filter: implement basic analysis of urls

Analysing of the url host part, some simple tests. Not usable as
squid filter yet.

1 files changed, 20 insertions, 0 deletions

diff --git a/blob.c b/blob.c
index 377ec62..a417a0b 100644
--- a/blob.c
+++ b/blob.c
@@ -180,3 +180,23 @@ blob_t blob_pull_cspn(blob_t *b, const blob_t reject)
 	*b = BLOB_NULL;
 	return t;
 }
+
+blob_t blob_expand_head(blob_t *b, blob_t limits, unsigned char sep)
+{
+	blob_t t = *b;
+	blob_t r;
+
+	if (t.ptr <= limits.ptr || t.ptr+t.len > limits.ptr+limits.len)
+		return BLOB_NULL;
+	while (t.ptr > limits.ptr && t.ptr[-1] == sep)
+		t.ptr--, t.len++;
+
+	r.ptr = t.ptr;
+	r.len = 0;
+	while (t.ptr > limits.ptr && t.ptr[-1] != sep) {
+		t.ptr--, t.len++;
+		r.ptr--, r.len++;
+	}
+	*b = t;
+	return r;
+}