hard wrap

[vmodamio]

Couldnt find whether this is implemented or not. Is there an option in conf.c to hard wrap to # cols ?

[kyx0r]

Hello, ughhh no. This is not how its implemented. Refer to the manual:

[#]gw Hard word wrap a line to # col limit. Default: 80
[#]gq Hard word wrap a buffer to # col limit. Default: 80

Use gw to word wrap a line or use gq to word wrap entire buffer.

[vmodamio]

In VIM for example, hard wrap with a textwidth of N, would let you type until the column N, and it wont break the line if you input a space afterwards. Any other character and will put the linebreak before the whole word. Now that you say, seems a little bit more complex than I though. Because the space character will be hidden from the editor (you wont see space in the first column of the next line).
Sorry to bother you with naive questions. I havent go throw your code exhaustively and I dont know which data structure you implemented for the buffer. How then can you store infinite length lines?!

[kyx0r]

Well, the beauty of Nextvi is that you don't have to make it as complex as in VIM. You are free to explore the problem space and find a way that works better by virtue of a simpler algorithm. Figure out the data and constraints and make it work how you want it to work.There is nothing putting sticks in your wheels here.

Please keep asking the naive questions as this is where most people give up. Don't give up, keep learning and soon enough you will understand everything.

On a technical side, the buffer is exactly what it sounds, its a buffer - array. More academic - a vector. Dynamically realizable array which can grow to infinity so long as there is enough memory.

Pay attention to how sbuf is implemented:
https://github.com/kyx0r/nextvi/blob/master/vi.h#L43

And look at the application of sbuf *sb in this function:
https://github.com/kyx0r/nextvi/blob/master/led.c#L412

[vmodamio]

Thank you! Ok, so if I look at the sbuf struct definition, the text is hold in a contiguos pice of memory of size s_sz, from which you keep track of the string length with s_n.
Then the buffer can reallocate more memory on demand.
If I might ask: So the full text is hold in one single sbuf? or there is one sbuf per line (apart of extra strings for commands, ex, current line, etc). How does this perform for searching, or editing big files, or yank/paste big chuncks of text?
Another thing is the uf8 encoding... characters have variable byte-size right? So this would complicate the wrapping mechanism.
The reason I need to start understanding the code from scratch is well, because it is fascinating, and second, because I would love to implement (a very stripped down version) it in RTOS (with very little RAM).

[kyx0r]

For insert mode, there is one buffer holding everything including prefix part, in between part and postfix part.
In general though, the lbuf holds lines of text, char * per line (not sbuf). So when it comes down to need to actually
edit the buffer, char * in lbuf gets replaced with sbuf->s and additional info for undo/redo tracking.

As for utf-8 portion, you can decide if you need this or not. Nextvi supports utf-8 everywhere. If you decide to use existing functions you will have that automatically.

[kyx0r]

To help you understand, this is the place where sb->s replaces the line(s) in lbuf.
https://github.com/kyx0r/nextvi/blob/master/vi.c#L1060

[kyx0r]

@vmodamio If you don't want word wrap, you can do a hard line wrap using this command:

:led:g/^/&80|\:grp2\:ka\:1,1,2?;>.(.)\:;'a\:&K\:grp:led

Change the 80 to your desired wrap column. This command only does one step, you can loop it using ? with the condition that you want or manually re-run it how ever many times you need.