Tarantool development patches archive
 help / color / mirror / Atom feed
From: "n.pettik" <korablev@tarantool.org>
To: tarantool-patches@freelists.org
Cc: Vladislav Shpilevoy <v.shpilevoy@tarantool.org>
Subject: [tarantool-patches] Re: [PATCH 3/6] test: fix sqltester methods to drop all tables/views
Date: Tue, 11 Dec 2018 21:29:23 +0300	[thread overview]
Message-ID: <E61B4DC9-6D7C-4795-A3CB-2B3697C6D182@tarantool.org> (raw)
In-Reply-To: <9464fa79-cbd0-7a46-60b7-98394a5a3766@tarantool.org>



> On 10 Dec 2018, at 17:16, Vladislav Shpilevoy <v.shpilevoy@tarantool.org> wrote:
> 
> Thanks for the patch! See 3 comments below.
> 
> On 10/12/2018 00:30, Nikita Pettik wrote:
>> Since we are going to remove string of SQL "CREATE TABLE ..." statement
>> from space's opts, lets rework methods in sqltester to drop all tables
>> and views in order to avoid relying on this parameter.
>> Part of #2647
>> ---
>>  test/sql-tap/drop_all.test.lua        |  4 ++--
>>  test/sql-tap/identifier_case.test.lua |  6 +++---
>>  test/sql-tap/lua/sqltester.lua        | 38 ++++++++++++++---------------------
>>  3 files changed, 20 insertions(+), 28 deletions(-)
>> diff --git a/test/sql-tap/lua/sqltester.lua b/test/sql-tap/lua/sqltester.lua
>> index 8751ef820..672d2e67d 100644
>> --- a/test/sql-tap/lua/sqltester.lua
>> +++ b/test/sql-tap/lua/sqltester.lua
>> @@ -277,7 +277,7 @@ end
>>  test.catchsql2 = catchsql2
>>    -- Show the VDBE program for an SQL statement but omit the Trace
>> --- opcode at the beginning.  This procedure can be used to prove
>> +-- opcode at the beginning.  This procedure can be used to PROVE
> 
> 1. ???

Oops, fixed:

diff --git a/test/sql-tap/lua/sqltester.lua b/test/sql-tap/lua/sqltester.lua
index 672d2e67d..076510932 100644
--- a/test/sql-tap/lua/sqltester.lua
+++ b/test/sql-tap/lua/sqltester.lua
@@ -277,7 +277,7 @@ end
 test.catchsql2 = catchsql2
 
 -- Show the VDBE program for an SQL statement but omit the Trace
--- opcode at the beginning.  This procedure can be used to PROVE
+-- opcode at the beginning.  This procedure can be used to prove

> 
>>  -- that different SQL statements generate exactly the same VDBE code.
>>  local function explain_no_trace(self, sql)
>>      tr = execsql(self, "EXPLAIN "..sql)
>> @@ -288,35 +288,27 @@ local function explain_no_trace(self, sql)
>>  end
>>  test.explain_no_trace = explain_no_trace
>>  json = require("json")
>> -function test.find_spaces_by_sql_statement(self, statement)
>> -    local spaces = box.space._space:select{}
>> -    local res = {}
>> -    for _, space in ipairs(spaces) do
>> -        local name_num = 3-- [3] is space name
>> -        local options_num = 6-- [6] is space options
>> -        if not space[name_num]:startswith("_") and space[options_num]["sql"] then
>> -            if string.find(space[options_num]["sql"], statement) then
>> -                table.insert(res, space[name_num])
>> -            end
>> -        end
>> -    end
>> -    return res
>> -end
>>    function test.drop_all_tables(self)
>> -    local tables = test:find_spaces_by_sql_statement("CREATE TABLE")
>> -    for _, table in ipairs(tables) do
>> -        test:execsql(string.format([[DROP TABLE "%s";]], table))
>> +    local entry_count = 0
>> +    for _, v in box.space._space:pairs() do
>> +        if v[1] >= 512 then
>> +            test:execsql(string.format([[DROP TABLE "%s";]], v[3]))
>> +            entry_count = entry_count + 1
> 
> 2. Since DD integration is finished, can we do box.space[v3]:drop()
> instead of executing SQL? I think it can help us to find more bugs.
> And it is faster.

Seems like:

@@ -293,7 +293,7 @@ function test.drop_all_tables(self)
     local entry_count = 0
     for _, v in box.space._space:pairs() do
         if v[1] >= 512 then
-            test:execsql(string.format([[DROP TABLE "%s";]], v[3]))
+            box.space[v[3]]:drop()
             entry_count = entry_count + 1
         end
     end
@@ -303,8 +303,8 @@ end
 function test.drop_all_views(self)
     local entry_count = 0
     for _, v in box.space._space:pairs() do
-        if v[1] > 512 and v[6]["view"] == true then
-            test:execsql(string.format([[DROP VIEW "%s";]], v[3]))
+        if v[1] > 512 and v[6].view == true then
+            box.space[v[3]]:drop()
             entry_count = entry_count + 1
         end

> 
>> +        end
>>      end
>> -    return tables
>> +    return entry_count
>>  end
>>    function test.drop_all_views(self)
>> -    local views = test:find_spaces_by_sql_statement("CREATE VIEW")
>> -    for _, view in ipairs(views) do
>> -        test:execsql(string.format([[DROP VIEW "%s";]], view))
>> +    local entry_count = 0
>> +    for _, v in box.space._space:pairs() do
>> +        if v[1] > 512 and v[6]["view"] == true then
> 
> 3. IMO .view looks better than ["view"], but I do not insist. Up to
> you.

I don’t really care, so let's use your approach (see above).

  reply	other threads:[~2018-12-11 18:29 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-12-09 21:30 [tarantool-patches] [PATCH 0/6] Remove string of SQL statement from opts Nikita Pettik
2018-12-09 21:30 ` [tarantool-patches] [PATCH 1/6] sql: avoid calling sql_encode_table_opts() during trigger creation Nikita Pettik
2018-12-10 14:17   ` [tarantool-patches] " Vladislav Shpilevoy
2018-12-11 18:29     ` n.pettik
2018-12-09 21:30 ` [tarantool-patches] [PATCH 2/6] sql: don't update SQL string during renaming Nikita Pettik
2018-12-10 14:16   ` [tarantool-patches] " Vladislav Shpilevoy
2018-12-11 18:29     ` n.pettik
2018-12-12 12:36       ` Vladislav Shpilevoy
2018-12-13 12:42         ` n.pettik
2018-12-09 21:30 ` [tarantool-patches] [PATCH 3/6] test: fix sqltester methods to drop all tables/views Nikita Pettik
2018-12-10 14:16   ` [tarantool-patches] " Vladislav Shpilevoy
2018-12-11 18:29     ` n.pettik [this message]
2018-12-09 21:30 ` [tarantool-patches] [PATCH 4/6] sql: don't add string of 'CREATE TABLE...' to space opts Nikita Pettik
2018-12-10 14:17   ` [tarantool-patches] " Vladislav Shpilevoy
2018-12-11 18:29     ` n.pettik
2018-12-09 21:30 ` [tarantool-patches] [PATCH 5/6] sql: don't add string of 'CREATE INDEX ...' to index opts Nikita Pettik
2018-12-10 14:18   ` [tarantool-patches] " Vladislav Shpilevoy
2018-12-11 18:29     ` n.pettik
2018-12-09 21:30 ` [tarantool-patches] [PATCH 6/6] Remove SQL string from " Nikita Pettik
2018-12-25 13:45 ` [tarantool-patches] Re: [PATCH 0/6] Remove string of SQL statement from opts Kirill Yukhin

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=E61B4DC9-6D7C-4795-A3CB-2B3697C6D182@tarantool.org \
    --to=korablev@tarantool.org \
    --cc=tarantool-patches@freelists.org \
    --cc=v.shpilevoy@tarantool.org \
    --subject='[tarantool-patches] Re: [PATCH 3/6] test: fix sqltester methods to drop all tables/views' \
    /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