The link you provided in your earlier post goes to a login page.
I don't have an account nor do I to wish to create one.
Perhaps you could provide an example or two of ChatGPT input and output (with respect to finding search terms).
| [reply] |
Well - it took a little asking but it come up with "subdirectory tree" or "subdirectory hierarchy". I'm not sure these are right and certainly don't correspond to kcott's "trie" but they do give a better term for Google...
Here's the conversation with my input in italics
Is there a name for this sort of directory structure where the directories are named after parts of the filename? Like this /var/images/f/o/foo.jpg
Yes, this type of directory structure is often referred to as a "hashed directory structure" or "hash-based directory structure". The directory names are usually created by applying a hash function to some part of the file name or contents, such as the first few characters or a checksum, to ensure that files are evenly distributed across the directories and to avoid having too many files in any one directory. This type of directory structure is commonly used in large-scale systems that need to store and manage a large number of files efficiently, such as content delivery networks, file-sharing platforms, and database systems.
That is not what I mean. In this structure there is no hashing. The first letter ('f' in the example) is the first directory then the second letter 'o' is the next level and so on. Like this - /f/o/foo.jpg
My apologies for misunderstanding your question. The directory structure you described is known as a "subdirectory tree" or "subdirectory hierarchy". It is a way of organizing files into a hierarchical structure by breaking down the filename into its constituent parts and using them to create nested directories. This type of structure can be useful for organizing large collections of files, as it allows for easy navigation and retrieval of files based on their names or other attributes. It is also commonly used in web applications, where files are often organized by user or content type.
| [reply] |
Thanks for taking the time to check that out.
I use DuckDuckGo (DDG). I used both "subdirectory tree" and "subdirectory hierarchy" as search terms.
I expanded to several (27") screenfuls of results for each;
I did only scan these but nothing leapt out as related to what the OP asked.
Did you try Google or any other search engine?
My real intelligence saw f/o/foo and guessed at a potential problem with what was actually asked;
I changed that to f/fo/foo, made a comparison with T/TI/TIMB, and answered "trie".
ChatGPT's artifical intelligence was unable to do that;
or, at least, was unable to ask "Did you mean f/fo/foo?" or add some query along those lines.
If I make a typo in a search term, DDG might show something along the lines of:
"Few results for psuedocode; did you mean pseudocode?".
The f/o/foo vs. f/fo/foo issue is more subtle.
I've known about tries and CPAN tarballs for years, and so was able to make an intelligent guess;
ChatGPT's AI can't make that intuitive leap from one to the other,
although I suspect it could answer a direct question about either.
An interesting exercise; however, I won't be signing up with ChatGPT to help with my internet searches anytime soon. :-)
[Aside:
Quite by accident, I came across this YouTube video,
"Let's build GPT: from scratch, in code, spelled out.",
a week or two ago.
It's about two hours long; I only watched the first few minutes; the code is Python; it may be of interest to some.
]
| [reply]
[d/l]
[select] |