Tarantool development patches archive
 help / color / mirror / Atom feed
From: Vladimir Davydov via Tarantool-patches <tarantool-patches@dev.tarantool.org>
To: Vladislav Shpilevoy <v.shpilevoy@tarantool.org>
Cc: tarantool-patches@dev.tarantool.org
Subject: Re: [Tarantool-patches] [PATCH] net.box: allow to store user-defined fields in future object
Date: Fri, 13 Aug 2021 13:01:46 +0300	[thread overview]
Message-ID: <20210813100146.s7fvpw47hmczfer5@esperanza> (raw)
In-Reply-To: <c3f8c9e2-f768-c003-a437-b23ba618e11c@tarantool.org>

On Thu, Aug 12, 2021 at 09:02:46PM +0300, Vladislav Shpilevoy wrote:
> > +
> > +static int
> > +luaT_netbox_request_newindex(struct lua_State *L)
> > +{
> > +	struct netbox_request *request = luaT_check_netbox_request(L, 1);
> > +	struct mh_strnptr_t *h = request->index;
> > +	size_t field_name_len;
> > +	const char *field_name = lua_tolstring(L, 2, &field_name_len);
> > +	if (field_name == NULL)
> > +		return luaL_error(L, "invalid index");
> > +	int field_value_ref = luaL_ref(L, LUA_REGISTRYINDEX);
> > +	if (field_value_ref == LUA_REFNIL) {
> > +		/* The field is set to nil. Delete it from the index. */
> 
> 1. But it is not deleted. It is just ignored, which is fine, but
> the comment is wrong then.
> 
> > +		if (h == NULL)
> > +			return 0;
> > +		mh_int_t k = mh_strnptr_find_inp(h, field_name,
> > +						 field_name_len);
> > +		int ref = (intptr_t)mh_strnptr_node(h, k)->val;
> > +		luaL_unref(L, LUA_REGISTRYINDEX, ref);
> > +		mh_strnptr_del(h, k, NULL);
> > +		return 0;
> > +	}
> > +	if (h == NULL) {
> > +		/* Lazily create the index on the first invocation. */
> > +		h = request->index = mh_strnptr_new();
> > +		if (h == NULL) {
> > +			luaL_unref(L, LUA_REGISTRYINDEX, field_value_ref);
> > +			return luaL_error(L, "out of memory");
> > +		}
> > +	}
> > +	/* Insert a reference to the new value into the index. */
> > +	struct mh_strnptr_node_t node = {
> > +		.str = field_name,
> > +		.len = field_name_len,
> > +		.hash = mh_strn_hash(field_name, field_name_len),
> 
> 2. lua_hashstring() might be faster. You could reuse already
> calculated hash value from Lua. Although for that 'field_name'
> must be of exactly string type (lua_tolstring() works not only for
> strings). So if you would make a more precise type check, you
> could avoid hash calculation completely. Or use lua_hash() for
> non-string values' strings. (But I might be wrong about how
> lua_hashstring() works for non-string values, you would need to
> double-check).
> 
> > +		.val = (void *)(intptr_t)field_value_ref,
> 
> 3. Can you simply cast to void * right away? Why do you need the
> intermediate intptr_t cast?

I reworked the patch to use a Lua table instead of mhash, as per
the suggestion by imun@, so these comments are not relevant anymore.
Please see v3:

https://lists.tarantool.org/pipermail/tarantool-patches/2021-August/025458.html

  parent reply	other threads:[~2021-08-13 10:01 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-08-10 16:53 Vladimir Davydov via Tarantool-patches
2021-08-10 17:15 ` Cyrill Gorcunov via Tarantool-patches
2021-08-11  7:49   ` Vladimir Davydov via Tarantool-patches
2021-08-12 17:07     ` [Tarantool-patches] [PATCH v2] " Vladimir Davydov via Tarantool-patches
2021-08-12 18:02 ` [Tarantool-patches] [PATCH] " Vladislav Shpilevoy via Tarantool-patches
2021-08-12 18:13   ` Vladislav Shpilevoy via Tarantool-patches
2021-08-13 10:01   ` Vladimir Davydov via Tarantool-patches [this message]
2021-08-12 18:13 ` Igor Munkin via Tarantool-patches
2021-08-13  9:57   ` Vladimir Davydov via Tarantool-patches
2021-08-13  9:59     ` [Tarantool-patches] [PATCH v3] " Vladimir Davydov via Tarantool-patches
2021-08-14 11:17       ` Igor Munkin via Tarantool-patches
2021-08-16  8:27         ` Vladimir Davydov via Tarantool-patches
2021-08-16  8:56           ` Igor Munkin via Tarantool-patches
2021-08-16 11:32             ` Vitaliia Ioffe via Tarantool-patches
2021-08-16 11:35               ` Vladimir Davydov via Tarantool-patches

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210813100146.s7fvpw47hmczfer5@esperanza \
    --to=tarantool-patches@dev.tarantool.org \
    --cc=v.shpilevoy@tarantool.org \
    --cc=vdavydov@tarantool.org \
    --subject='Re: [Tarantool-patches] [PATCH] net.box: allow to store user-defined fields in future object' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox