From: Urban Wallasch <urban.wallasch@freenet.de>
Date: Wed, 13 Nov 2019 14:59:57 +0000 (+0100)
Subject: * Added mem_mem_bmh() using the Boyer-Moore-Horspool algorithm.
X-Git-Url: https://git.packet-gain.de/?a=commitdiff_plain;h=95722ce60212d5b06723fd3a42e235568a99ef5b;p=oddbits.git

* Added mem_mem_bmh() using the Boyer-Moore-Horspool algorithm.
---

diff --git a/str/str.h b/str/str.h
index 6270d2b..86d299f 100644
--- a/str/str.h
+++ b/str/str.h
@@ -27,14 +27,15 @@
  *
  * str_istr  - locate a substring ignoring case
  *
+ * mem_mem_bmh,
  * mem_mem,
  * mem_str,
  * mem_istr  - locate a pattern in memory
  *
  * str_delim - locate the first occurrence of a delimiter in a string
-
+ *
  * str_skip  - skip initial sequence of specified characters in a string
-
+ *
  * str_split - break down a string into separate parts
  *
  * str_ltrim,
@@ -164,24 +165,54 @@ static inline char *str_istr(const char *haystack, const char *needle) {
 
 
 /*
- * mem_mem, mem_str, mem_istr - locate a pattern in memory
+ * mem_mem_bmh, mem_mem, mem_str, mem_istr - locate a pattern in memory
  *
- * The mem_mem() function finds the first occurrence of the pattern
- * needle of length n in an haystack of size h.
+ * The mem_mem_bmh() and mem_mem() functions find the first occurrence
+ * of the pattern needle of length n in an haystack of size h.  The
+ * mem_mem_bmh() function uses the Boyer-Moore-Horspool search algorithm,
+ * see: https://en.wikipedia.org/wiki/Boyer–Moore–Horspool_algorithm
  *
  * The mem_str() function is similar, except that it expects a string
  * for needle, ignoring the terminating null character.  The mem_istr()
  * function additionally ignores the case of characters in both needle
  * and haystack.
  *
- * The mem_mem(), mem_str() and mem_istr() functions return a pointer
- * to the beginning of the located pattern, or NULL if the pattern is
- * not found.  If the search pattern is of zero length, the functions
- * return haystack.
+ * The mem_mem_bmh(), mem_mem(), mem_str() and mem_istr() functions
+ * return a pointer to the beginning of the located pattern, or NULL if
+ * the pattern was not found.  If the search pattern is of zero length,
+ * the functions return haystack.
  */
+static inline void *mem_mem_bmh(const void *haystack, size_t h,
+                                const void *needle, size_t n) {
+    size_t k, skip[256];
+    const uint8_t *hst = (const uint8_t *)haystack;
+    const uint8_t *ndl = (const uint8_t *)needle;
+
+    if (n == 0)
+        return (void *)haystack;
+    /* Set up the finite state machine we use. */
+    for (k = 0; k < 256; ++k)
+        skip[k] = n;
+    for (k = 0; k < n - 1; ++k)
+        skip[ndl[k]] = n - k - 1;
+    /* Do the search. */
+    for (k = n - 1; k < h; k += skip[hst[k]]) {
+        int i, j;
+        for (j = n - 1, i = k; j >= 0 && hst[i] == ndl[j]; j--)
+            i--;
+        if (j == -1)
+            return (void *)(hst + i + 1);
+    }
+    return NULL;
+}
+
 static inline void *mem_mem(const void *haystack, size_t h,
                             const void *needle, size_t n) {
     if (h >= n) {
+        /* Resort to the BMH algorithm under suitable conditions: */
+        if (n > 2 && n * (h - n) > 256)
+            return mem_mem_bmh(haystack, h, needle, n);
+        /* Perform naive search: */
         const char *p = haystack;
         h -= n;
         do {
diff --git a/str/str_test.c b/str/str_test.c
index b08b322..3520fd1 100644
--- a/str/str_test.c
+++ b/str/str_test.c
@@ -142,6 +142,18 @@ int main(void) {
         T(mem_mem(m1, sizeof m1, (char[]){-1}, 1) == m1 + 8);
         T(mem_mem(m1, sizeof m1, "xc", 3) == NULL);
         T(mem_mem(m1, sizeof m1, "xc", 2) == m1 + 6);
+        const char m3[] = "Lorem ipsum dolor sit amet, consectetur "
+                "adipiscing elit, sed do eiusmod tempor incididunt ut "
+                "labore et dolore magna aliqua. Ut enim ad minim veniam, "
+                "quis nostrud exercitation ullamco laboris nisi ut aliquip "
+                "ex ea commodo consequat. Duis aute irure dolor in "
+                "reprehenderit in voluptate velit esse cillum dolore eu "
+                "fugiat nulla pariatur. Excepteur sint occaecat cupidatat "
+                "non proident, sunt in culpa qui officia deserunt mollit "
+                "anim id est laborum.";
+        T(mem_mem_bmh(m3, strlen(m3), "", 0) == m3);
+        T(mem_mem_bmh(m3, strlen(m3), "FOOBAR", 6) == NULL);
+        T(mem_mem_bmh(m3, strlen(m3), "officia", 7) == m3 + 401);
     }
 
     /*****************************************/