Fix rbtree foreach traversal#9
Merged
Merged
Conversation
Reviewer's GuideAdds targeted test coverage for om_rbtree_foreach traversal and fixes the internal recursive implementation to correctly traverse all nodes and honor early-termination semantics. Flow diagram for updated _om_rbtree_foreach traversal and early terminationflowchart TD
A_start["_om_rbtree_foreach(node, fun, arg)"] --> B_isNull{"node == NULL?"}
B_isNull -- "yes" --> B_retTrue["return true"]
B_isNull -- "no" --> C_callLeft["call _om_rbtree_foreach(node.left, fun, arg)"]
C_callLeft --> C_leftResult{"left returned true?"}
C_leftResult -- "no" --> C_retFalse["return false"]
C_leftResult -- "yes" --> D_callFun["call fun(node, arg)"]
D_callFun --> D_funResult{"fun returned true?"}
D_funResult -- "no" --> D_retFalse["return false"]
D_funResult -- "yes" --> E_callRight["call _om_rbtree_foreach(node.right, fun, arg)"]
E_callRight --> E_retRight["return value from right subtree"]
File-Level Changes
Possibly linked issues
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
Jiu-xiao
changed the title
Jiu-xiao
changed the title
![sourcery-ai[bot]](https://avatars.githubusercontent.com/in/48477?s=60&v=4){kind=small}
There was a problem hiding this comment.
Hey - I've found 3 issues
Prompt for AI Agents
Please address the comments from this code review:
## Individual Comments
### Comment 1
<location path="test/om_test_base.c" line_range="138-136" />
<code_context>
}
END_TEST
+START_TEST(_RBT_FOREACH) {
+ RBT_ROOT(rbt);
+ om_rbt_node_t node = {.key = "B"};
+ int count = 0;
+
+ om_rbtree_insert(&rbt, &node);
+ om_rbtree_foreach(&rbt, _rbt_foreach_once, &count);
+
+ ck_assert_msg(count == 1, "单节点遍历应执行1次回调，实际执行%d次", count);
+}
+END_TEST
+
Suite* make_om_base_suite(void) {
</code_context>
<issue_to_address>
**suggestion (testing):** Add tests to cover multi-node traversals and verify full in-order behavior, not just the single-node case.
Given the traversal bug fixed in this PR, this test should exercise more than the single-node case. Please build a small multi-node tree (e.g., three nodes that force visiting both left and right branches) and assert that all nodes are visited and that the visit sequence is strictly in-order by key (for example, inserting keys "A", "B", "C" and collecting the visited keys in the callback to check the exact order).
</issue_to_address>
### Comment 2
<location path="test/om_test_base.c" line_range="144" />
<code_context>
+ int count = 0;
+
+ om_rbtree_insert(&rbt, &node);
+ om_rbtree_foreach(&rbt, _rbt_foreach_once, &count);
+
+ ck_assert_msg(count == 1, "单节点遍历应执行1次回调，实际执行%d次", count);
</code_context>
<issue_to_address>
**suggestion (testing):** Add a test that verifies early termination when the callback returns false.
Since the tests only use callbacks that return `true`, they don't verify the new short-circuit behavior. Please add a test where the callback returns `false` after visiting the first (or second) node, and assert both that traversal stops invoking the callback after that node (e.g., via a call counter or collected keys) and that the `false` result is propagated (or at least that no further nodes are visited).
Suggested implementation:
```c
static bool _rbt_foreach_early_stop(om_rbt_node_t* node, void* data) {
int* count = (int*)data;
(*count)++;
/* 在访问到第2个节点后中止遍历 */
if (*count >= 2) {
return false;
}
return true;
}
START_TEST(_RBT_FOREACH) {
```
```c
END_TEST
START_TEST(_RBT_FOREACH_EARLY_STOP) {
RBT_ROOT(rbt);
om_rbt_node_t node_a = {.key = "A"};
om_rbt_node_t node_b = {.key = "B"};
om_rbt_node_t node_c = {.key = "C"};
int count = 0;
om_rbtree_insert(&rbt, &node_b);
om_rbtree_insert(&rbt, &node_a);
om_rbtree_insert(&rbt, &node_c);
om_rbtree_foreach(&rbt, _rbt_foreach_early_stop, &count);
ck_assert_msg(count == 2, "回调返回false后应只访问前2个节点，实际访问%d个", count);
}
END_TEST
Suite* make_om_base_suite(void) {
```
```c
tcase_add_test(tc_rbt, _RBT);
tcase_add_test(tc_rbt, _RBT_FOREACH);
tcase_add_test(tc_rbt, _RBT_FOREACH_EARLY_STOP);
```
1. This patch assumes `bool` is already available (e.g., via `#include <stdbool.h>` in this file or a common header). If not, add that include near the top of `om_test_base.c`.
2. The exact callback signature for `om_rbtree_foreach` was inferred from `_rbt_foreach_once`; if it differs (e.g., const qualifiers or different parameter order), adjust `_rbt_foreach_early_stop` to match it.
3. If your traversal guarantees a different visit order or if the tree may contain fewer than 3 nodes in some configurations, you may want to relax the assertion to `ck_assert_int_le(count, 2)` and separately assert `count >= 1`.
</issue_to_address>
### Comment 3
<location path="test/om_test_base.c" line_range="139" />
<code_context>
END_TEST
+START_TEST(_RBT_FOREACH) {
+ RBT_ROOT(rbt);
+ om_rbt_node_t node = {.key = "B"};
+ int count = 0;
</code_context>
<issue_to_address>
**suggestion (testing):** Consider adding a dedicated test for traversing an empty tree.
Since the fix now returns `true` when `node == NULL`, please add a test that calls `om_rbtree_foreach` on an empty `RBT_ROOT` and verifies that the callback is never invoked (e.g., a counter stays at 0). This will protect the `node == NULL` base case from regressions.
Suggested implementation:
```c
START_TEST(_RBT_FOREACH) {
RBT_ROOT(rbt);
om_rbt_node_t node = {.key = "B"};
int count = 0;
om_rbtree_insert(&rbt, &node);
om_rbtree_foreach(&rbt, _rbt_foreach_once, &count);
ck_assert_msg(count == 1, "单节点遍历应执行1次回调，实际执行%d次", count);
}
END_TEST
START_TEST(_RBT_FOREACH_EMPTY) {
RBT_ROOT(rbt);
int count = 0;
om_rbtree_foreach(&rbt, _rbt_foreach_once, &count);
ck_assert_msg(count == 0, "空树遍历不应执行回调，实际执行%d次", count);
}
END_TEST
```
To actually run the new `_RBT_FOREACH_EMPTY` test, you will also need to register it in the suite definition (likely in `make_om_base_suite` or wherever other `START_TEST` cases are added with `tcase_add_test`). Add a corresponding `tcase_add_test(tc, _RBT_FOREACH_EMPTY);` next to the existing `_RBT_FOREACH` registration.
</issue_to_address>Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.
| @@ -127,6 +135,18 @@ START_TEST(_RBT) { | |||
| } | |||
| END_TEST | |||
There was a problem hiding this comment.
suggestion (testing): Add tests to cover multi-node traversals and verify full in-order behavior, not just the single-node case.
Given the traversal bug fixed in this PR, this test should exercise more than the single-node case. Please build a small multi-node tree (e.g., three nodes that force visiting both left and right branches) and assert that all nodes are visited and that the visit sequence is strictly in-order by key (for example, inserting keys "A", "B", "C" and collecting the visited keys in the callback to check the exact order).
| int count = 0; | ||
|
|
||
| om_rbtree_insert(&rbt, &node); | ||
| om_rbtree_foreach(&rbt, _rbt_foreach_once, &count); |
There was a problem hiding this comment.
suggestion (testing): Add a test that verifies early termination when the callback returns false.
Since the tests only use callbacks that return true, they don't verify the new short-circuit behavior. Please add a test where the callback returns false after visiting the first (or second) node, and assert both that traversal stops invoking the callback after that node (e.g., via a call counter or collected keys) and that the false result is propagated (or at least that no further nodes are visited).
Suggested implementation:
static bool _rbt_foreach_early_stop(om_rbt_node_t* node, void* data) {
int* count = (int*)data;
(*count)++;
/* 在访问到第2个节点后中止遍历 */
if (*count >= 2) {
return false;
}
return true;
}
START_TEST(_RBT_FOREACH) {END_TEST
START_TEST(_RBT_FOREACH_EARLY_STOP) {
RBT_ROOT(rbt);
om_rbt_node_t node_a = {.key = "A"};
om_rbt_node_t node_b = {.key = "B"};
om_rbt_node_t node_c = {.key = "C"};
int count = 0;
om_rbtree_insert(&rbt, &node_b);
om_rbtree_insert(&rbt, &node_a);
om_rbtree_insert(&rbt, &node_c);
om_rbtree_foreach(&rbt, _rbt_foreach_early_stop, &count);
ck_assert_msg(count == 2, "回调返回false后应只访问前2个节点，实际访问%d个", count);
}
END_TEST
Suite* make_om_base_suite(void) { tcase_add_test(tc_rbt, _RBT);
tcase_add_test(tc_rbt, _RBT_FOREACH);
tcase_add_test(tc_rbt, _RBT_FOREACH_EARLY_STOP);- This patch assumes
boolis already available (e.g., via#include <stdbool.h>in this file or a common header). If not, add that include near the top ofom_test_base.c. - The exact callback signature for
om_rbtree_foreachwas inferred from_rbt_foreach_once; if it differs (e.g., const qualifiers or different parameter order), adjust_rbt_foreach_early_stopto match it. - If your traversal guarantees a different visit order or if the tree may contain fewer than 3 nodes in some configurations, you may want to relax the assertion to
ck_assert_int_le(count, 2)and separately assertcount >= 1.
| END_TEST | ||
|
|
||
| START_TEST(_RBT_FOREACH) { | ||
| RBT_ROOT(rbt); |
There was a problem hiding this comment.
suggestion (testing): Consider adding a dedicated test for traversing an empty tree.
Since the fix now returns true when node == NULL, please add a test that calls om_rbtree_foreach on an empty RBT_ROOT and verifies that the callback is never invoked (e.g., a counter stays at 0). This will protect the node == NULL base case from regressions.
Suggested implementation:
START_TEST(_RBT_FOREACH) {
RBT_ROOT(rbt);
om_rbt_node_t node = {.key = "B"};
int count = 0;
om_rbtree_insert(&rbt, &node);
om_rbtree_foreach(&rbt, _rbt_foreach_once, &count);
ck_assert_msg(count == 1, "单节点遍历应执行1次回调，实际执行%d次", count);
}
END_TEST
START_TEST(_RBT_FOREACH_EMPTY) {
RBT_ROOT(rbt);
int count = 0;
om_rbtree_foreach(&rbt, _rbt_foreach_once, &count);
ck_assert_msg(count == 0, "空树遍历不应执行回调，实际执行%d次", count);
}
END_TESTTo actually run the new _RBT_FOREACH_EMPTY test, you will also need to register it in the suite definition (likely in make_om_base_suite or wherever other START_TEST cases are added with tcase_add_test). Add a corresponding tcase_add_test(tc, _RBT_FOREACH_EMPTY); next to the existing _RBT_FOREACH registration.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary by Sourcery
Fix red-black tree foreach traversal behavior and add a regression test for single-node iteration.
Bug Fixes:
Tests:
Fixes #8