chore(tests): heavy integration to use large dataset

Jon-Becker · Jun 6, 2024 · fae2273 · fae2273
1 parent 773acc6
commit fae2273
Show file tree

Hide file tree

Showing 3 changed files with 103 additions and 111 deletions.
diff --git a/.github/workflows/heavy-integration.yml b/.github/workflows/heavy-integration.yml
@@ -1,45 +1,52 @@
 name: heavy integration
 
 on:
-    schedule:
-        # Runs at 10PM utc
-        - cron: "0 22 * * *"
-    workflow_dispatch:
+  schedule:
+    # Runs at 10PM utc
+    - cron: "0 22 * * *"
+  workflow_dispatch:
 
 concurrency:
-    group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
-    cancel-in-progress: true
+  group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
+  cancel-in-progress: true
 
 env:
-    CARGO_TERM_COLOR: always
+  CARGO_TERM_COLOR: always
 
 jobs:
-    heavy-integration:
-        name: heavy (long-running) integration tests
-        runs-on: ubuntu-latest
-        timeout-minutes: 120
+  heavy-integration:
+    name: heavy (long-running) integration tests
+    runs-on: ubuntu-latest
+    timeout-minutes: 120
 
-        steps:
-            - uses: actions/checkout@v3
-            - uses: dtolnay/rust-toolchain@stable
-            - uses: taiki-e/install-action@nextest
-            - name: Run Tests
-              run: |
-                  cargo nextest r --no-fail-fast --release --nocapture -- --ignored
+    steps:
+      - uses: actions/checkout@v3
+      - uses: dtolnay/rust-toolchain@stable
+      - uses: taiki-e/install-action@nextest
+      - name: Fetch Dataset
+        run: |
+          # download from https://jbecker.dev/data/largest1k.tar.gz
+          wget https://jbecker.dev/data/largest1k.tar.gz
 
-    # If any of the jobs fail, this will create a high-priority issue to signal so.
-    issue:
-        name: Open an issue
-        runs-on: ubuntu-latest
-        needs: heavy-integration
-        if: ${{ failure() }}
-        steps:
-            - uses: actions/checkout@v4
-            - uses: JasonEtco/create-an-issue@v2
-              env:
-                  GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-                  WORKFLOW_URL: |
-                      ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}
-              with:
-                  update_existing: true
-                  filename: .github/INTEGRATION_FAILURE.md
+          # extract the dataset
+          tar -xvf largest1k.tar.gz
+      - name: Run Tests
+        run: |
+          cargo nextest r --no-fail-fast --release --nocapture -- --ignored
+
+  # If any of the jobs fail, this will create a high-priority issue to signal so.
+  issue:
+    name: Open an issue
+    runs-on: ubuntu-latest
+    needs: heavy-integration
+    if: ${{ failure() }}
+    steps:
+      - uses: actions/checkout@v4
+      - uses: JasonEtco/create-an-issue@v2
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          WORKFLOW_URL: |
+            ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}
+        with:
+          update_existing: true
+          filename: .github/INTEGRATION_FAILURE.md
diff --git a/.gitignore b/.gitignore
@@ -34,3 +34,5 @@ false/*
 
 *.svg
 *.sh
+
+largest1k
diff --git a/crates/core/tests/test_decompile.rs b/crates/core/tests/test_decompile.rs
@@ -1,8 +1,11 @@
 #[cfg(test)]
 mod integration_tests {
+    use std::{path::PathBuf, process::Command};
+
     use alloy_json_abi::JsonAbi;
     use heimdall_common::utils::io::file::delete_path;
-    use heimdall_decompiler::{decompile, DecompilerArgs};
+    use heimdall_decompiler::{decompile, DecompilerArgs, DecompilerArgsBuilder};
+    use serde_json::Value;
 
     #[tokio::test]
     async fn test_decompile_precompile() {
@@ -172,91 +175,71 @@ mod integration_tests {
     #[tokio::test]
     #[ignore]
     async fn heavy_test_decompile_thorough() {
-        let rpc_url = std::env::var("RPC_URL").unwrap_or_else(|_| {
-            println!("RPC_URL not set, skipping test");
+        let root_dir = PathBuf::from(env!("CARGO_MANIFEST_DIR"))
+            .parent()
+            .expect("no parent")
+            .parent()
+            .expect("no parent")
+            .to_owned();
+
+        // if the ./largest1k directory does not exist, download it from https://jbecker.dev/data/largest1k.tar.gz
+        let dataset_dir = root_dir.join("largest1k");
+        if !dataset_dir.exists() {
+            eprintln!("dataset not found in root, skipping test");
             std::process::exit(0);
-        });
+        }
 
-        let contracts = [
-            "0xdAC17F958D2ee523a2206206994597C13D831ec7",
-            "0x3fC91A3afd70395Cd496C647d5a6CC9D4B2b7FAD",
-            "0x7a250d5630B4cF539739dF2C5dAcb4c659F2488D",
-            "0xA3C66393049fAB4830C330Dfe658f94A4de0A122",
-            "0xA0b86991c6218b36c1d19D4a2e9Eb0cE3606eB48",
-            "0x32400084C286CF3E17e7B677ea9583e60a000324",
-            "0x00000000000000ADc04C56Bf30aC9d3c0aAF14dC",
-            "0xC02aaA39b223FE8D0A0e5C4F27eAD9083C756Cc2",
-            "0x881D40237659C251811CEC9c364ef91dC08D300C",
-            "0x1111111254EEB25477B68fb85Ed929f73A960582",
-            "0x6b75d8AF000000e20B7a7DDf000Ba900b4009A80",
-            "0xDef1C0ded9bec7F1a1670819833240f027b25EfF",
-            "0xaBEA9132b05A70803a4E85094fD0e1800777fBEF",
-            "0x6B175474E89094C44Da98b954EedeAC495271d0F",
-            "0xae0Ee0A63A2cE6BaeEFFE56e7714FB4EFE48D419",
-            "0x1a0ad011913A150f69f6A19DF447A0CfD9551054",
-            "0x29469395eAf6f95920E59F858042f0e28D98a20B",
-            "0xA69babEF1cA67A37Ffaf7a485DfFF3382056e78C",
-            "0x95aD61b0a150d79219dCF64E1E6Cc01f0B64C4cE",
-            "0xA9D1e08C7793af67e9d92fe308d5697FB81d3E43",
-            "0x3dB52cE065f728011Ac6732222270b3F2360d919",
-            "0x000000000000Ad05Ccc4F10045630fb830B95127",
-            "0x253553366Da8546fC250F225fe3d25d0C782303b",
-            "0x65f2F6Fba44e5AbeFD90C2aEE52B11a243FC7A16",
-            "0x1f9840a85d5aF5bf1D1762F925BDADdC4201F984",
-            "0x7D1AfA7B718fb893dB30A3aBc0Cfc608AaCfeBB0",
-            "0xb0fcB43D3701f00aFD2Fb3d5f577a806F551D2F2",
-            "0x0000000000A39bb272e79075ade125fd351887Ac",
-            "0xEf1c6E67703c7BD7107eed8303Fbe6EC2554BF6B",
-            "0x68b3465833fb72A70ecDF485E0e4C7bD8665Fc45",
-            "0x98C3d3183C4b8A650614ad179A1a98be0a8d6B8E",
-            "0x514910771AF9Ca656af840dff83E8264EcF986CA",
-            "0x06450dEe7FD2Fb8E39061434BAbCFC05599a6Fb8",
-            "0x6982508145454Ce325dDbE47a25d4ec3d2311933",
-            "0x2a3DD3EB832aF982ec71669E178424b10Dca2EDe",
-            "0xa24787320ede4CC19D800bf87B41Ab9539c4dA9D",
-            "0x473037de59cf9484632f4A27B509CFE8d4a31404",
-            "0xFD14567eaf9ba941cB8c8a94eEC14831ca7fD1b4",
-            "0x4d224452801ACEd8B2F0aebE155379bb5D594381",
-            "0xDe30da39c46104798bB5aA3fe8B9e0e1F348163F",
-            "0x7Fc66500c84A76Ad7e9c93437bFc5Ac33E2DDaE9",
-            "0x388C818CA8B9251b393131C08a736A67ccB19297",
-            "0x3999D2c5207C06BBC5cf8A6bEa52966cabB76d41",
-            "0x3b3ae790Df4F312e745D270119c6052904FB6790",
-            "0xB517850510997a34b4DdC8c3797B4F83fAd510c4",
-            "0x902F09715B6303d4173037652FA7377e5b98089E",
-            "0x5a54fe5234E811466D5366846283323c954310B2",
-            "0xd1d2Eb1B1e90B638588728b4130137D262C87cae",
-            "0x95e05e2Abbd26943874ac000D87C3D9e115B543c",
-            "0x00000000A991C429eE2Ec6df19d40fe0c80088B8",
-            "0x737673b5e0a3c68adf4c1a87bca5623cfc537ec3",
-            "0x940259178FbF021e625510919BC2FF0B944E5613",
-            "0xff612db0583be8d5498731e4e32bc12e08fa6292",
-            "0xd5FEa30Ed719693Ec8848Dc7501b582F5de6a5BB",
-            "0x4C727a07246A70862e45B2E58fcd82c0eD5Eda85",
-            "0x9baa53dD2aB408D9135e549831C06E5c6407bF1d",
-        ];
+        // list files in root_dir
+        let contracts = std::fs::read_dir(dataset_dir)
+            .expect("failed to read dataset directory")
+            .map(|res| {
+                // HashMap from filename (without extension) to bytecode (from serde_json::Value)
+                res.map(|e| {
+                    let path = e.path();
+                    let filename = path
+                        .file_stem()
+                        .expect("no file stem")
+                        .to_str()
+                        .expect("no file stem")
+                        .to_owned();
+
+                    // read contents as json and parse to serde_json::Value
+                    let contents_json: Value = serde_json::from_str(
+                        &std::fs::read_to_string(path).expect("failed to read file"),
+                    )
+                    .expect("failed to parse json");
+                    let bytecode = contents_json["code"].as_str().expect("no bytecode").to_owned();
+
+                    (filename, bytecode)
+                })
+            })
+            .collect::<Result<Vec<_>, std::io::Error>>()
+            .expect("failed to collect files");
 
         // define flag checks
         let mut is_function_covered = false;
         let mut is_event_covered = false;
         let mut is_require_covered = false;
         let mut is_error_covered = false;
 
-        for contract in contracts {
-            println!("Testing contract: {contract}");
-            let result = decompile(DecompilerArgs {
-                target: contract.to_string(),
-                rpc_url: rpc_url.to_string(),
-                default: true,
-                skip_resolving: true,
-                include_solidity: true,
-                include_yul: false,
-                output: String::from(""),
-                name: String::from(""),
-                timeout: 10000,
-            })
-            .await
-            .expect("failed to decompile");
+        for (contract_address, bytecode) in contracts {
+            println!("Testing contract: {contract_address}");
+            let args = DecompilerArgsBuilder::new()
+                .target(bytecode)
+                .skip_resolving(true)
+                .include_solidity(true)
+                .timeout(10000)
+                .output(String::from("./output/tests/decompile/integration"))
+                .build()
+                .expect("failed to build args");
+
+            let result = decompile(args)
+                .await
+                .map_err(|e| {
+                    eprintln!("failed to decompile {contract_address}: {e}");
+                    e
+                })
+                .expect("failed to decompile");
 
             // assert that the number of opening and closing brackets, parentheses, and curly braces
             // are equal