Make serd_uri_string_length() precise and add it to public API

3 files changed, 61 insertions, 23 deletions

diff --git a/include/serd/serd.h b/include/serd/serd.h
index dd165360..af087ab5 100644
--- a/include/serd/serd.h
+++ b/include/serd/serd.h
@@ -577,6 +577,18 @@ bool
 serd_uri_is_within(SerdURIView r, SerdURIView base);
 
 /**
+   Return the length of `uri` as a string.
+
+   This can be used to get the expected number of bytes that will be written by
+   serd_write_uri().
+
+   @return A string length in bytes, not including the null terminator.
+*/
+SERD_API
+size_t
+serd_uri_string_length(SerdURIView uri);
+
+/**
    Write `uri` as a string to `sink`.
 
    This will call `sink` several times to emit the serialised URI.
@@ -584,7 +596,9 @@ serd_uri_is_within(SerdURIView r, SerdURIView base);
    @param uri URI to write as a string.
    @param sink Sink to write string output to.
    @param stream Opaque user argument to pass to `sink`.
-   @return The number of bytes written.
+
+   @return The number of bytes written, which is less than
+   `serd_uri_string_length(uri)` on error.
 */
 SERD_API
 size_t
diff --git a/src/node.c b/src/node.c
index 76c12235..0448c312 100644
--- a/src/node.c
+++ b/src/node.c
@@ -95,25 +95,6 @@ serd_node_check_padding(const SerdNode* node)
 }
 
 static size_t
-serd_uri_string_length(const SerdURIView* uri)
-{
-  size_t len = uri->path_prefix.len;
-
-#define ADD_LEN(field, n_delims)     \
-  if ((field).len) {                 \
-    len += (field).len + (n_delims); \
-  }
-
-  ADD_LEN(uri->path, 1)      // + possible leading `/'
-  ADD_LEN(uri->scheme, 1)    // + trailing `:'
-  ADD_LEN(uri->authority, 2) // + leading `//'
-  ADD_LEN(uri->query, 1)     // + leading `?'
-  ADD_LEN(uri->fragment, 1)  // + leading `#'
-
-  return len + 2; // + 2 for authority `//'
-}
-
-static size_t
 string_sink(const void* buf, size_t size, size_t nmemb, void* stream)
 {
   char** ptr = (char**)stream;
@@ -280,7 +261,7 @@ serd_new_typed_literal_uri(const SerdStringView str,
                            const SerdNodeFlags  flags,
                            const SerdURIView    datatype_uri)
 {
-  const size_t datatype_uri_len = serd_uri_string_length(&datatype_uri);
+  const size_t datatype_uri_len = serd_uri_string_length(datatype_uri);
   const size_t len              = serd_node_pad_size(str.len);
   const size_t total_len        = len + sizeof(SerdNode) + datatype_uri_len;
 
@@ -294,6 +275,7 @@ serd_new_typed_literal_uri(const SerdStringView str,
   char*           ptr           = serd_node_buffer(datatype_node);
 
   const size_t actual_len = serd_write_uri(datatype_uri, string_sink, &ptr);
+  assert(actual_len == datatype_uri_len);
 
   serd_node_buffer(datatype_node)[actual_len] = '\0';
   datatype_node->n_bytes                      = actual_len;
@@ -462,11 +444,13 @@ serd_new_uri(const SerdStringView str)
 SerdNode*
 serd_new_parsed_uri(const SerdURIView uri)
 {
-  const size_t    len        = serd_uri_string_length(&uri);
+  const size_t    len        = serd_uri_string_length(uri);
   SerdNode* const node       = serd_node_malloc(len, 0, SERD_URI);
   char*           ptr        = serd_node_buffer(node);
   const size_t    actual_len = serd_write_uri(uri, string_sink, &ptr);
 
+  assert(actual_len == len);
+
   serd_node_buffer(node)[actual_len] = '\0';
   node->n_bytes                      = actual_len;
 
@@ -478,11 +462,13 @@ static SerdNode*
 serd_new_from_uri(const SerdURIView uri, const SerdURIView base)
 {
   const SerdURIView abs_uri    = serd_resolve_uri(uri, base);
-  const size_t      len        = serd_uri_string_length(&abs_uri);
+  const size_t      len        = serd_uri_string_length(abs_uri);
   SerdNode*         node       = serd_node_malloc(len, 0, SERD_URI);
   char*             ptr        = serd_node_buffer(node);
   const size_t      actual_len = serd_write_uri(abs_uri, string_sink, &ptr);
 
+  assert(actual_len == len);
+
   serd_node_buffer(node)[actual_len] = '\0';
   node->n_bytes                      = actual_len;
 
diff --git a/src/uri.c b/src/uri.c
index 6b029284..5b1ed598 100644
--- a/src/uri.c
+++ b/src/uri.c
@@ -410,6 +410,44 @@ serd_uri_is_within(const SerdURIView uri, const SerdURIView base)
   return true;
 }
 
+size_t
+serd_uri_string_length(const SerdURIView uri)
+{
+  size_t len = 0;
+
+  if (uri.scheme.buf) {
+    len += uri.scheme.len + 1;
+  }
+
+  if (uri.authority.buf) {
+    const bool needs_extra_slash =
+      (uri.authority.len > 0 && uri_path_len(&uri) > 0 &&
+       uri_path_at(&uri, 0) != '/');
+
+    len += 2 + uri.authority.len + needs_extra_slash;
+  }
+
+  if (uri.path_prefix.buf) {
+    len += uri.path_prefix.len;
+  } else if (uri.path_prefix.len) {
+    len += 3 * uri.path_prefix.len;
+  }
+
+  if (uri.path.buf) {
+    len += uri.path.len;
+  }
+
+  if (uri.query.buf) {
+    len += uri.query.len + 1;
+  }
+
+  if (uri.fragment.buf) {
+    len += uri.fragment.len;
+  }
+
+  return len;
+}
+
 /// See http://tools.ietf.org/html/rfc3986#section-5.3
 size_t
 serd_write_uri(const SerdURIView uri, SerdWriteFunc sink, void* stream)