Adding decimal literals

[aupiff]

fixes #90

READY TO MERGE.

NOTE:

• An ordinary decimal literal is assumed to to be of type uint256. To write a uint8 literal one must add the suffix u8, as in 5u8.

[aupiff]

I am working on automated gethtests in another branch.

I'll test various functions which use decimal literals on a local blockchain and then include those tests and the script to run them in separate PR. The feature-automated-geth-test branch is in the very early stages, but it should be ready by the end of the week. I'll be confident about the correctness of this decimal literals PR by Thursday.

[pirapira]

Did you have a look at src/exec/endToEnd.ml? It uses eth to test the compiled bytecode.

[pirapira]

The changes in this PR look good so far.

[aupiff]

@pirapira src/exec/endToEnd.ml looks good, but I have issues running make endToEnd.

ocamlbuild -use-ocamlfind -Is src/basics,src/ast,src/parse,src/lib,src/codegen -package batteries -package cryptokit -package rope -package rpclib -package rpclib.unix -package unix -package rpclib.json -package ppx_deriving -package ppx_deriving_rpc -package hex -use-menhir src/exec/endToEnd.native
Finished, 0 targets (0 cached) in 00:00:00.
+ /Users/aupiff/.opam/system/bin/ocamlfind ocamldep -package batteries -package cryptokit -package hex -package rope -package hex -package ppx_deriving_rpc -package ppx_deriving -package rpclib.json -package unix -package rpclib.unix -package rpclib -package rope -package cryptokit -package batteries -modules src/exec/endToEnd.ml > src/exec/endToEnd.ml.depends
ocamlfind: Package `ppx_deriving_rpc' not found                       
Command exited with code 2.        
Compilation unsuccessful after building 1 target (0 cached) in 00:00:00.
make: *** [endToEnd] Error 10

I'll write my own geth tests in bash for now and report on their results in a comment. Afterwards, I'll figure out how to compile add automated tests to endToEnd.ml

[aupiff]

Did some manually testing with geth, everything seems to work as expected. The contract below refers to A() from 025_declit_numeric.bbo

> contract.multiply5.call(2)       
10               
> contract.multiply7.call(3)       
21               
> contract.multiply7.call(30000)   
210000           
> contract.s.call()                
true             
> contract.ggg.call()              
7.771213812919199999999e+22        
> contract.g.call(324234)          
10               
> contract.f.call(1000)            
false            
> contract.f.call(4)               
true             
> contract.f.call(5)               
false            
> contract.gg.call(23)             
12138138405817283711        

contract A ()
{
    case(bool f(uint256 a))
    {
        return a < 5 then become A();
    }
    case(uint256 g(uint256 a))
    {
        return 1 + 9 then become A();
    }
    case(uint256 gg(uint256 a))
    {
        return 12138129191999999999 + 9213817283712 then become A();
    }
    case(uint256 ggg(uint256 a))
    {
        return 77712138129191999999999 - 9 then become A();
    }
    case(bool s())
    {
        return 239842934 > 289302 then become A();
    }
    case(uint8 i(uint8 a))
    {
        return 12u8 + 5u8 then become A();
    }
    case(uint8 j())
    {
        return 12u8 - 5u8 then become A();
    }
    case(bool k())
    {
        return 12u8 > 5u8 then become A();
    }
    case(bool q(uint8 a))
    {
        return a < 5u8 then become A();
    }
    case(uint8 multiply5(uint8 a))
    {
        return a * 5u8 then become A();
    }
    case(uint256 multiply7(uint256 a))
    {
        return a * 7 then become A();
    }
}

[pirapira]

The compilation error about endToEnd can be solved by this line: https://github.com/pirapira/bamboo/blob/master/.travis.yml#L40

Oh, now I notice (WIP) is gone. I'll have a look.

[aupiff]

@pirapira just fixed some silly bugs and got endToEnd to compile, thanks for letting me know what deps I was missing.

[pirapira]

Documentation is missing about endToEnd.

[pirapira]

I created an issue #210 about the documentation.

[pirapira]

👍

[pirapira]

(* uint8 has at most three digits *)

[pirapira]

Will you add a similar uint256_too_big.bbo?

[pirapira]

I found very minor things, but the PR is already good to merge.

[pirapira]

I'll deal with my own comments.