Currently, we simply include the libyaml source directory into the main
project. The problem is that libyaml uses ICU. If ICU is built outside
the main project cmake config, as it's the case with the static-build,
both the main project and libyaml cmake configs will use the same ICU
version. However, if we build ICU in the main project, as we intend to
do to resolve #9242, it may not work. To fix that, we need to use the
external project API to build libyaml.

Needed for #9242

NO_DOC=build
NO_TEST=build
NO_CHANGELOG=build

There's no need in it because we closed Tarantool symbols long ago and
now export only those symbols that are specified in exports files. Let's
remove the useless code.

Needed for #9242

NO_DOC=build
NO_TEST=build
NO_CHANGELOG=build

The tests are TAP compatible and applicable to all Tarantool builds so
there's no need to run them with ctest. We just need to add a couple
skip conditions:
 - The luarocks test shouldn't be run on dynamic builds because luarocks
   modules aren't embedded there.
 - The traceback test should be run only if ENABLE_BACKTRACE was set at
   build time.

Part of #9242

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

NO_DOC=tests
NO_TEST=tests
NO_CHANGELOG=tests

unit test, in case when assertion failed

NO_DOC=test
NO_CHANGELOG=test

See the doc bot request for the description.

Benchmark results:

NO_WRAP
$ taskset 0x2 compare.py benchmarks ./memtx.perftest.old ./memtx.perftest.new \
  --benchmark_min_warmup_time=10 \
  --benchmark_repetitions=30 \
  --benchmark_report_aggregates_only=true \
  --benchmark_filter=TreeReplaceRandomExistingKeys
[...]
Comparing ./memtx.perftest.old to ./memtx.perftest.new
Benchmark                                                           Time             CPU      Time Old      Time New       CPU Old       CPU New
------------------------------------------------------------------------------------------------------------------------------------------------
MemtxFixture/TreeReplaceRandomExistingKeys_mean                  +0.0097         +0.0097          1073          1084          1073          1084
MemtxFixture/TreeReplaceRandomExistingKeys_median                +0.0075         +0.0075          1062          1070          1062          1070
MemtxFixture/TreeReplaceRandomExistingKeys_stddev                -0.1207         -0.1208            56            49            56            49
MemtxFixture/TreeReplaceRandomExistingKeys_cv                    -0.1291         -0.1292             0             0             0             0
NO_WRAP

Closes #6762

@TarantoolBot document
Title: Document `space:stat()`
Product: Tarantool
Since: 3.0
Root document: https://www.tarantool.io/en/doc/latest/reference/reference_lua/box_space/

space_object:stat() - Get statistics on the memory usage

Returns a table with the cumulative statistics on the memory usage by
tuples in the space. The statistics is grouped by arena types: "memtx"
or "malloc".
For a detailed description of each field see `tuple_object:info()`.

Note: The statistics is collected only for memtx storage engine. For other
types of spaces, an empty table is returned.

Example:

```
tarantool> box.space.test:stat()
---
- tuple:
    memtx:
      data_size: 5100699
      header_size: 96
      field_map_size: 40
      waste_size: 143093
    malloc:
      data_size: 18850077
      header_size: 70
      field_map_size: 28
      waste_size: 0
...
```

See the doc bot request for the description.

Part of #6762

NO_CHANGELOG=next commit

@TarantoolBot document
Title: Document `tuple:info()`
Product: Tarantool
Since: 3.0
Root document: https://www.tarantool.io/en/doc/latest/reference/reference_lua/box_tuple/

tuple_object:info() - Get information about the tuple

Returns a table with the following fields:

 * data_size - Size of the MessagePack data in the tuple.
   This number equals to number returned by `tuple_object:bsize()`.
 * header_size - Size of the internal tuple header.
 * field_map_size - Size of the field map.
   Field map is used to speed up access to indexed fields of the tuple.
 * waste_size - The amount of excess memory used to store the tuple
   in mempool.
 * arena - Type of the arena where the tuple is allocated.
   Possible values are: "memtx", "malloc", "runtime".

It is similar to TreeGetRandomExistingKeys, but performs box_replace()
instead of box_index_get().

Needed for #6762

NO_DOC=benchmark
NO_TEST=benchmark
NO_CHANGELOG=benchmark

Before this patch, nameInUsingClause() expected old_col to be non-NULL,
but this may not be the case. This patch fixes this feature accordingly.

Follow-up #4467

NO_DOC=bugfix
NO_CHANGELOG=feature was not released yet

This patch replaces the SQL mechanism for setting expression default
values with the BOX mechanism.

Closes #8793

@TarantoolBot document
Title: field default value in SQL

SQL now uses the default mechanism for BOX fields. The main difference
visible to the user is that now all NULL values explicitly inserted into
a field with a default value will be replaced with a default value,
which is different from the original SQL behavior. Additionally, default
values are now inserted instead of NULL after the SQL BEFORE INSERT
trigger but before the BOX BEFORE INSERT trigger. Previously,
SQL default values were inserted before the SQL BEFORE INSERT trigger.

This patch moves the VDBE function creation code from
vdbe_emit_ck_constraint_create() to the new function
vdbe_emit_create_function(). Because of this, the error message changes
slightly.

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

This patch replaces the SQL mechanism for setting literal default values
with the BOX mechanism.

Part of #8793

NO_DOC=will be added later
NO_CHANGELOG=will be added later

After this patch, non-numeric values will not be accepted by the
parser as default values if they have '+' or '-' before them.

NO_DOC=will be added later

After this patch SQL_EXPR functions will be allowed to set as default
values.

Part of #8793

@TarantoolBot document
Title: SQL_EXPR functions as default values

SQL_EXPR functions can now be set as default values. Note that all
arguments to the SQL_EXPR functions will be replaced by the space field
'default' value. This only matters for the SQL_EXPR used as default from
Lua, since in SQL it is not allowed to have arguments in the default
expression.

This patch removes support for default values in SHOW CREATE TABLE,
since it is not always possible to set the described default value in
SQL. Support for literals and SQL_EXPR functions with zero arguments is
potentially possible, and this support may be introduced later.

Part of #8793

NO_DOC=will be introduced later

Prior to this patch it was possible to lose BOX field default values
after ALTER TABLE ADD COLUMN.

NO_DOC=bugfix

* Revert "Update cur_L on exceptional path (arm)"
* Revert "arm64: fix cur_L restoration on error throw"
* Revert "Update cur_L on exceptional path"
* Revert "Fix cur_L tracking on exceptional path"
* Restore cur_L for specific Lua/C API use case.
* Fix Clang build.

Closes #6323
Part of #9145

NO_DOC=LuaJIT submodule bump
NO_TEST=LuaJIT submodule bump

See the documentation request below for details about the feature.

The template variables are substituted in three different places.

* On deducing configuration of the current instance from the cluster
  configuration.
* On calculation of URIs of replicaset peers to form
  `box.cfg({replication = <...>})` value.
* On calculation of URIs of all cluster members for generation of
  `vshard.{router,storage}.cfg()` configuration.

That's why we need a lot of testing code around this small feature.

Part of #8862

@TarantoolBot document
Title: config: support {{ replicaset_name }} and {{ group_name }}

The declarative configuration supports `{{ instance_name }}` template
variable from the beginning.

Now, two more ones were added: `{{ replicaset_name }}` and
`{{ group_name }}`.

All the template variables are set according to the
instance/replicaset/group topology from the given configuration.

Example:

NO_WRAP
```yaml
console:
  socket: '{{ group_name }}/{{ replicaset_name }}/{{ instance_name }}.admin'

groups:
  g:
    replicasets:
      r:
        instances:
          i: {}
```
NO_WRAP

The admin console will be at `g/r/i.admin` for instance `i`.

Follows up commit d5c874e1 ("main: rewrite help message").

Thanks to Mergen Imeev (@ImeevMA) for pointing it out!

Part of #8862

NO_DOC=The documentation request is present in d5c874e1.
NO_CHANGELOG=same as NO_DOC

This commit aims several goals.

First, we're going to recommend `tt` as a tool to run and manage
tarantool instances. The tool is going to replace `tarantoolctl` and it
seems valuable to sync recipes across different parts of the
documentation (including this help message).

Second, the action options and modifier options are separated now. At
least one action option should be specified. I hope, it is more obvious
from the help message now.

Third, single line descriptions are replaced with a bit more extended
ones to give a user basic context to understand an influence of an
option.

Updated the man page accordingly: synchronized the options, removed
`tarantoolctl`.

Part of #8862

NO_TEST=the help message is not an API for a developer
NO_CHANGELOG=nothing new here, just better wording here and there

@TarantoolBot document
Title: Update CLI options documentation

https://www.tarantool.io/en/doc/latest/reference/configuration/#command-options
seems to need an update. See the linked commit for details and updated
`tarantool --help` message.

Prior to this patch, there were many possible code execution options
that were not covered by tests. After this commit, any assert(false)
inside box_wait_limbo_acked cause a crash.

Closes #7318

NO_DOC=test
NO_CHANGELOG=test

Though the range of values of the number being converted to string in the
tuple format map test never exceeds 10, GCC 12 issues warnings about the
size of the buffer to which the string version of the number is written:

NO_WRAP
```
test/unit/tuple_format_map.c:256:53: error: ‘%zu’ directive output may be truncated writing between 1 and 19 bytes into a region of size 4 [-Werror=format-
truncation=]
  256 |                 snprintf(name, lengthof(name), "test%zu", i);
      |                                                     ^~~
test/unit/tuple_format_map.c:256:48: note: directive argument in the range [0, 9223372036854775807]
  256 |                 snprintf(name, lengthof(name), "test%zu", i);
      |                                                ^~~~~~~~~
test/unit/tuple_format_map.c:256:17: note: ‘snprintf’ output between 6 and 24 bytes into a destination of size 8
  256 |                 snprintf(name, lengthof(name), "test%zu", i);
      |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
```
NO_WRAP

Let's make GCC happy and increase all the buffer sizes to 32.

NO_CHANGELOG=<build fix>
NO_DOC=<build fix>
NO_TEST=<build fix>

Closes #9321

NO_DOC=gdb extension
NO_TEST=gdb extension
NO_CHANGELOG=gdb extension

Closes #8054

@TarantoolBot document
Title: Document Lua helpers for encoding/decoding IPROTO packets

The following new constants and functions were added to the `box.iproto`
namespace:

 - `GREETING_SIZE`: Size of a Tarantool greeting message

 - `GREETING_PROTOCOL_LEN_MAX`: Max length of a protocol string that can
   be encoded in a Tarantool greeting message.

 - `GREETING_SALT_LEN_MAX`: Max length of a salt string that can be
   encoded in a Tarantool greeting message.

 - `box.iproto.encode_greeting({version = x, uuid = x, salt = x})`:
   Encodes a Tarantool greeting message. Takes a table. Returns a
   string. Raises on error. The protocol is set to "Binary" (IPROTO).

 - `box.iproto.decode_greeting(string)`: Decodes a Tarantool greeting
   message. Takes a string. Returns a table with the following fields:
   `protocol`, `version`, `uuid`, `salt`. Raises on error. The input
   string must be exactly `GREETING_SIZE` bytes long.

 - `box.iproto.encode_packet(header[, body])`: Encodes an IPROTO packet.
   Takes a packet header and optionally a body given as a table or a
   string. A table argument will be encoded in MsgPack. A string
   argument will be copied as is (it's supposed to contain valid MsgPack
   but it isn't enforced). Returns a string. Raises on error.

 - `box.iproto.decode_packet(string[, pos])`: Decodes an IPROTO packet.
   Takes a string containing one or more encoded IPROTO packets and
   optionally a position in the string to start decoding from. If the
   position is omitted, the function will start decoding from the
   beginning of the string. On success returns the decoded packet
   header, body, and the position in the string where decoding stopped.
   Both header and body are returned as `msgpack.object`. The body may
   be absent (set to nil). If the input is truncated, returns nil and
   the min number of bytes required to decode the packet. On failure,
   raises an error.

For examples, see [`test/app-luatest/iproto_encoder_test.lua`][1].

[1]: https://github.com/tarantool/tarantool/blob/master/test/app-luatest/iproto_encoder_test.lua

This patch add second lookup for constraint name in the DROP CONSTRAINT
statement.

Part of #4467

@TarantoolBot document
Title: Second lookup for names in SQL

Since all names are now case sensitive, a second lookup is now supported
to maintain backward compatibility. A second lookup by name will be
performed if the object was not found by name in the first lookup and if
the name is not enclosed in double quotation marks. The name for the
second lookup is normalized using old rules.

This patch add second lookup for index name in the following cases:
- index in INDEXED BY clause;
- index in DROP INDEX statement;
- index in PRAGMA statement.

Part of #4467

NO_DOC=will be added later
NO_CHANGELOG=will be added later

This patch adds second lookup for function names in expressions.

Part of #4467

NO_DOC=will be added later
NO_CHANGELOG=will be added later

This patch add second lookup for column name in the following cases:
- columns in tuple foreign key creation clause;
- columns in field foreign key creation clause;
- columns in primary key creation clause;
- columns in unique constraint creation clause;
- columns in expressions;
- columns in UPDATE TABLE statement;
- columns in INSERT INTO statement;
- columns in CREATE INDEX statement;

Also, second lookup for table name in expressions also now supported.

Part of #4467

NO_DOC=will be added later
NO_CHANGELOG=will be added later

This patch add second lookup for space name in the following cases:
- foreign table in tuple foreign key creation clause;
- foreign table in field foreign key creation clause;
- table in DROP TABLE statement;
- view in DROP VIEW statement;
- table in SELECT FROM statement;
- table in DELETE FROM statement;
- table in TRUNCATE TABLE statement;
- table in UPDATE TABLE statement;
- table in INSERT INTO statement;
- table in CREATE INDEX statement;
- table in DROP INDEX statement;
- table in PRAGMA statement;
- table in SHOW CREATE TABLE statement;
- table in CREATE TRIGGER statement;
- table in ALTER TABLE statements;

Table names in expressions will be supported in subsequent commits.

Part of #4467

NO_DOC=will be added later
NO_CHANGELOG=will be added later

Part of #4467

NO_DOC=will be added later
NO_CHANGELOG=will be added later

This patch introduces a second lookup for the savepoint name in the
"RELEASE savepoint_name" and "ROLLBACK TO savepoint_name" statements.

Part of #4467

NO_DOC=will be added later
NO_CHANGELOG=will be added later

This patch makes identifiers in SQL case sensitive. Prior to this patch,
these identifiers were normalized unless they were enclosed in double
quotes.

Part of #4467

NO_DOC=will be added later
NO_CHANGELOG=will be added later

This patch introduces the sql_coll_id_by_expr() function. This function
is used to search for a collation based on a given expression.

Needed for #4467

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

This patch introduces the sql_coll_id_by_token() function. This function
is used to search for a collation based on a given token.

Needed for #4467

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

This patch introduces the sql_fieldno_by_item() function. This function
is used to search for a space field based on a given element of the
struct ExprList.

Needed for #4467

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

This patch introduces the sql_fieldno_by_expr() function. This function
is used to search for a space field based on a given expression.

Needed for #4467

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

This patch introduces the sql_fieldno_by_id() function. This function is
used to search for a space field based on a given element of the
struct IdList.

Needed for #4467

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

This patch introduces the sql_index_id_by_src() function. This function
is used to search for an index based on a given element of the
struct SrcList.

Needed for #4467

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

This patch introduces the sql_space_by_src() function. This function is
used to search for a space based on a given element of the
struct SrcList. The main purpose of this function is to introduce a
second lookup for spaces, so this function is not used where a second
lookup is not required.

Needed for #4467

NO_DOC=refactoring
NO_TEST=refactoring
NO_CHANGELOG=refactoring

Before this patch, the existence of an space for a
"SHOW CREATE TABLE table_name" statement was checked at runtime. This is
quite inconvenient for #4467, so this will now be checked when creating
the VDBE.

Needed for #4467

NO_DOC=no user-visible changes
NO_TEST=no user-visible changes
NO_CHANGELOG=no user-visible changes