feat: add minimal ast bindings (gakonst#1167)

* feat: add minimal ast bindings

* feat: add recursive nodes

ethers-solc/src/artifact_output/configurable.rs

@@ -8,7 +8,7 @@ use crate::{
             BytecodeOutputSelection, ContractOutputSelection, EvmOutputSelection,
             EwasmOutputSelection,
         },
-        CompactContractBytecodeCow, DevDoc, Evm, Ewasm, FunctionDebugData, GasEstimates,
+        Ast, CompactContractBytecodeCow, DevDoc, Evm, Ewasm, FunctionDebugData, GasEstimates,
         LosslessAbi, Metadata, Offsets, Settings, StorageLayout, UserDoc,
     },
     ArtifactOutput, SolcConfig, SolcError, SourceFile,
@@ -52,8 +52,8 @@ pub struct ConfigurableContractArtifact {
     pub ir_optimized: Option<String>,
     #[serde(default, skip_serializing_if = "Option::is_none")]
     pub ewasm: Option<Ewasm>,
-    #[serde(default, skip_serializing_if = "serde_json::Value::is_null")]
-    pub ast: serde_json::Value,
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub ast: Option<Ast>,
 }
 
 impl ConfigurableContractArtifact {
@@ -303,7 +303,7 @@ impl ArtifactOutput for ConfigurableArtifacts {
             ir: artifact_ir,
             ir_optimized: artifact_ir_optimized,
             ewasm: artifact_ewasm,
-            ast: source_file.map(|s| s.ast.clone()).unwrap_or_default(),
+            ast: source_file.and_then(|s| s.ast.clone()),
         }
     }
 }

ethers-solc/src/artifacts/ast.rs

@@ -0,0 +1,223 @@
+//! Bindings for solc's `ast` output field
+
+use crate::artifacts::serde_helpers;
+use serde::{Deserialize, Serialize};
+use std::{collections::BTreeMap, fmt, fmt::Write, str::FromStr};
+
+/// Represents the AST field in the solc output
+#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+pub struct Ast {
+    #[serde(rename = "absolutePath")]
+    pub absolute_path: String,
+    pub id: usize,
+    #[serde(default, rename = "exportedSymbols")]
+    pub exported_symbols: BTreeMap<String, Vec<usize>>,
+    #[serde(rename = "nodeType")]
+    pub node_type: NodeType,
+    #[serde(with = "serde_helpers::display_from_str")]
+    pub src: SourceLocation,
+    #[serde(default, skip_serializing_if = "Vec::is_empty")]
+    pub nodes: Vec<Node>,
+    #[serde(flatten)]
+    pub other: BTreeMap<String, serde_json::Value>,
+}
+
+#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+pub struct Node {
+    pub id: usize,
+    #[serde(rename = "nodeType")]
+    pub node_type: NodeType,
+    #[serde(with = "serde_helpers::display_from_str")]
+    pub src: SourceLocation,
+    #[serde(default, skip_serializing_if = "Vec::is_empty")]
+    pub nodes: Vec<Node>,
+    #[serde(flatten)]
+    pub other: BTreeMap<String, serde_json::Value>,
+}
+
+/// Represents the source location of a node : `<start>:<length>:<index>`
+///
+/// The `length` and `index` can be -1 which is represented as `None`
+#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+pub struct SourceLocation {
+    pub start: usize,
+    pub length: Option<usize>,
+    pub index: Option<usize>,
+}
+
+impl FromStr for SourceLocation {
+    type Err = String;
+
+    fn from_str(s: &str) -> Result<Self, Self::Err> {
+        let invalid_location = move || format!("{} invalid source location", s);
+
+        let mut split = s.split(':');
+        let start = split
+            .next()
+            .ok_or_else(invalid_location)?
+            .parse::<usize>()
+            .map_err(|_| invalid_location())?;
+        let length = split
+            .next()
+            .ok_or_else(invalid_location)?
+            .parse::<isize>()
+            .map_err(|_| invalid_location())?;
+        let index = split
+            .next()
+            .ok_or_else(invalid_location)?
+            .parse::<isize>()
+            .map_err(|_| invalid_location())?;
+
+        let length = if length < 0 { None } else { Some(length as usize) };
+        let index = if index < 0 { None } else { Some(index as usize) };
+
+        Ok(Self { start, length, index })
+    }
+}
+
+impl fmt::Display for SourceLocation {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        self.start.fmt(f)?;
+        f.write_char(':')?;
+        if let Some(length) = self.length {
+            length.fmt(f)?;
+        } else {
+            f.write_str("-1")?;
+        }
+        f.write_char(':')?;
+        if let Some(index) = self.index {
+            index.fmt(f)?;
+        } else {
+            f.write_str("-1")?;
+        }
+        Ok(())
+    }
+}
+
+#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(untagged)]
+pub enum NodeType {
+    YulAssignment,
+    YulBlock,
+    YulExpressionStatement,
+    YulForLoop,
+    YulIf,
+    YulVariableDeclaration,
+    YulFunctionDefinition,
+    SourceUnit,
+    PragmaDirective,
+    ContractDefinition,
+    EventDefinition,
+    ErrorDefinition,
+    Other(String),
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn can_parse_ast() {
+        let ast = r#"
+        {
+  "absolutePath": "input.sol",
+  "exportedSymbols":
+  {
+    "Ballot":
+    [
+      2
+    ],
+    "Ballot2":
+    [
+      3
+    ],
+    "Ballot3":
+    [
+      4
+    ]
+  },
+  "id": 5,
+  "nodeType": "SourceUnit",
+  "nodes":
+  [
+    {
+      "id": 1,
+      "literals":
+      [
+        "solidity",
+        ">=",
+        "0.4",
+        ".0"
+      ],
+      "nodeType": "PragmaDirective",
+      "src": "1:24:0"
+    },
+    {
+      "abstract": false,
+      "baseContracts": [],
+      "canonicalName": "Ballot",
+      "contractDependencies": [],
+      "contractKind": "contract",
+      "fullyImplemented": true,
+      "id": 2,
+      "linearizedBaseContracts":
+      [
+        2
+      ],
+      "name": "Ballot",
+      "nameLocation": "36:6:0",
+      "nodeType": "ContractDefinition",
+      "nodes": [],
+      "scope": 5,
+      "src": "27:20:0",
+      "usedErrors": []
+    },
+    {
+      "abstract": false,
+      "baseContracts": [],
+      "canonicalName": "Ballot2",
+      "contractDependencies": [],
+      "contractKind": "contract",
+      "fullyImplemented": true,
+      "id": 3,
+      "linearizedBaseContracts":
+      [
+        3
+      ],
+      "name": "Ballot2",
+      "nameLocation": "58:7:0",
+      "nodeType": "ContractDefinition",
+      "nodes": [],
+      "scope": 5,
+      "src": "49:21:0",
+      "usedErrors": []
+    },
+    {
+      "abstract": false,
+      "baseContracts": [],
+      "canonicalName": "Ballot3",
+      "contractDependencies": [],
+      "contractKind": "contract",
+      "fullyImplemented": true,
+      "id": 4,
+      "linearizedBaseContracts":
+      [
+        4
+      ],
+      "name": "Ballot3",
+      "nameLocation": "81:7:0",
+      "nodeType": "ContractDefinition",
+      "nodes": [],
+      "scope": 5,
+      "src": "72:21:0",
+      "usedErrors": []
+    }
+  ],
+  "src": "1:92:0"
+}
+        "#;
+        let _ast: Ast = serde_json::from_str(ast).unwrap();
+
+        dbg!(serde_json::from_str::<serde_json::Value>("{}").unwrap());
+    }
+}

ethers-solc/src/artifacts/mod.rs

@@ -15,6 +15,8 @@ use crate::{compile::*, error::SolcIoError, remappings::Remapping, utils};
 
 use serde::{de::Visitor, Deserialize, Deserializer, Serialize, Serializer};
 
+pub mod ast;
+pub use ast::*;
 pub mod bytecode;
 pub mod contract;
 pub mod output_selection;
@@ -1402,8 +1404,8 @@ pub struct SecondarySourceLocation {
 #[derive(Clone, Debug, Serialize, Deserialize, Eq, PartialEq)]
 pub struct SourceFile {
     pub id: u32,
-    #[serde(default)]
-    pub ast: serde_json::Value,
+    #[serde(default, with = "serde_helpers::empty_json_object_opt")]
+    pub ast: Option<Ast>,
 }
 
 /// A wrapper type for a list of source files

ethers-solc/src/artifacts/serde_helpers.rs

@@ -3,14 +3,14 @@
 use ethers_core::types::Bytes;
 use serde::{Deserialize, Deserializer};
 
-pub fn deserialize_bytes<'de, D>(d: D) -> std::result::Result<Bytes, D::Error>
+pub fn deserialize_bytes<'de, D>(d: D) -> Result<Bytes, D::Error>
 where
     D: Deserializer<'de>,
 {
     String::deserialize(d)?.parse::<Bytes>().map_err(|e| serde::de::Error::custom(e.to_string()))
 }
 
-pub fn deserialize_opt_bytes<'de, D>(d: D) -> std::result::Result<Option<Bytes>, D::Error>
+pub fn deserialize_opt_bytes<'de, D>(d: D) -> Result<Option<Bytes>, D::Error>
 where
     D: Deserializer<'de>,
 {
@@ -70,6 +70,43 @@ pub mod json_string_opt {
     }
 }
 
+/// deserializes empty json object `{}` as `None`
+pub mod empty_json_object_opt {
+    use serde::{
+        de::{self, DeserializeOwned},
+        ser, Deserialize, Deserializer, Serialize, Serializer,
+    };
+
+    pub fn serialize<T, S>(value: &Option<T>, serializer: S) -> Result<S::Ok, S::Error>
+    where
+        S: Serializer,
+        T: Serialize,
+    {
+        if let Some(value) = value {
+            let value = serde_json::to_string(value).map_err(ser::Error::custom)?;
+            serializer.serialize_str(&value)
+        } else {
+            let empty = serde_json::Value::Object(Default::default());
+            serde_json::Value::serialize(&empty, serializer)
+        }
+    }
+
+    pub fn deserialize<'de, T, D>(deserializer: D) -> Result<Option<T>, D::Error>
+    where
+        D: Deserializer<'de>,
+        T: DeserializeOwned,
+    {
+        let json = serde_json::Value::deserialize(deserializer)?;
+        if json.is_null() {
+            return Ok(None)
+        }
+        if json.as_object().map(|obj| obj.is_empty()).unwrap_or_default() {
+            return Ok(None)
+        }
+        serde_json::from_value(json).map_err(de::Error::custom).map(Some)
+    }
+}
+
 /// serde support for string
 pub mod string_bytes {
     use serde::{Deserialize, Deserializer, Serializer};
@@ -128,6 +165,28 @@ pub mod display_from_str_opt {
     }
 }
 
+pub mod display_from_str {
+    use serde::{de, Deserialize, Deserializer, Serializer};
+    use std::{fmt, str::FromStr};
+
+    pub fn serialize<T, S>(value: &T, serializer: S) -> Result<S::Ok, S::Error>
+    where
+        T: fmt::Display,
+        S: Serializer,
+    {
+        serializer.collect_str(value)
+    }
+
+    pub fn deserialize<'de, T, D>(deserializer: D) -> Result<T, D::Error>
+    where
+        D: Deserializer<'de>,
+        T: FromStr,
+        T::Err: fmt::Display,
+    {
+        String::deserialize(deserializer)?.parse().map_err(de::Error::custom)
+    }
+}
+
 /// (De)serialize vec of tuples as map
 pub mod tuple_vec_map {
     use serde::{de::DeserializeOwned, Deserialize, Deserializer, Serialize, Serializer};

ethers-solc/src/compile/project.rs

@@ -692,9 +692,9 @@ mod tests {
         assert_eq!(state.output.sources.len(), 3);
         for (f, source) in &state.output.sources {
             if f.ends_with("A.sol") {
-                assert!(source.ast.is_object());
+                assert!(source.ast.is_some());
             } else {
-                assert!(source.ast.is_null());
+                assert!(source.ast.is_none());
             }
         }