apply, seq, libcommon: Improvements to the code internals by takusuman · Pull Request #76 · Projeto-Pindorama/heirloom-ng

takusuman · 2025-11-24T12:55:49Z

Closes #75.

…trmenta.h

Also made it faster per replacing sprintf() calls with sputchar() from strmenta.h/sputs.h.

… at apply.c

…ply.c. At the cost of some CPU cycles, of course.

I haven't tested this yet since I'm doing my groceries at this precise moment.

…pply.c.

free(cmdl) for every return by buildcmd().

The comportment remains.

"Bazinga!" 🤓

"char *argv[]" -> "char *const argv[]".

Use pfmt(3) instead of printf(3) because it treats the non-literal string properly and doesn't create warnings. When using MM_NOSTD, it comports just like printf(3). PERHAPS it could be better to use vprintf(3) instead.

Now using "& MM_NOGET" so pfmt(3) comports just like printf(3).

Now using "^ MM_NOGET" so pfmt(3) comports just like printf(3).

…*'" at seq.c.

Also sneakly add back the 'static's for global variables.

arthurbacci · 2025-12-31T12:41:06Z

By the way, regarding #34, I think we shall start compiling programs with -Wshadow, -Wunused, -Wpedantic, -Wall and -Wextra when testing. ~~Perhaps -Wall with -Wextra are sufficient, but I hadn't checked the GCC docs yet.~~

I don't even compile what I code, the environment isn't working very well here.

arthurbacci · 2025-12-31T12:49:57Z

For some reason f4fc181 removes void main() from seq.c and not 9dda692.

takusuman · 2025-12-31T21:08:02Z

By the way, regarding #34, I think we shall start compiling programs with -Wshadow, -Wunused, -Wpedantic, -Wall and -Wextra when testing. ~~Perhaps -Wall with -Wextra are sufficient, but I hadn't checked the GCC docs yet.~~

I don't even compile what I code, the environment isn't working very well here.

Check your build/mk.config later.

For some reason f4fc181 removes void main() from seq.c and not 9dda692.

Does it make a huge difference? This will be all squashed into a single commit later.

arthurbacci · 2025-12-31T21:52:32Z

By the way, regarding #34, I think we shall start compiling programs with -Wshadow, -Wunused, -Wpedantic, -Wall and -Wextra when testing. ~~Perhaps -Wall with -Wextra are sufficient, but I hadn't checked the GCC docs yet.~~

I don't even compile what I code, the environment isn't working very well here.

Check your build/mk.config later.

For some reason f4fc181 removes void main() from seq.c and not 9dda692.

Does it make a huge difference? This will be all squashed into a single commit later.

Why are you squashing PRs?

takusuman · 2026-01-01T00:25:36Z

By the way, regarding #34, I think we shall start compiling programs with -Wshadow, -Wunused, -Wpedantic, -Wall and -Wextra when testing. ~~Perhaps -Wall with -Wextra are sufficient, but I hadn't checked the GCC docs yet.~~

I don't even compile what I code, the environment isn't working very well here.

Check your build/mk.config later.

For some reason f4fc181 removes void main() from seq.c and not 9dda692.

Does it make a huge difference? This will be all squashed into a single commit later.

Why are you squashing PRs?

Often a ton of commits that do one thing.
This didn't happen at large PR for the 20240220-fix branch.

takusuman · 2026-01-01T01:52:57Z

@arthurbacci Let's go, last release of the year.

@arthurbacci

@arthurbacci check.

Also sneakly add parenthesis to a multiplication in a function call, so we can start creating consistent rules of style.

@arthurbacci

@arthurbacci Perhaps I could say sorry for the parenthesis at index + (number - 1).

takusuman · 2026-01-04T08:58:38Z

@arthurbacci Also removed the ch = cmd[c] thing.

Better late than never.

takusuman · 2026-01-09T21:25:34Z

I think we would also need to do some fixes at libcommon/strjoin.c, @arthurbacci.

This function has undefined behavior if elems[0] is a null pointer. It should just allocate and return an empty string in this case. Also UB if malloc fails.

chqrlie at StackOverflow

I think the first is not the case since the for loop having *elems as a condition would already make the function stop if a NULL is found. For the second, I will do mea culpa and admit that I should've tested if joinbufp != NULL before continuing. Simple fixes in any case.

…lems[0] is NULL.

takusuman · 2026-01-10T18:44:59Z

Added a if (!elems[0]) because, if it weren't for it, we'd be returning an empty string instead of a NULL.

I hope my cellphone's digital keyboard hasn't garbled a thing.

takusuman added 12 commits October 7, 2024 22:03

chore: putchar()-oid macro for copying a character into a string at s…

ffabc6d

…trmenta.h

fix: Memory fault when using an invalid magic number at apply(1B).

fbca101

Also made it faster per replacing sprintf() calls with sputchar() from strmenta.h/sputs.h.

Merge branch 'master' into apply-pick-impl

c9eeccf

fix: free() *cmd and *shpath, also check for calloc()/malloc() errors…

19e0f6d

… at apply.c

Merge branch 'master' into apply-pick-impl

37df0f0

fix: Use sizeof(char), not sizeof(char *) at apply.c.

c5164d5

chore: Estimate tightly the number of characters used by cmdbuf at ap…

bc53e0a

…ply.c. At the cost of some CPU cycles, of course.

chore: Initial attempt to compress the block of code at apply.c.

3436312

fix: forgot to count spaces when enamo at apply.c.

a35f157

I haven't tested this yet since I'm doing my groceries at this precise moment.

chore: Comments messed up by GBoard, also have better formatting at a…

9825832

…pply.c.

feat: Use CFLAGSU at apply(1B) compilation.

697767e

fix: Allocate the right amount of memory at seq.c.

0d4b771

takusuman changed the title ~~apply: Improvements to the code internals~~ apply, seq: Improvements to the code internals Nov 25, 2025

arthurbacci reviewed Nov 26, 2025

View reviewed changes

Comment thread apply/apply.c Outdated

chore: Some ternary a e s t h e t h i c at apply.c.

b202db6

takusuman marked this pull request as ready for review November 27, 2025 02:45

takusuman added 14 commits November 28, 2025 05:32

fix: Possible memory leak at apply.c;

86973a3

free(cmdl) for every return by buildcmd().

feat: Why not -e for exiting on error at apply(1B)?

702a90c

chore: Better formatting for apply.c, I think.

33e41c2

chore: Or this looks better (at apply.c)?

f13159c

The comportment remains.

docs: Mention '-e' at apply(1B) manual page.

4e1ccd4

chore: Replace ternary at apply.c.

21318dd

chore: Made the code less clerverish and more legible at apply.c.

1381e83

"Bazinga!" 🤓

fix: "unsafe buffer usage" at seq.c.

520dd99

"char *argv[]" -> "char *const argv[]".

fix: "format string is not a string literal" at seq.c.

cb665ac

Use pfmt(3) instead of printf(3) because it treats the non-literal string properly and doesn't create warnings. When using MM_NOSTD, it comports just like printf(3). PERHAPS it could be better to use vprintf(3) instead.

fix: the last fix at seq.c.

39ad148

Now using "& MM_NOGET" so pfmt(3) comports just like printf(3).

fix: the last fix at seq.c.

82140ad

Now using "^ MM_NOGET" so pfmt(3) comports just like printf(3).

fix: "format specifies type 'double' but the argument has type 'char …

239d000

…*'" at seq.c.

fix: Type mismatching at seq.c.

0264122

fix: My dyslexia at apply(1B) and apply.c.

450b7c9

takusuman added 5 commits December 29, 2025 14:05

fix: A handful of warnings at apply.c.

f4fc181

Also sneakly add back the 'static's for global variables.

chore: Remove void main() from seq.c.

9dda692

chore: All still static at seq.c.

6a5173d

Merge branch 'master' into apply-pick-impl

53dfc56

chore: Enforce that now apply and seq are pure C99.

bdde15a

takusuman added 2 commits December 31, 2025 21:04

chore: const doesn't make a difference anymore, also return at seq.c.

c62c738

chore: Remove """unnecessary""" parenthesis at apply.c.

d63556f

arthurbacci requested changes Jan 2, 2026

View reviewed changes

takusuman added 2 commits January 2, 2026 18:11

chore: This just feels erred... at apply.c.

903a510

@arthurbacci check.

chore: This feels like torture... at apply.c.

889427d

arthurbacci requested changes Jan 3, 2026

View reviewed changes

Comment thread apply/apply.c Outdated

Comment thread apply/apply.c Outdated

Comment thread apply/apply.c Outdated

Comment thread seq/seq.c Outdated

Comment thread seq/seq.c Outdated

takusuman added 3 commits January 3, 2026 15:37

fix: 'cmdbuflen * sizeof(char)' at apply.c.

d2a059a

chore: Don't use 'ch' for storing cmd[c] at apply.c.

ccaea0c

Also sneakly add parenthesis to a multiplication in a function call, so we can start creating consistent rules of style.

chore: Remove, in fact, unnecessary parts of the code at apply.c.

059605f

@arthurbacci Perhaps I could say sorry for the parenthesis at index + (number - 1).

chore: Merry Christmas and a happy New Year!

5df5d11

Better late than never.

arthurbacci reviewed Jan 6, 2026

View reviewed changes

Comment thread apply/apply.c Outdated

fix: For libcommon/strjoin.c: check for malloc, also return NULL if e…

3260946

…lems[0] is NULL.

takusuman changed the title ~~apply, seq: Improvements to the code internals~~ apply, seq, libcommon: Improvements to the code internals Jan 10, 2026

chore: More unnecessary parenthesis at apply.c.

db5b5dc

takusuman requested a review from arthurbacci January 10, 2026 18:51

chore: More "unnecessary parenthesis", but at seq.c.

cd19262

I hope my cellphone's digital keyboard hasn't garbled a thing.

Conversation

takusuman commented Nov 24, 2025

Uh oh!

Uh oh!

arthurbacci commented Dec 31, 2025

Uh oh!

arthurbacci commented Dec 31, 2025

Uh oh!

takusuman commented Dec 31, 2025

Uh oh!

arthurbacci commented Dec 31, 2025

Uh oh!

takusuman commented Jan 1, 2026

Uh oh!

takusuman commented Jan 1, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

takusuman commented Jan 4, 2026

Uh oh!

Uh oh!

takusuman commented Jan 9, 2026

Uh oh!

takusuman commented Jan 10, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants