From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <vdavydov.dev@gmail.com>
Date: Fri, 21 Jun 2019 18:53:07 +0300
From: Vladimir Davydov <vdavydov.dev@gmail.com>
Subject: Re: [PATCH 2/2] decimal: expose decimal type to lua.
Message-ID: <20190621155307.z6pnpfbgqgni6zud@esperanza>
References: <cover.1560958964.git.sergepetrenko@tarantool.org>
 <89c41adbdf7452c1ef4a8f478363fe2ccb95a848.1560958964.git.sergepetrenko@tarantool.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <89c41adbdf7452c1ef4a8f478363fe2ccb95a848.1560958964.git.sergepetrenko@tarantool.org>
To: Serge Petrenko <sergepetrenko@tarantool.org>
Cc: tarantool-patches@freelists.org
List-ID: <tarantool-patches.dev.tarantool.org>

On Wed, Jun 19, 2019 at 06:58:05PM +0300, Serge Petrenko wrote:
> Add a decimal library to lua.
> 
> Part of #692
> 
> @TarantoolBot document
> Title: Document decimal module in lua.
> 
> First of all, you have to require the package via
> `decimal = require('decimal')`
> Now you can construct decimals via `tonumber` method.
> Decimals may be constructed from lua numbers, strings, unsigned and
> signed 64 bit integers.
> Decimal is a fixed-point type with maximum 38 digits of precision. All
> the calculations are exact, so, be careful when constructing decimals
> from lua numbers: they may hold only 15 decimal digits of precision.
> You are advised to construct decimals from strings, since strings
> represent decimals exactly, and vice versa.
> 
> ```
> a = decimal.tonumber(123e-7)
> b = decimal.tonumber('123.456')
> c = decimal.tonumber('123.456e2')
> d = decimal.tonumber(123ULL)
> e = decimal.tonumber(2)
> ```

tonumber is a confusing name IMO. Let's rename it to decimal.new()

> The allowed operations are addition, subtraction, division,
> multiplication and power. If at least one of the operands is decimal,
> decimal operations are performed. The other operand may be either
> decimal or string, containing a number representation, or a lua number.
> When the operation is called as `decimal.opname`, both operands may be
> strings or lua numbers, e.g. `decimal.add(23, '123.456') == 146.456`:
> ```

Can these functions fail (overflow, underflow)? What happens on error?
Please mention in this document.

> tarantool> a + b
> ---
> - '123.456012300000000'
> ...
> 
> tarantool> decimal.add(a,b)
> ---
> - '123.456012300000000'
> ...
> 
> tarantool> c - d
> ---
> - '12222.6'
> ...
> 
> tarantool> decimal.sub(c,d)

I don't think we need decimal.add/sub/div/mul methods as long as we have
corresponding operators.

> The following math functions are also supported:
> log10, ln, exp, sqrt, pow. When specified as
> `decimal.opname()`, operations may be performed on
> strings and lua numbers.
> ```
> f = decimal.tonumber(100)
> tarantool> f:log10()

This looks weird. I think we should only allow decimal.log10(x),
not x:log10().

> ---
>  src/CMakeLists.txt        |   1 +
>  src/lua/decimal.c         | 327 ++++++++++++++++++++++++++++++++++++++
>  src/lua/decimal.h         |  39 +++++
>  src/lua/init.c            |   2 +
>  test/app/decimal.result   | 172 ++++++++++++++++++++
>  test/app/decimal.test.lua |  50 ++++++
>  6 files changed, 591 insertions(+)
>  create mode 100644 src/lua/decimal.c
>  create mode 100644 src/lua/decimal.h
>  create mode 100644 test/app/decimal.result
>  create mode 100644 test/app/decimal.test.lua

Please consider implementing this using Lua ffi. Take a look at
src/lua/uuid.lua for example. They say that ffi implementation is
generally faster than Lua C, because it doesn't break JIT traces.
It might be worth benchmarking the two approaches.