Update expr.c

[JebediahS]

Filling the stack should start from element with index 0.

[ohkimur]

Thanks for the PR! The change from pre-increment/decrement to post-increment/decrement looks correct and fixes the off‑by‑one behavior:

• In push, stack[stack_pointer++] = element; is the right choice if stack_pointer tracks the next free slot.
• In pop, return stack[--stack_pointer]; correctly moves back to the last pushed element before reading.

A couple of suggestions to strengthen this and future contributions:

• Add brief comments documenting the stack invariant: “stack_pointer points to the next free slot.”
• Include small tests (or a demo main) for: push/pop single item, push until full, pop until empty, and verify strict LIFO.
• Consider returning a status from push/pop (e.g., int/bool with out‑param) instead of printing errors; returning 0 from pop is ambiguous if 0.0 is a valid value.
• Use size_t for stack_pointer and comparisons with STACK_SIZE.
• Optionally guard against overflow/underflow with asserts in debug builds.

Overall, nice, focused fix with clear correctness improvement. If you’re up for it, similar sanity checks and invariants across other stack/queue exercises would be super valuable.

[JebediahS]

Thanks for the PR! The change from pre-increment/decrement to post-increment/decrement looks correct and fixes the off‑by‑one behavior:

* In `push`, `stack[stack_pointer++] = element;` is the right choice if `stack_pointer` tracks the next free slot.

* In `pop`, `return stack[--stack_pointer];` correctly moves back to the last pushed element before reading.

A couple of suggestions to strengthen this and future contributions:

* Add brief comments documenting the stack invariant: “`stack_pointer` points to the next free slot.”

* Include small tests (or a demo `main`) for: push/pop single item, push until full, pop until empty, and verify strict LIFO.

* Consider returning a status from `push`/`pop` (e.g., int/bool with out‑param) instead of printing errors; returning `0` from `pop` is ambiguous if `0.0` is a valid value.

* Use `size_t` for `stack_pointer` and comparisons with `STACK_SIZE`.

* Optionally guard against overflow/underflow with asserts in debug builds.

Overall, nice, focused fix with clear correctness improvement. If you’re up for it, similar sanity checks and invariants across other stack/queue exercises would be super valuable.

Thank you for the suggestions, I would be happy to contribute more (if my time allows)!

[ohkimur]

@JebediahS Hello, are you still interested to contribute on this PR, or shall I close it?

[JebediahS]

@JebediahS Hello, are you still interested to contribute on this PR, or shall I close it?

Sorry for the delay, still interested!

[ohkimur]

@JebediahS Hello, are you still interested to contribute on this PR, or shall I close it?

Sorry for the delay, still interested!

Great! Whenever you want to merge this PR, just make sure you checkout the new CONTRIBUTIONS.md file, as this a new addition to the repo so it makes the code more consistent and a bit more modern.

[ohkimur]

@JebediahS Also, you'll need to merge the new changes from main first in your local branch to solve the merge conflicts.

[JebediahS]

Thanks for the PR! The change from pre-increment/decrement to post-increment/decrement looks correct and fixes the off‑by‑one behavior:

* In `push`, `stack[stack_pointer++] = element;` is the right choice if `stack_pointer` tracks the next free slot.

* In `pop`, `return stack[--stack_pointer];` correctly moves back to the last pushed element before reading.

A couple of suggestions to strengthen this and future contributions:

* Add brief comments documenting the stack invariant: “`stack_pointer` points to the next free slot.”

* Include small tests (or a demo `main`) for: push/pop single item, push until full, pop until empty, and verify strict LIFO.

* Consider returning a status from `push`/`pop` (e.g., int/bool with out‑param) instead of printing errors; returning `0` from `pop` is ambiguous if `0.0` is a valid value.

* Use `size_t` for `stack_pointer` and comparisons with `STACK_SIZE`.

* Optionally guard against overflow/underflow with asserts in debug builds.

Overall, nice, focused fix with clear correctness improvement. If you’re up for it, similar sanity checks and invariants across other stack/queue exercises would be super valuable.

To not mess up your style: In which form do you expect the tests to be set up? Should there be a test-folder exercise_X_XX/test/ folder with a expr_test.c or a shell expr_test.sh script which covers the critical cases?