From e9970edd775a617226007d3d3cb0fbbc39ed3948 Mon Sep 17 00:00:00 2001
From: Gyubong <jopemachine@naver.com>
Date: Tue, 23 Aug 2022 21:45:01 +0900
Subject: [PATCH] Add built-in `breakpoint` function (#4122)

---
 Lib/test/test_builtin.py  |  5 ++-
 vm/src/stdlib/builtins.rs | 13 ++++++--
 vm/src/stdlib/sys.rs      | 67 ++++++++++++++++++++++++++++++++++++++-
 3 files changed, 80 insertions(+), 5 deletions(-)

diff --git a/Lib/test/test_builtin.py b/Lib/test/test_builtin.py
index 92a60af70..835e7f0f8 100644
--- a/Lib/test/test_builtin.py
+++ b/Lib/test/test_builtin.py
@@ -1900,7 +1900,6 @@ class BuiltinTest(unittest.TestCase):
             self.assertFalse(not NotImplemented)
 
 
-@unittest.skip("TODO: RUSTPYTHON, AttributeError: module 'sys' has no attribute '__breakpointhook__'")
 class TestBreakpoint(unittest.TestCase):
     def setUp(self):
         # These tests require a clean slate environment.  For example, if the
@@ -2005,6 +2004,10 @@ class TestBreakpoint(unittest.TestCase):
             breakpoint()
             mock.assert_not_called()
 
+    def test_runtime_error_when_hook_is_lost(self):
+        del sys.breakpointhook
+        with self.assertRaises(RuntimeError):
+            breakpoint()
 
 @unittest.skipUnless(pty, "the pty and signal modules must be available")
 class PtyTests(unittest.TestCase):
diff --git a/vm/src/stdlib/builtins.rs b/vm/src/stdlib/builtins.rs
index 5ea84a0a0..d8c4aa7ae 100644
--- a/vm/src/stdlib/builtins.rs
+++ b/vm/src/stdlib/builtins.rs
@@ -80,8 +80,6 @@ mod builtins {
         }
     }
 
-    // builtin_breakpoint
-
     #[pyfunction]
     fn callable(obj: PyObjectRef, vm: &VirtualMachine) -> bool {
         vm.is_callable(&obj)
@@ -363,7 +361,16 @@ mod builtins {
         obj.hash(vm)
     }
 
-    // builtin_help
+    #[pyfunction]
+    fn breakpoint(args: FuncArgs, vm: &VirtualMachine) -> PyResult {
+        match vm
+            .sys_module
+            .get_attr(vm.ctx.intern_str("breakpointhook"), vm)
+        {
+            Ok(hook) => vm.invoke(hook.as_ref(), args),
+            Err(_) => Err(vm.new_runtime_error("lost sys.breakpointhook".to_owned())),
+        }
+    }
 
     #[pyfunction]
     fn hex(number: PyIntRef) -> String {
diff --git a/vm/src/stdlib/sys.rs b/vm/src/stdlib/sys.rs
index ec56ec926..dc6ddc2b0 100644
--- a/vm/src/stdlib/sys.rs
+++ b/vm/src/stdlib/sys.rs
@@ -13,13 +13,18 @@ mod sys {
         frame::FrameRef,
         function::{FuncArgs, OptionalArg, PosArgs},
         stdlib::builtins,
+        stdlib::warnings::warn,
         types::PyStructSequence,
         version,
         vm::{Settings, VirtualMachine},
         AsObject, PyObjectRef, PyRef, PyRefExact, PyResult,
     };
     use num_traits::ToPrimitive;
-    use std::{env, mem, path, sync::atomic::Ordering};
+    use std::{
+        env::{self, VarError},
+        mem, path,
+        sync::atomic::Ordering,
+    };
 
     // not the same as CPython (e.g. rust's x86_x64-unknown-linux-gnu is just x86_64-linux-gnu)
     // but hopefully that's just an implementation detail? TODO: copy CPython's multiarch exactly,
@@ -316,6 +321,66 @@ mod sys {
         vm.write_exception(&mut crate::py_io::PyWriter(stderr, vm), &exc)
     }
 
+    #[pyfunction(name = "__breakpointhook__")]
+    #[pyfunction]
+    pub fn breakpointhook(args: FuncArgs, vm: &VirtualMachine) -> PyResult {
+        let envar = std::env::var("PYTHONBREAKPOINT")
+            .and_then(|envar| {
+                if envar.is_empty() {
+                    Err(VarError::NotPresent)
+                } else {
+                    Ok(envar)
+                }
+            })
+            .unwrap_or_else(|_| "pdb.set_trace".to_owned());
+
+        if envar.eq("0") {
+            return Ok(vm.ctx.none());
+        };
+
+        let print_unimportable_module_warn = || {
+            warn(
+                vm.ctx.exceptions.runtime_warning,
+                format!(
+                    "Ignoring unimportable $PYTHONBREAKPOINT: \"{}\"",
+                    envar.to_owned(),
+                ),
+                0,
+                vm,
+            )
+            .unwrap();
+            Ok(vm.ctx.none())
+        };
+
+        let last = match envar.split('.').last() {
+            Some(last) => last,
+            None => {
+                return print_unimportable_module_warn();
+            }
+        };
+
+        let (modulepath, attrname) = if last.eq(&envar) {
+            ("builtins".to_owned(), envar.to_owned())
+        } else {
+            (
+                envar[..(envar.len() - last.len() - 1)].to_owned(),
+                last.to_owned(),
+            )
+        };
+
+        let module = match vm.import(modulepath, None, 0) {
+            Ok(module) => module,
+            Err(_) => {
+                return print_unimportable_module_warn();
+            }
+        };
+
+        match vm.get_attribute_opt(module, attrname) {
+            Ok(Some(hook)) => vm.invoke(hook.as_ref(), args),
+            _ => print_unimportable_module_warn(),
+        }
+    }
+
     #[pyfunction]
     fn exc_info(vm: &VirtualMachine) -> (PyObjectRef, PyObjectRef, PyObjectRef) {
         match vm.topmost_exception() {