[BUG] dynarray_set() increments length on failure

[seesee010]

Bug Description

When dynarray_set() fails because index is out of bounds, it incorrectly increments the length before returning false.

Location

src/access.c lines 28-30

Current Code

if (index >= array->capacity) {
    array->length++;  // ← Why increment on FAILURE?!
    return false;
}

Problem

1. Increments length when operation fails
2. Causes length corruption
3. Makes push() break because it relies on accurate length

Expected Behavior

Should return false WITHOUT modifying array state on failure.

Proposed Fix

if (index >= array->capacity) {
    return false;  // Just fail, don't corrupt state
}

Impact

• Severity: High
• Causes incorrect length tracking
• Breaks dynarray_push() functionality
• Violates principle: failed operations shouldn't modify state

Reproduction

DynArray *arr = dynarray_create(sizeof(int), 2);
int val = 10;
dynarray_set(arr, 5, &val);  // Should fail, but corrupts length
printf("Length: %zu\n", dynarray_length(arr));  // Shows 1 instead of 0!