Re: tangle option to not write a file with same contents?

[Max Nikulin]

On 28/10/2021 11:04, Greg Minshall wrote:
> i wonder if it would be reasonable to add an option such that, when
> tangling, `org-babel-tangle` would not write a file with the
> already-existing contents of the target file?

Are you going to celebrate a decade of the feature request?

https://list.orgmode.org/orgmode/CAFDswJtMzFZgRQqsFtiQNT7c1uFE7HV4kek5j9akx1yAdWZZEg@mail.gmail.com/T/#u
Holger Hoefling @ 2011-11-18 13:17 UTC
Not overwriting unchanged source code files when tangling

> this would be helpful, e.g., for those of us who use make(1)-based work
> flows.

I agree with you, make is wide spread and fast tool that is really 
convenient in simple cases.

Some hash-based build systems are mentioned in that thread. Since that 
time more more similar tools have appeared, e.g. buck, reimplementations 
of DJB's redo https://cr.yp.to/redo.html

> then, if this might generally be thought useful, i wonder if this should
> be implemented as specifically this, or whether we might implement a
> callback at the appropriate point in `org-babel-tangle` asking whether
> or not to proceed.  (then, the user's callback routine could do the
> comparison.)
>
> if we do "specifically this", i would suggest that this comparison be
> dead simple: read in the existing file's contents into some hidden
> buffer, and use `compare-buffer-substrings` to compare point-{min,max}
> of both.

Tom Gillespie mentioned that it is easy to lost modifications of tangled 
files created during debugging.
https://list.orgmode.org/orgmode/CA+G3_PNo8i8U9rOrC8BBydjEv9=LKgtJBopwX9J6vyNXdnjtXg@mail.gmail.com/T/#u

I think, some header may be added to tangled file containing hash of 
rest part of file. So the file may be checked for user modifications 
before overwriting, that hash may be compared with the new buffer to 
keep existing file in place.

It seems `compare-buffer-substrings` has more logic than just byte to 
byte comparison. Is it to handle alternatives for unicode character 
alternatives? For tangled buffer it should be size that is checked first...