From eaa56ad32ae9736c02c7b824e4bb63b16d6487da Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Motiejus=20Jak=C5=A1tys?= <motiejus@jakstys.lt>
Date: Sat, 9 Jul 2022 18:15:04 +0300
Subject: [PATCH] use bdz_read.c

---
 build.zig       |  3 +-
 deps/bdz_read.c | 82 +++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 83 insertions(+), 2 deletions(-)
 create mode 100755 deps/bdz_read.c

diff --git a/build.zig b/build.zig
index 8458e57..a516078 100644
--- a/build.zig
+++ b/build.zig
@@ -48,8 +48,7 @@ pub fn build(b: *zbs.Builder) void {
     bdz.setBuildMode(mode);
     bdz.linkLibC();
     bdz.addCSourceFiles(&.{
-        "deps/cmph/src/bdz.c",
-        "deps/cmph/src/bdz_ph.c",
+        "deps/bdz_read.c",
         "deps/cmph/src/hash.c",
         "deps/cmph/src/jenkins_hash.c",
     }, &.{
diff --git a/deps/bdz_read.c b/deps/bdz_read.c
new file mode 100755
index 0000000..f132053
--- /dev/null
+++ b/deps/bdz_read.c
@@ -0,0 +1,82 @@
+/*
+ * this is a trimmed-down version of bdz.c with only "search".
+ */
+#include "bdz.h"
+#include "cmph_structs.h"
+#include "bdz_structs.h"
+#include "bitbool.h"
+
+#define UNASSIGNED 3U
+
+// table used for looking up the number of assigned vertices  a 8-bit integer
+const cmph_uint8 bdz_lookup_table[] =
+{
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 2, 2, 2, 1,
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 2, 2, 2, 1,
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+4, 4, 4, 3, 4, 4, 4, 3, 4, 4, 4, 3, 3, 3, 3, 2,
+3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 2, 2, 2, 1,
+3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 2, 2, 2, 1,
+3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 2, 2, 2, 1,
+3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 2, 2, 2, 1,
+2, 2, 2, 1, 2, 2, 2, 1, 2, 2, 2, 1, 1, 1, 1, 0
+};
+
+static inline cmph_uint32 rank(cmph_uint32 b, cmph_uint32 * ranktable, cmph_uint8 * g, cmph_uint32 vertex)
+{
+	register cmph_uint32 index = vertex >> b;
+	register cmph_uint32 base_rank = ranktable[index];
+	register cmph_uint32 beg_idx_v = index << b;
+	register cmph_uint32 beg_idx_b = beg_idx_v >> 2;
+	register cmph_uint32 end_idx_b = vertex >> 2;
+	while(beg_idx_b < end_idx_b)
+	{
+		base_rank += bdz_lookup_table[*(g + beg_idx_b++)];
+
+	}
+	beg_idx_v = beg_idx_b << 2;
+	while(beg_idx_v < vertex)
+	{
+		if(GETVALUE(g, beg_idx_v) != UNASSIGNED) base_rank++;
+		beg_idx_v++;
+	}
+
+	return base_rank;
+}
+
+/** cmph_uint32 bdz_search(void *packed_mphf, const char *key, cmph_uint32 keylen);
+ *  \brief Use the packed mphf to do a search.
+ *  \param  packed_mphf pointer to the packed mphf
+ *  \param key key to be hashed
+ *  \param keylen key legth in bytes
+ *  \return The mphf value
+ */
+cmph_uint32 bdz_search_packed(void *packed_mphf, const char *key, cmph_uint32 keylen)
+{
+
+	register cmph_uint32 vertex;
+	register CMPH_HASH hl_type  = (CMPH_HASH)(*(cmph_uint32 *)packed_mphf);
+	register cmph_uint8 *hl_ptr = (cmph_uint8 *)(packed_mphf) + 4;
+
+	register cmph_uint32 *ranktable = (cmph_uint32*)(hl_ptr + hash_state_packed_size(hl_type));
+
+	register cmph_uint32 r = *ranktable++;
+	register cmph_uint32 ranktablesize = *ranktable++;
+	register cmph_uint8 * g = (cmph_uint8 *)(ranktable + ranktablesize);
+	register cmph_uint8 b = *g++;
+
+	cmph_uint32 hl[3];
+	hash_vector_packed(hl_ptr, hl_type, key, keylen, hl);
+	hl[0] = hl[0] % r;
+	hl[1] = hl[1] % r + r;
+	hl[2] = hl[2] % r + (r << 1);
+	vertex = hl[(GETVALUE(g, hl[0]) + GETVALUE(g, hl[1]) + GETVALUE(g, hl[2])) % 3];
+	return rank(b, ranktable, g, vertex);
+}