From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Wed, 3 Apr 2019 14:10:03 +0300 From: Vladimir Davydov Subject: Re: [tarantool-patches] Re: [PATCH v2 2/2] lua: add key_def lua module Message-ID: <20190403111003.x7vq7olda55tthgi@esperanza> References: <20190328020146.lluz4mg5tacpghwv@tkn_work_nb> <35ed4661-9789-7cf1-6627-2ced2a821939@tarantool.org> <6d915212-e80f-4a6d-d884-b838bf25f8a7@tarantool.org> <20190328112158.kpxsk6b55noicbes@tkn_work_nb> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20190328112158.kpxsk6b55noicbes@tkn_work_nb> To: Alexander Turenko , Kirill Shcherbatov Cc: tarantool-patches@freelists.org List-ID: On Thu, Mar 28, 2019 at 02:22:01PM +0300, Alexander Turenko wrote: > Changes I made (squashed and force-pushed): > > * Updated a comment (see below). > * Removed #4025 from commit messages as duplicate of #3398. > * Fixed the typo: `sk:pairs({1}))` -> `sk:pairs({1})` > > Vladimir, can you, please, look into the patchset? Patch 1 looks good to me. Pasting patch 2 here for review: > From 0ba422c6f9a4daa4716275ebb256a9e76e49aea1 Mon Sep 17 00:00:00 2001 > From: Kirill Shcherbatov > Date: Mon, 7 Jan 2019 19:12:50 +0300 > Subject: [PATCH] lua: add key_def lua module > > There are several reasons to add this module: > > * Factor out key parts parsing code from the tuples merger (#3276). > * Support comparing a tuple with a key / a tuple, support merging > key_defs from Lua (#3398). > * Support extracting a key from a tuple (#4025). > > The format of `parts` parameter in the `key_def.new(parts)` call is > compatible with the following structures: > > * box.space[...].index[...].parts; > * net_box_conn.space[...].index[...].parts. > > A key_def instance has the following methods: > > * :extract_key(tuple) -> key (as tuple) > * :compare(tuple_a, tuple_b) -> number > * :compare_with_key(tuple, key) -> number What number do these functions return? > * :merge(another_key_def) -> new key_def instance What does 'merge' do? > * :totable() -> table Does this function return key_def parts? In what format? Please elaborate the comments. Also, key_def.compare() sounds like it compares key definitions, not tuples. May be, we should move these functions to box.tuple module? Also, returning 1, 0, -1 to Lua looks uncommon. May be, we'd better introduce 'equal', 'greater', 'less', etc helpers returning bool? I'm not strongly against the proposed API, but I think we should agree on it with other members of the team, potential users, and Kostja. Also, how is this module supposed to handle multikey indexes? > > Note functions that accept tuple(s) also allow to pass Lua > table(s) instead. > > Needed for #3276. > Fixes #3398. > Fixes #4025. > > @TarantoolBot document > Title: lua: key_def module > > See the commit message for an API reference. Rather than referring to the commit message, please move @TarantoolBot request above. > > Example for extract_key(): > > ```lua > -- Remove values got from a secondary non-unique index. > local key_def_lib = require('key_def') > local s = box.schema.space.create('test') > local pk = s:create_index('pk') > local sk = s:create_index('test', {unique = false, parts = { > {2, 'number', path = 'a'}, {2, 'number', path = 'b'}}}) > s:insert{1, {a = 1, b = 1}} > s:insert{2, {a = 1, b = 2}} > local key_def = key_def_lib.new(pk.parts) > for _, tuple in sk:pairs({1})) do > local key = key_def:extract_key(tuple) > pk:delete(key) > end > ``` > > diff --git a/src/box/lua/key_def.c b/src/box/lua/key_def.c > new file mode 100644 > index 00000000..4df6c204 > --- /dev/null > +++ b/src/box/lua/key_def.c > +/** > + * Set key_part_def from a table on top of a Lua stack. > + * > + * When successful return 0, otherwise return -1 and set a diag. > + */ > +static int > +luaT_key_def_set_part(struct lua_State *L, struct key_part_def *parts, > + int part_idx, struct region *region) What is the region needed for? Please mention in the comment. > +{ > + struct key_part_def *part = &parts[part_idx]; > + *part = key_part_def_default; > + > + /* Set part->fieldno. */ > + lua_pushstring(L, "fieldno"); > + lua_gettable(L, -2); > + if (lua_isnil(L, -1)) { > + diag_set(IllegalParams, "fieldno must not be nil"); > + return -1; > + } > + /* > + * Transform one-based Lua fieldno to zero-based > + * fieldno to use in key_def_new(). > + */ > + part->fieldno = lua_tointeger(L, -1) - TUPLE_INDEX_BASE; > + lua_pop(L, 1); > + > + /* Set part->type. */ > + lua_pushstring(L, "type"); > + lua_gettable(L, -2); > + if (lua_isnil(L, -1)) { > + diag_set(IllegalParams, "type must not be nil"); > + return -1; > + } > + size_t type_len; > + const char *type_name = lua_tolstring(L, -1, &type_len); > + lua_pop(L, 1); > + part->type = field_type_by_name(type_name, type_len); > + switch (part->type) { > + case FIELD_TYPE_ANY: > + case FIELD_TYPE_ARRAY: > + case FIELD_TYPE_MAP: > + /* Tuple comparators don't support these types. */ > + diag_set(IllegalParams, "Unsupported field type: %s", > + type_name); > + return -1; > + case field_type_MAX: > + diag_set(IllegalParams, "Unknown field type: %s", type_name); > + return -1; > + default: > + /* Pass though. */ > + break; > + } > + > + /* Set part->is_nullable and part->nullable_action. */ > + lua_pushstring(L, "is_nullable"); > + lua_gettable(L, -2); > + if (!lua_isnil(L, -1) && lua_toboolean(L, -1) != 0) { > + part->is_nullable = true; > + part->nullable_action = ON_CONFLICT_ACTION_NONE; > + } > + lua_pop(L, 1); > + > + /* > + * Set part->coll_id using collation_id. > + * > + * The value will be checked in key_def_new(). > + */ > + lua_pushstring(L, "collation_id"); > + lua_gettable(L, -2); > + if (!lua_isnil(L, -1)) > + part->coll_id = lua_tointeger(L, -1); > + lua_pop(L, 1); > + > + /* Set part->coll_id using collation. */ > + lua_pushstring(L, "collation"); > + lua_gettable(L, -2); > + if (!lua_isnil(L, -1)) { > + /* Check for conflicting options. */ > + if (part->coll_id != COLL_NONE) { > + diag_set(IllegalParams, "Conflicting options: " > + "collation_id and collation"); > + return -1; > + } > + > + size_t coll_name_len; > + const char *coll_name = lua_tolstring(L, -1, &coll_name_len); > + struct coll_id *coll_id = coll_by_name(coll_name, > + coll_name_len); > + if (coll_id == NULL) { > + diag_set(IllegalParams, "Unknown collation: \"%s\"", > + coll_name); > + return -1; > + } > + part->coll_id = coll_id->id; > + } > + lua_pop(L, 1); > + > + /* Set part->path (JSON path). */ > + lua_pushstring(L, "path"); > + lua_gettable(L, -2); > + if (!lua_isnil(L, -1)) { > + size_t path_len; > + const char *path = lua_tolstring(L, -1, &path_len); > + if (json_path_validate(path, path_len, TUPLE_INDEX_BASE) != 0) { > + diag_set(ClientError, ER_WRONG_INDEX_OPTIONS, > + part_idx + TUPLE_INDEX_BASE, "invalid path"); Mixing ER_WRONG_INDEX_OPTIONS and IllegalParams looks ugly. Please fix. > + return -1; > + } > + char *tmp = region_alloc(region, path_len + 1); > + if (tmp == NULL) { > + diag_set(OutOfMemory, path_len + 1, "region", "path"); > + return -1; > + } > + /* > + * lua_tolstring() guarantees that a string have > + * trailing '\0'. > + */ > + memcpy(tmp, path, path_len + 1); > + part->path = tmp; > + } else { > + part->path = NULL; > + } > + lua_pop(L, 1); > + return 0; > +} > + > +void > +lbox_push_key_part(struct lua_State *L, const struct key_part *part) > +{ > + lua_newtable(L); > + > + lua_pushstring(L, field_type_strs[part->type]); > + lua_setfield(L, -2, "type"); > + > + lua_pushnumber(L, part->fieldno + TUPLE_INDEX_BASE); > + lua_setfield(L, -2, "fieldno"); > + > + if (part->path != NULL) { > + lua_pushlstring(L, part->path, part->path_len); > + lua_setfield(L, -2, "path"); > + } > + > + lua_pushboolean(L, key_part_is_nullable(part)); > + lua_setfield(L, -2, "is_nullable"); > + > + if (part->coll_id != COLL_NONE) { > + struct coll_id *coll_id = coll_by_id(part->coll_id); > + assert(coll_id != NULL); > + lua_pushstring(L, coll_id->name); > + lua_setfield(L, -2, "collation"); > + } > +} > + > +struct key_def * > +check_key_def(struct lua_State *L, int idx) Please prefix the name with lbox_ or... I dunno - the naming looks inconsistent: luaT_key_def_set_part, lbox_push_key_part, check_key_def. Is there some kind of pattern? > +{ > + if (lua_type(L, idx) != LUA_TCDATA) > + return NULL; > + > + uint32_t cdata_type; > + struct key_def **key_def_ptr = luaL_checkcdata(L, idx, &cdata_type); > + if (key_def_ptr == NULL || cdata_type != key_def_type_id) > + return NULL; > + return *key_def_ptr; > +} > + > +/** > + * Free a key_def from a Lua code. > + */ > +static int > +lbox_key_def_gc(struct lua_State *L) > +{ > + struct key_def *key_def = check_key_def(L, 1); > + if (key_def == NULL) > + return 0; > + box_key_def_delete(key_def); > + return 0; > +} > + > +/** > + * Take existent tuple from LUA stack or build a new tuple with > + * default format from table, check for compatibility with a > + * given key_def. Take tuple reference pointer on success. > + */ > +static struct tuple * > +lbox_key_def_check_tuple(struct lua_State *L, struct key_def *key_def, int idx) > +{ > + struct tuple *tuple = luaT_istuple(L, idx); > + if (tuple == NULL) > + tuple = luaT_tuple_new(L, idx, box_tuple_format_default()); > + if (tuple == NULL) > + return NULL; > + /* Check that tuple match with the key definition. */ > + uint32_t min_field_count = > + tuple_format_min_field_count(&key_def, 1, NULL, 0); > + uint32_t field_count = tuple_field_count(tuple); > + if (field_count < min_field_count) { > + diag_set(ClientError, ER_NO_SUCH_FIELD_NO, field_count + 1); > + return NULL; > + } > + for (uint32_t idx = 0; idx < key_def->part_count; idx++) { > + struct key_part *part = &key_def->parts[idx]; > + const char *field = tuple_field_by_part(tuple, part); > + if (field == NULL) { > + assert(key_def->has_optional_parts); > + continue; > + } > + if (key_part_validate(part->type, field, idx, > + key_part_is_nullable(part)) != 0) > + return NULL; > + } > + tuple_ref(tuple); > + return tuple; > +} The code checking a tuple against key_def should live somewhere in src/box - chances are high that we miss lua/key_def.c when we extend key_def struct again. Anyway, this check is incorrect: tarantool> s = box.schema.space.create('test') --- ... tarantool> i = box.space.test:create_index('pk', {parts = {{1, 'unsigned', path = 'a'}, {1, 'unsigned', path = 'b'} }}) --- ... tarantool> k = require('key_def').new(i.parts) --- ... tarantool> k:extract_key({1, 2, 3}) tarantool: /home/vlad/src/tarantool/src/box/lua/key_def.c:246: lbox_key_def_check_tuple: Assertion `key_def->has_optional_parts' failed. Please fix it and don't forget to add a test case. Probably, we should reuse tuple_validate() for checking a tuple against a key_def so as not to implement the same code again. > + > +static int > +lbox_key_def_extract_key(struct lua_State *L) > +{ > + struct key_def *key_def; > + if (lua_gettop(L) != 2 || (key_def = check_key_def(L, 1)) == NULL) > + return luaL_error(L, "Usage: key_def:extract_key(tuple)"); > + > + struct tuple *tuple; > + if ((tuple = lbox_key_def_check_tuple(L, key_def, 2)) == NULL) > + return luaT_error(L); > + > + uint32_t key_size; > + char *key = tuple_extract_key(tuple, key_def, &key_size); > + tuple_unref(tuple); > + if (key == NULL) > + return luaT_error(L); > + > + struct tuple *ret = > + box_tuple_new(box_tuple_format_default(), key, key + key_size); > + if (ret == NULL) > + return luaT_error(L); > + luaT_pushtuple(L, ret); > + return 1; > +} > + > +static int > +lbox_key_def_compare(struct lua_State *L) > +{ > + struct key_def *key_def; > + if (lua_gettop(L) != 3 || (key_def = check_key_def(L, 1)) == NULL) { > + return luaL_error(L, "Usage: key_def:" > + "compare(tuple_a, tuple_b)"); > + } > + > + struct tuple *tuple_a, *tuple_b; > + if ((tuple_a = lbox_key_def_check_tuple(L, key_def, 2)) == NULL) > + return luaT_error(L); > + if ((tuple_b = lbox_key_def_check_tuple(L, key_def, 3)) == NULL) { > + tuple_unref(tuple_a); > + return luaT_error(L); > + } > + > + int rc = tuple_compare(tuple_a, tuple_b, key_def); > + tuple_unref(tuple_a); > + tuple_unref(tuple_b); > + lua_pushinteger(L, rc); > + return 1; > +} > + > +static int > +lbox_key_def_compare_with_key(struct lua_State *L) > +{ > + struct key_def *key_def; > + if (lua_gettop(L) != 3 || (key_def = check_key_def(L, 1)) == NULL) { > + return luaL_error(L, "Usage: key_def:" > + "compare_with_key(tuple, key)"); > + } > + > + struct tuple *tuple, *key_tuple = NULL; > + struct tuple_format *format = box_tuple_format_default(); > + if ((tuple = lbox_key_def_check_tuple(L, key_def, 2)) == NULL) > + return luaT_error(L); > + if ((key_tuple = luaT_tuple_new(L, 3, format)) == NULL) { > + tuple_unref(tuple); > + return luaT_error(L); > + } > + tuple_ref(key_tuple); > + > + const char *key = tuple_data(key_tuple); Creating a tuple even though we just need msgpack doesn't look good to me. Can we encode msgpack without creating a tuple? How does index.get handle it? > + assert(mp_typeof(*key) == MP_ARRAY); > + uint32_t part_count = mp_decode_array(&key); > + if (key_validate_parts(key_def, key, part_count, true) != 0) { > + tuple_unref(tuple); > + tuple_unref(key_tuple); > + return luaT_error(L); > + } > + > + int rc = tuple_compare_with_key(tuple, key, part_count, key_def); > + tuple_unref(tuple); > + tuple_unref(key_tuple); > + lua_pushinteger(L, rc); > + return 1; > +} > + > +static int > +lbox_key_def_merge(struct lua_State *L) > +{ > + struct key_def *key_def_a, *key_def_b; > + if (lua_gettop(L) != 2 || (key_def_a = check_key_def(L, 1)) == NULL || > + (key_def_b = check_key_def(L, 2)) == NULL) > + return luaL_error(L, "Usage: key_def:merge(second_key_def)"); > + > + struct key_def *new_key_def = key_def_merge(key_def_a, key_def_b); > + if (new_key_def == NULL) > + return luaT_error(L); > + > + *(struct key_def **) luaL_pushcdata(L, key_def_type_id) = new_key_def; > + lua_pushcfunction(L, lbox_key_def_gc); > + luaL_setcdatagc(L, -2); > + return 1; > +} > + > +static int > +lbox_key_def_to_table(struct lua_State *L) > +{ > + struct key_def *key_def; > + if (lua_gettop(L) != 1 || (key_def = check_key_def(L, 1)) == NULL) > + return luaL_error(L, "Usage: key_def:totable()"); > + > + lua_createtable(L, key_def->part_count, 0); > + for (uint32_t i = 0; i < key_def->part_count; ++i) { > + lbox_push_key_part(L, &key_def->parts[i]); > + lua_rawseti(L, -2, i + 1); > + } > + return 1; > +} > + > +/** > + * Create a new key_def from a Lua table. > + * > + * Expected a table of key parts on the Lua stack. The format is > + * the same as box.space.<...>.index.<...>.parts or corresponding > + * net.box's one. > + * > + * Push the new key_def as cdata to a Lua stack. > + */ > +static int > +lbox_key_def_new(struct lua_State *L) > +{ > + if (lua_gettop(L) != 1 || lua_istable(L, 1) != 1) > + return luaL_error(L, "Bad params, use: key_def.new({" > + "{fieldno = fieldno, type = type" > + "[, is_nullable = ]" > + "[, path = ]" > + "[, collation_id = ]" > + "[, collation = ]}, ...}"); > + > + uint32_t part_count = lua_objlen(L, 1); > + const ssize_t parts_size = sizeof(struct key_part_def) * part_count; > + > + struct region *region = &fiber()->gc; > + size_t region_svp = region_used(region); > + struct key_part_def *parts = region_alloc(region, parts_size); > + if (parts == NULL) { > + diag_set(OutOfMemory, parts_size, "region", "parts"); > + return luaT_error(L); > + } > + > + for (uint32_t i = 0; i < part_count; ++i) { > + lua_pushinteger(L, i + 1); > + lua_gettable(L, 1); > + if (luaT_key_def_set_part(L, parts, i, region) != 0) { > + region_truncate(region, region_svp); > + return luaT_error(L); > + } > + } > + > + struct key_def *key_def = key_def_new(parts, part_count); > + region_truncate(region, region_svp); > + if (key_def == NULL) > + return luaT_error(L); > + > + /* > + * Calculate minimal field count of tuples with specified > + * key and update key_def optionality to use correct > + * compare/extract functions. > + */ > + uint32_t min_field_count = > + tuple_format_min_field_count(&key_def, 1, NULL, 0); > + key_def_update_optionality(key_def, min_field_count); > + > + *(struct key_def **) luaL_pushcdata(L, key_def_type_id) = key_def; > + lua_pushcfunction(L, lbox_key_def_gc); > + luaL_setcdatagc(L, -2); > + > + return 1; > +} > + > +LUA_API int > +luaopen_key_def(struct lua_State *L) > +{ > + luaL_cdef(L, "struct key_def;"); > + key_def_type_id = luaL_ctypeid(L, "struct key_def&"); > + > + /* Export C functions to Lua. */ > + static const struct luaL_Reg meta[] = { > + {"new", lbox_key_def_new}, > + {NULL, NULL} > + }; > + luaL_register_module(L, "key_def", meta); > + > + lua_newtable(L); /* key_def.internal */ > + lua_pushcfunction(L, lbox_key_def_extract_key); > + lua_setfield(L, -2, "extract_key"); > + lua_pushcfunction(L, lbox_key_def_compare); > + lua_setfield(L, -2, "compare"); > + lua_pushcfunction(L, lbox_key_def_compare_with_key); > + lua_setfield(L, -2, "compare_with_key"); > + lua_pushcfunction(L, lbox_key_def_merge); > + lua_setfield(L, -2, "merge"); > + lua_pushcfunction(L, lbox_key_def_to_table); > + lua_setfield(L, -2, "totable"); > + lua_setfield(L, -2, "internal"); Why 'internal'? We use them as is as key_def methods. E.g. box.tuple.* methods aren't internal. > + > + return 1; > +} > diff --git a/src/box/lua/space.cc b/src/box/lua/space.cc > index ca793e42..6eecbc75 100644 > --- a/src/box/lua/space.cc > +++ b/src/box/lua/space.cc > @@ -30,6 +30,7 @@ > */ > #include "box/lua/space.h" > #include "box/lua/tuple.h" > +#include "box/lua/key_def.h" > #include "box/sql/sqlLimit.h" > #include "lua/utils.h" > #include "lua/trigger.h" > @@ -286,32 +287,7 @@ lbox_fillspace(struct lua_State *L, struct space *space, int i) > > for (uint32_t j = 0; j < index_def->key_def->part_count; j++) { > lua_pushnumber(L, j + 1); > - lua_newtable(L); > - const struct key_part *part = > - &index_def->key_def->parts[j]; > - > - lua_pushstring(L, field_type_strs[part->type]); > - lua_setfield(L, -2, "type"); > - > - lua_pushnumber(L, part->fieldno + TUPLE_INDEX_BASE); > - lua_setfield(L, -2, "fieldno"); > - > - if (part->path != NULL) { > - lua_pushlstring(L, part->path, part->path_len); > - lua_setfield(L, -2, "path"); > - } > - > - lua_pushboolean(L, key_part_is_nullable(part)); > - lua_setfield(L, -2, "is_nullable"); > - > - if (part->coll_id != COLL_NONE) { > - struct coll_id *coll_id = > - coll_by_id(part->coll_id); > - assert(coll_id != NULL); > - lua_pushstring(L, coll_id->name); > - lua_setfield(L, -2, "collation"); > - } > - > + lbox_push_key_part(L, &index_def->key_def->parts[j]); > lua_settable(L, -3); /* index[k].parts[j] */ Why not factor out lbox_push_key_def instead? > } > > diff --git a/test/box-tap/key_def.test.lua b/test/box-tap/key_def.test.lua > new file mode 100755 > index 00000000..1270f82e > --- /dev/null > +++ b/test/box-tap/key_def.test.lua > @@ -0,0 +1,370 @@ > +#!/usr/bin/env tarantool > + > +local tap = require('tap') > +local ffi = require('ffi') > +local json = require('json') > +local fun = require('fun') > +local key_def_lib = require('key_def') > + > +local usage_error = 'Bad params, use: key_def.new({' .. > + '{fieldno = fieldno, type = type' .. > + '[, is_nullable = ]' .. > + '[, path = ]' .. > + '[, collation_id = ]' .. > + '[, collation = ]}, ...}' > + > +local function coll_not_found(fieldno, collation) > + if type(collation) == 'number' then > + return ('Wrong index options (field %d): ' .. > + 'collation was not found by ID'):format(fieldno) > + end > + > + return ('Unknown collation: "%s"'):format(collation) > +end > + > +local function set_key_part_defaults(parts) > + local res = {} > + for i, part in ipairs(parts) do > + res[i] = table.copy(part) > + if res[i].is_nullable == nil then > + res[i].is_nullable = false > + end > + end > + return res > +end > + > +local key_def_new_cases = { > + -- Cases to call before box.cfg{}. > + { > + 'Pass a field on an unknown type', > + parts = {{ > + fieldno = 2, > + type = 'unknown', > + }}, > + exp_err = 'Unknown field type: unknown', > + }, > + { > + 'Try to use collation_id before box.cfg{}', > + parts = {{ > + fieldno = 1, > + type = 'string', > + collation_id = 2, > + }}, > + exp_err = coll_not_found(1, 2), > + }, > + { > + 'Try to use collation before box.cfg{}', > + parts = {{ > + fieldno = 1, > + type = 'string', > + collation = 'unicode_ci', > + }}, > + exp_err = coll_not_found(1, 'unicode_ci'), > + }, > + function() > + -- For collations. > + box.cfg{} > + end, > + -- Cases to call after box.cfg{}. > + { > + 'Try to use both collation_id and collation', > + parts = {{ > + fieldno = 1, > + type = 'string', > + collation_id = 2, > + collation = 'unicode_ci', > + }}, > + exp_err = 'Conflicting options: collation_id and collation', > + }, > + { > + 'Unknown collation_id', > + parts = {{ > + fieldno = 1, > + type = 'string', > + collation_id = 999, > + }}, > + exp_err = coll_not_found(1, 999), > + }, > + { > + 'Unknown collation name', > + parts = {{ > + fieldno = 1, > + type = 'string', > + collation = 'unknown', > + }}, > + exp_err = 'Unknown collation: "unknown"', > + }, > + { > + 'Bad parts parameter type', > + parts = 1, > + exp_err = usage_error, > + }, > + { > + 'No parameters', > + params = {}, > + exp_err = usage_error, > + }, > + { > + 'Two parameters', > + params = {{}, {}}, > + exp_err = usage_error, > + }, > + { > + 'Invalid JSON path', > + parts = {{ > + fieldno = 1, > + type = 'string', > + path = '[3[', > + }}, > + exp_err = 'Wrong index options (field 1): invalid path', > + }, > + { > + 'Success case; zero parts', > + parts = {}, > + exp_err = nil, > + }, > + { > + 'Success case; one part', > + parts = {{ > + fieldno = 1, > + type = 'string', > + }}, > + exp_err = nil, > + }, > + { > + 'Success case; one part with a JSON path', > + parts = {{ > + fieldno = 1, > + type = 'string', > + path = '[3]', > + }}, > + exp_err = nil, > + }, > +} > + > +local test = tap.test('key_def') > + > +test:plan(#key_def_new_cases - 1 + 7) > +for _, case in ipairs(key_def_new_cases) do > + if type(case) == 'function' then > + case() > + else > + local ok, res > + if case.params then > + ok, res = pcall(key_def_lib.new, unpack(case.params)) > + else > + ok, res = pcall(key_def_lib.new, case.parts) > + end > + if case.exp_err == nil then > + ok = ok and type(res) == 'cdata' and > + ffi.istype('struct key_def', res) > + test:ok(ok, case[1]) > + else > + local err = tostring(res) -- cdata -> string > + test:is_deeply({ok, err}, {false, case.exp_err}, case[1]) > + end > + end > +end > + > +-- Prepare source data for test cases. > +local parts_a = { > + {type = 'unsigned', fieldno = 1}, > +} > +local parts_b = { > + {type = 'number', fieldno = 2}, > + {type = 'number', fieldno = 3}, > +} > +local parts_c = { > + {type = 'scalar', fieldno = 2}, > + {type = 'scalar', fieldno = 1}, > + {type = 'string', fieldno = 4, is_nullable = true}, > +} > +local key_def_a = key_def_lib.new(parts_a) > +local key_def_b = key_def_lib.new(parts_b) > +local key_def_c = key_def_lib.new(parts_c) > +local tuple_a = box.tuple.new({1, 1, 22}) > +local tuple_b = box.tuple.new({2, 1, 11}) > +local tuple_c = box.tuple.new({3, 1, 22}) > + > +-- Case: extract_key(). > +test:test('extract_key()', function(test) > + test:plan(9) > + > + test:is_deeply(key_def_a:extract_key(tuple_a):totable(), {1}, 'case 1') > + test:is_deeply(key_def_b:extract_key(tuple_a):totable(), {1, 22}, 'case 2') > + > + -- JSON path. > + local res = key_def_lib.new({ > + {type = 'string', fieldno = 1, path = 'a.b'}, > + }):extract_key(box.tuple.new({{a = {b = 'foo'}}})):totable() > + test:is_deeply(res, {'foo'}, 'JSON path (tuple argument)') > + > + local res = key_def_lib.new({ > + {type = 'string', fieldno = 1, path = 'a.b'}, > + }):extract_key({{a = {b = 'foo'}}}):totable() > + test:is_deeply(res, {'foo'}, 'JSON path (table argument)') I like key_def_new_cases - they are very easy to read or extend. I don't quite like the tests below, because they refer to objects created a few screens above (tuple_a, key_def_a, etc). Could you please rewrite them in a similar to key_def_new_cases fashion, without referring to any predefined variables? Also, it looks like you don't test json tuple comparison properly. Please add more test cases.