Adding Move On Sui Semgrep grammer (#508)

* First instance of adding semgrep grammars for Move-on-Sui language.

Currently everything in the tree-sitter grammer is fully supported and all tests pass.

However, several things are missing from the tree-sitter grammar today:
- Proper support for annotations
- Support for scripts
- The most recent changes to the Move language

* Moving Move on Sui tree-sitter to my branch for accelerated developement

* Added a package to stats, this has a lot of tests that are updated with move as this is a vscode highigher

* Fixing Macro parsing bug due to meta variables eating the $ and 1 only upcase chars

* updating tree-sitter-move-on-sui repo

* Fixing Macros so now macro variables that start with $ parse.

* Simplifying _identifier_or_metavariable

* Adding ellipsis to module_access

.gitmodules

@@ -123,4 +123,4 @@
 	url = https://github.com/Decurity/tree-sitter-circom.git
 [submodule "lang/semgrep-grammars/src/tree-sitter-move-on-sui"]
 	path = lang/semgrep-grammars/src/tree-sitter-move-on-sui
-	url = https://github.com/tzakian/tree-sitter-move.git
+	url = https://github.com/maxmysten/tree-sitter-move.git

lang/Makefile

@@ -138,7 +138,8 @@ STAT_LANGUAGES2 = \
   rust \
   solidity \
   tsx \
-  typescript
+  typescript \
+  move-on-sui
 
 STAT_LANGUAGES = $(STAT_LANGUAGES1) $(STAT_LANGUAGES2)
 

lang/move-on-sui/projects.txt

@@ -4,3 +4,4 @@
 https://github.com/AftermathFinance/move-interfaces
 https://github.com/Bucket-Protocol/bucket-interface
 https://github.com/MystenLabs/sui-axelar
+https://github.com/MystenLabs/prettier-move/

lang/move-on-sui/test/ok/address.move

@@ -0,0 +1,79 @@
+// Copyright (c) Mysten Labs, Inc.
+// SPDX-License-Identifier: Apache-2.0
+
+module sui::address {
+    use sui::hex;
+    use std::ascii;
+    use std::bcs;
+    use std::string;
+
+    /// The length of an address, in bytes
+    const LENGTH: u64 = 32;
+
+    // The largest integer that can be represented with 32 bytes: 2^(8*32) - 1
+    const MAX: u256 = 115792089237316195423570985008687907853269984665640564039457584007913129639935;
+
+    #[allow(unused_const)]
+    /// Error from `from_bytes` when it is supplied too many or too few bytes.
+    const EAddressParseError: u64 = 0;
+
+    /// Convert `a` into a u256 by interpreting `a` as the bytes of a big-endian integer
+    /// (e.g., `to_u256(0x1) == 1`)
+    public native fun to_u256(a: address): u256;
+
+    spec to_u256 {
+        pragma opaque;
+        // TODO: stub to be replaced by actual abort conditions if any
+        aborts_if [abstract] true;
+        // TODO: specify actual function behavior
+    }
+
+    /// Convert `n` into an address by encoding it as a big-endian integer (e.g., `from_u256(1) = @0x1`)
+    /// Aborts if `n` > `MAX_ADDRESS`
+    public native fun from_u256(n: u256): address;
+
+    spec from_u256 {
+        pragma opaque;
+        // TODO: stub to be replaced by actual abort conditions if any
+        aborts_if [abstract] true;
+        // TODO: specify actual function behavior
+    }
+
+    /// Convert `bytes` into an address.
+    /// Aborts with `EAddressParseError` if the length of `bytes` is not 32
+    public native fun from_bytes(bytes: vector<u8>): address;
+
+    spec from_bytes {
+        pragma opaque;
+        // TODO: stub to be replaced by actual abort conditions if any
+        aborts_if [abstract] true;
+        // TODO: specify actual function behavior
+    }
+
+    /// Convert `a` into BCS-encoded bytes.
+    public fun to_bytes(a: address): vector<u8> {
+        bcs::to_bytes(&a)
+    }
+
+    /// Convert `a` to a hex-encoded ASCII string
+    public fun to_ascii_string(a: address): ascii::String {
+        ascii::string(hex::encode(to_bytes(a)))
+    }
+
+    /// Convert `a` to a hex-encoded ASCII string
+    public fun to_string(a: address): string::String {
+        string::from_ascii(to_ascii_string(a))
+    }
+
+    /// Length of a Sui address in bytes
+    public fun length(): u64 {
+        LENGTH
+    }
+
+    /// Largest possible address
+    public fun max(): u256 {
+        MAX
+    }
+
+    spec module { pragma verify = false; }
+}

lang/move-on-sui/test/ok/annotations.move

@@ -0,0 +1,9 @@
+#[allow(unused_const, unused_variable)]
+
+#[allow(unused_const, unused_variable)]
+#[test, expected_failure(abort_code = ::staking::staking_tests::ENotImplemented)]
+#[test, expected_failure(abort_code = other_module::ENotFound)]
+#[expected_failure(arithmetic_error, location = pkg_addr::other_module)]
+
+// #[allow(unused_const, unused_variable)]
+module c::k {}

lang/move-on-sui/test/ok/as_expressions.move

@@ -0,0 +1,71 @@
+module foo::bar {
+    fun f(): u64 {
+        1 as u64;
+        1 + 2 as u64;
+        (1 + 2) as u64;
+        1 + (2 as u64);
+        v[1 as u64];
+        1 as u64 + 2;
+        (1 as u64) + 2;
+    }
+
+    fun simple(cond: bool, x: u32) {
+        (if (cond) x else { x }) as u32;
+        (if (cond) x else { x } as u32);
+        (loop { break 0 }) as u32;
+        (loop { break 0 } as u32);
+        ('l: { 0 }) as u32;
+        ('l: { 0 } as u32);
+        (return) as u32;
+        (return as u32);
+        loop {
+            (break) as u32;
+            (break as u32);
+            (continue) as u32;
+            (continue as u32);
+        };
+        (0 as u32) as u32;
+        (0 as u32 as u32);
+    }
+
+    public fun f_imm(s: &S): &u64 { &s.f }
+
+    fun ops(x: u8, y: u8) {
+        (x + y as u32);
+        (x - y as u32);
+        (x * y as u32);
+        (x / y as u32);
+        (x % y as u32);
+        (x & y as u32);
+        (x | y as u32);
+        (x ^ y as u32);
+        (x << y as u32);
+        (x >> y as u32);
+
+        (x + y) as u32;
+        (x - y) as u32;
+        (x * y) as u32;
+        (x / y) as u32;
+        (x % y) as u32;
+        (x & y) as u32;
+        (x | y) as u32;
+        (x ^ y) as u32;
+        (x << y) as u32;
+        (x >> y) as u32;
+    }
+
+
+    public struct S has copy, drop { f: u64 }
+
+    fun dotted(cond: bool, mut s: S) {
+        (1 + s.f) as u32;
+        (1 + s.f as u32);
+        (1 + S { f: 0 }.f) as u32;
+        (1 + S { f: 0 }.f as u32);
+        (*if (cond) { &0 } else { &mut 0 }) as u32;
+        // this case still does not work
+        // (*if (cond) { &0 } else { &mut 0 } as u32);
+        (*if (cond) { &s } else {&mut s}.f_imm()) as u32;
+        (*if (cond) { &s } else {&mut s}.f_imm() as u32);
+    }
+}

lang/move-on-sui/test/ok/ascii.move

@@ -0,0 +1,151 @@
+// Copyright (c) Mysten Labs, Inc.
+// SPDX-License-Identifier: Apache-2.0
+
+/// The `ASCII` module defines basic string and char newtypes in Move that verify
+/// that characters are valid ASCII, and that strings consist of only valid ASCII characters.
+module std::ascii {
+    use std::vector;
+    use std::option::{Self, Option};
+
+    /// An invalid ASCII character was encountered when creating an ASCII string.
+    const EINVALID_ASCII_CHARACTER: u64 = 0x10000;
+
+   /// The `String` struct holds a vector of bytes that all represent
+   /// valid ASCII characters. Note that these ASCII characters may not all
+   /// be printable. To determine if a `String` contains only "printable"
+   /// characters you should use the `all_characters_printable` predicate
+   /// defined in this module.
+   struct String has copy, drop, store {
+       bytes: vector<u8>,
+   }
+   spec String {
+       invariant forall i in 0..len(bytes): is_valid_char(bytes[i]);
+   }
+
+   /// An ASCII character.
+   struct Char has copy, drop, store {
+       byte: u8,
+   }
+   spec Char {
+       invariant is_valid_char(byte);
+   }
+
+    /// Convert a `byte` into a `Char` that is checked to make sure it is valid ASCII.
+    public fun char(byte: u8): Char {
+        assert!(is_valid_char(byte), EINVALID_ASCII_CHARACTER);
+        Char { byte }
+    }
+    spec char {
+        aborts_if !is_valid_char(byte) with EINVALID_ASCII_CHARACTER;
+    }
+
+    /// Convert a vector of bytes `bytes` into an `String`. Aborts if
+    /// `bytes` contains non-ASCII characters.
+    public fun string(bytes: vector<u8>): String {
+       let x = try_string(bytes);
+       assert!(
+            option::is_some(&x),
+            EINVALID_ASCII_CHARACTER
+       );
+       option::destroy_some(x)
+    }
+    spec string {
+        aborts_if exists i in 0..len(bytes): !is_valid_char(bytes[i]) with EINVALID_ASCII_CHARACTER;
+    }
+
+    /// Convert a vector of bytes `bytes` into an `String`. Returns
+    /// `Some(<ascii_string>)` if the `bytes` contains all valid ASCII
+    /// characters. Otherwise returns `None`.
+    public fun try_string(bytes: vector<u8>): Option<String> {
+       let len = vector::length(&bytes);
+       let i = 0;
+       while ({
+           spec {
+               invariant i <= len;
+               invariant forall j in 0..i: is_valid_char(bytes[j]);
+           };
+           i < len
+       }) {
+           let possible_byte = *vector::borrow(&bytes, i);
+           if (!is_valid_char(possible_byte)) return option::none();
+           i = i + 1;
+       };
+       spec {
+           assert i == len;
+           assert forall j in 0..len: is_valid_char(bytes[j]);
+       };
+       option::some(String { bytes })
+    }
+
+    /// Returns `true` if all characters in `string` are printable characters
+    /// Returns `false` otherwise. Not all `String`s are printable strings.
+    public fun all_characters_printable(string: &String): bool {
+       let len = vector::length(&string.bytes);
+       let i = 0;
+       while ({
+           spec {
+               invariant i <= len;
+               invariant forall j in 0..i: is_printable_char(string.bytes[j]);
+           };
+           i < len
+       }) {
+           let byte = *vector::borrow(&string.bytes, i);
+           if (!is_printable_char(byte)) return false;
+           i = i + 1;
+       };
+       spec {
+           assert i == len;
+           assert forall j in 0..len: is_printable_char(string.bytes[j]);
+       };
+       true
+    }
+    spec all_characters_printable {
+        ensures result ==> (forall j in 0..len(string.bytes): is_printable_char(string.bytes[j]));
+    }
+
+    public fun push_char(string: &mut String, char: Char) {
+        vector::push_back(&mut string.bytes, char.byte);
+    }
+    spec push_char {
+        ensures len(string.bytes) == len(old(string.bytes)) + 1;
+    }
+
+    public fun pop_char(string: &mut String): Char {
+        Char { byte: vector::pop_back(&mut string.bytes) }
+    }
+    spec pop_char {
+        ensures len(string.bytes) == len(old(string.bytes)) - 1;
+    }
+
+    public fun length(string: &String): u64 {
+        vector::length(as_bytes(string))
+    }
+
+    /// Get the inner bytes of the `string` as a reference
+    public fun as_bytes(string: &String): &vector<u8> {
+       &string.bytes
+    }
+
+    /// Unpack the `string` to get its backing bytes
+    public fun into_bytes(string: String): vector<u8> {
+       let String { bytes } = string;
+       bytes
+    }
+
+    /// Unpack the `char` into its underlying byte.
+    public fun byte(char: Char): u8 {
+       let Char { byte } = char;
+       byte
+    }
+
+    /// Returns `true` if `b` is a valid ASCII character. Returns `false` otherwise.
+    public fun is_valid_char(b: u8): bool {
+       b <= 0x7F
+    }
+
+    /// Returns `true` if `byte` is an printable ASCII character. Returns `false` otherwise.
+    public fun is_printable_char(byte: u8): bool {
+       byte >= 0x20 && // Disallow metacharacters
+       byte <= 0x7E // Don't allow DEL metacharacter
+    }
+}