

[@zinebaassila]

🙌 Wow, thank you so much for the clear and perfect explanation of how GitHub's Code Search works! You've made it so easy to understand. 🚀 Your knowledge-sharing is greatly appreciated! 💯

[@core_dump]

Great to hear!

[@uchennaofoma4624]

This is my first comment on your channel 😀, I've been following keenly for a while now. But yeah, it blows my mind how crazy goof engineers are, guess I've got a lot of learning to do huh?

[@core_dump]

Thanks... do watch the other videos as well! And share it with your friends!

[@shashankpathak63]

Lazy loading of indices is done like searching the value space for a key ? And if searching is done is any meta-heuristic sort of search is performed or a simple binary search on this sorted space works? I had one more doubt how similarity between two git repos are calculated do they use text similarity or since as you mentioned they are git objects instead of text file they have some similarity index evaluation? Btw an awesome video got to learn a lot !!

[@user-dc9zo7ek5j]

I don't get why they are not using grep to search within files instead of this complexity. The github search functionality always sucked. I tried a couple of times to search code on github and even though I've written specific word to search for, it never returned the results I've wanted. Typing dot (one of the most used symbols) to guide the search, returns no results, although the code exists, example "user.GetAge", double quotation also. We're used to searching inside by content, not "relative" or suggestions. Their solution forces me to clone the whole repo and search whatever I want with grep myself.

[@core_dump]

In the linked blog, they have mentioned reasons why they can't use grep. I am using the new search system for a few days, it's not perfect, but it's a major improvement from the previous one.

[@khanra17]

🦀