enh: 代码结构

MtFmT-Lib · Jul 25, 2023 · 2641384 · 2641384
1 parent 4c18940
commit 2641384
Show file tree

Hide file tree

Showing 2 changed files with 41 additions and 76 deletions.
diff --git a/src/mm_heap.c b/src/mm_heap.c
@@ -312,6 +312,9 @@ static void* heap_re_allocate_impl(
  */
 static void heap_free_impl(Heap* heap, void* mem)
 {
+    if (mem == NULL) {
+        return;
+    }
     FreeBlock* block = (FreeBlock*)((uptr_t)(mem) - sizeof(FreeBlock));
     // 本内存块大小
     heap_size_t origin_sz = block->size;

diff --git a/src/mm_pattern.c b/src/mm_pattern.c
@@ -36,12 +36,10 @@
 static mstr_result_t mstr_find_simple(
     isize_t*, const char*, usize_t, usize_t, const char*, usize_t
 );
-static mstr_result_t patt_match_naive(
-    isize_t*, const mstr_char_t*, usize_t, const mstr_char_t*, usize_t
-);
-static mstr_result_t patt_match_large(
+static mstr_result_t patt_match_impl(
     isize_t*, const mstr_char_t*, usize_t, const mstr_char_t*, usize_t
 );
+static void make_jump_table(uint8_t*, const mstr_char_t*, usize_t);
 
 //
 // public:
@@ -159,54 +157,54 @@ static mstr_result_t mstr_find_simple(
         *result = 0;
         return MStr_Ok;
     }
-    if (pattern_cnt < BM_THRESHOLD_CNT) {
-        ret_code = patt_match_naive(
-            &find_res,
-            main_str + begin_pos,
-            main_str_cnt - begin_pos,
-            pattern,
-            pattern_cnt
-        );
-    }
-    else {
-        ret_code = patt_match_large(
-            &find_res,
-            main_str + begin_pos,
-            main_str_cnt - begin_pos,
-            pattern,
-            pattern_cnt
-        );
-    }
+    ret_code = patt_match_impl(
+        &find_res,
+        main_str + begin_pos,
+        main_str_cnt - begin_pos,
+        pattern,
+        pattern_cnt
+    );
     if (MSTR_SUCC(ret_code)) {
         *result = find_res;
     }
     return ret_code;
 }
 
 /**
- * @brief 暴力方式的子串匹配
+ * @brief 子串匹配
  *
  * @param[out] res: 匹配结果
  * @param[in] mstr: 主串
  * @param[in] mstr_cnt: 主串长度
  * @param[in] patt: 模式串
  * @param[in] patt_len: 模式串长度
  */
-static mstr_result_t patt_match_naive(
+static mstr_result_t patt_match_impl(
     isize_t* res,
     const mstr_char_t* mstr,
     usize_t mstr_cnt,
     const mstr_char_t* patt,
     usize_t patt_cnt
 )
 {
+    uint8_t* bad_char_arr = NULL;
     usize_t match_cnt = patt_cnt - 1;
     const mstr_char_t* m_b = mstr;
     const mstr_char_t* m_p = mstr;
     const mstr_char_t* m_end = mstr + mstr_cnt;
     mstr_assert(patt_cnt > 0);
     // 默认为没有找到
     *res = -1;
+    // 分配bad char数组, 并构造跳转表
+    if (patt_cnt >= BM_THRESHOLD_CNT) {
+        bad_char_arr = (uint8_t*)mstr_heap_alloc(
+            sizeof(*bad_char_arr) * BM_CHAR_INDEX_MAX
+        );
+        if (bad_char_arr == NULL) {
+            return MStr_Err_HeapTooSmall;
+        }
+        make_jump_table(bad_char_arr, patt, patt_cnt);
+    }
     // 匹配子串
     while (m_p < m_end - match_cnt) {
         isize_t i = (isize_t)match_cnt;
@@ -220,52 +218,39 @@ static mstr_result_t patt_match_naive(
             break;
         }
         else {
-            // 匹配失败, 跳过1个字符的距离
-            m_p += 1;
+            // substring的第i个字符匹配失败, m_p + i处失败
+            // 移动next_pos到匹配失败的字符最后一次在sub
+            // string中出现的位置处
+            usize_t arr_idx = (usize_t)(uint8_t)m_p[i];
+            mstr_assert(arr_idx >= 0 && arr_idx <= BM_CHAR_INDEX_MAX);
+            m_p += bad_char_arr ? (usize_t)bad_char_arr[arr_idx] : 1;
         }
     }
+    if (bad_char_arr != NULL) {
+        mstr_heap_free(bad_char_arr);
+    }
     return MStr_Ok;
 }
 
 /**
- * @brief 基于Boyer-Moore实现的字符串匹配
+ * @brief 按照模式串 patt 构造跳转表
  *
- * @param[out] res: 匹配结果
- * @param[in] mstr: 主串
- * @param[in] mstr_len: 主串长度
+ * @param[out] table: 跳转表
  * @param[in] patt: 模式串
- * @param[in] patt_cnt: 模式串长度
+ * @param[in] patt_cnt: 模式串的字符数
  */
-static mstr_result_t patt_match_large(
-    isize_t* res,
-    const mstr_char_t* mstr,
-    usize_t mstr_cnt,
-    const mstr_char_t* patt,
-    usize_t patt_cnt
+static void make_jump_table(
+    uint8_t* table, const mstr_char_t* patt, usize_t patt_cnt
 )
 {
-    uint8_t* bad_char_arr;
-    usize_t match_cnt = patt_cnt - 1;
-    const mstr_char_t* m_b = mstr;
-    const mstr_char_t* m_p = mstr;
-    const mstr_char_t* m_end = mstr + mstr_cnt;
     const usize_t BAD_CHAR_MAX_OFFSET = 255;
     mstr_assert(patt_cnt > 0);
-    // 默认为没有找到
-    *res = -1;
-    // 分配bad char数组
-    bad_char_arr = (uint8_t*)mstr_heap_alloc(
-        sizeof(*bad_char_arr) * BM_CHAR_INDEX_MAX
-    );
-    if (bad_char_arr == NULL) {
-        return MStr_Err_HeapTooSmall;
-    }
     // 默认情况, 没有出现在sub string的字符移动substring.length长度
     for (usize_t i = 0; i < BM_CHAR_INDEX_MAX; i += 1) {
         usize_t offset = patt_cnt > BAD_CHAR_MAX_OFFSET ?
                              BAD_CHAR_MAX_OFFSET :
                              patt_cnt;
-        bad_char_arr[i] = (uint8_t)offset;
+        table[i] = (uint8_t)offset;
     }
     // 不然, bad_char[ch]就是ch在sub string中最后一次出现的位置
     for (usize_t i = 0; i < patt_cnt - 1; i += 1) {
@@ -274,29 +259,6 @@ static mstr_result_t patt_match_large(
             offset > BAD_CHAR_MAX_OFFSET ? BAD_CHAR_MAX_OFFSET : offset;
         usize_t ch_idx = (usize_t)(uint8_t)patt[i];
         mstr_assert(ch_idx >= 0 && ch_idx <= BM_CHAR_INDEX_MAX);
-        bad_char_arr[ch_idx] = (uint8_t)li_off;
-    }
-    // 匹配子串
-    while (m_p < m_end - match_cnt) {
-        isize_t i = (isize_t)match_cnt;
-        // 倒着来匹配
-        while (i >= 0 && patt[i] == m_p[i]) {
-            i -= 1;
-        }
-        if (i < 0) {
-            // 匹配成功
-            *res = (isize_t)(m_p - m_b);
-            break;
-        }
-        else {
-            // substring的第i个字符匹配失败, m_p + i处失败
-            // 移动next_pos到匹配失败的字符最后一次在sub
-            // string中出现的位置处
-            usize_t arr_idx = (usize_t)(uint8_t)m_p[i];
-            mstr_assert(arr_idx >= 0 && arr_idx <= BM_CHAR_INDEX_MAX);
-            m_p += (usize_t)bad_char_arr[arr_idx];
-        }
+        table[ch_idx] = (uint8_t)li_off;
     }
-    mstr_heap_free(bad_char_arr);
-    return MStr_Ok;
 }