CG-10465 raw text edit (#170)

# Motivation
File(name="doc.mdx").edit(...) and .replace(...) should edit in place.
Right now it's appending.

# Content
New File methods.

# Testing
UT, manual testing.

# Please check the following before marking your PR as ready for review

- [x] I have added tests for my changes
- [x] I have updated the documentation or added new documentation as
needed
- [x] I have read and agree to the [Contributor License
Agreement](../CLA.md)

Co-authored-by: codegen-bot <[email protected]>

Author: tomcodgen

src/codegen/sdk/core/file.py

@@ -344,6 +344,73 @@ def parse(self, G: "CodebaseGraph") -> None:
     def _compute_dependencies(self, *args, **kwargs) -> None:
         pass
 
+    @writer
+    def edit(self, new_src: str, fix_indentation: bool = False, priority: int = 0, dedupe: bool = True) -> None:
+        """Replace the source of this file with new_src.
+
+        For non-source files, replaces the entire content. For source files, delegates to the parent
+        Editable implementation which uses TreeSitter nodes for precise editing.
+
+        Args:
+            new_src (str): The new source text to replace the current text with.
+            fix_indentation (bool): If True, adjusts the indentation of new_src to match the current
+                text's indentation level. Only applies to source files. Defaults to False.
+            priority (int): The priority of the edit transaction. Higher priority edits are
+                applied first. Defaults to 0.
+            dedupe (bool): If True, deduplicates identical transactions. Defaults to True.
+
+        Raises:
+            ValueError: If attempting to edit a binary file.
+
+        Returns:
+            None
+        """
+        if self.is_binary:
+            msg = "Cannot replace content in binary files"
+            raise ValueError(msg)
+
+        if self.ts_node is None or not isinstance(self, SourceFile):
+            self._edit_byte_range(new_src, 0, len(self.content_bytes), priority, dedupe)
+        else:
+            super().edit(new_src, fix_indentation, priority, dedupe)
+
+    @writer
+    def replace(self, old: str, new: str, count: int = -1, is_regex: bool = False, priority: int = 0) -> int:
+        """Replace occurrences of text in the file.
+
+        For non-source files, performs a direct string replacement. For source files, delegates to the
+        parent Editable implementation which uses TreeSitter nodes for precise replacements.
+
+        Args:
+            old (str): The text to be replaced.
+            new (str): The text to replace with.
+            count (int): Maximum number of replacements to make. -1 means replace all occurrences.
+                Only applies to source files. Defaults to -1.
+            is_regex (bool): If True, treat 'old' as a regular expression pattern.
+                Only applies to source files. Defaults to False.
+            priority (int): The priority of the edit transaction. Higher priority edits are
+                applied first. Defaults to 0.
+
+        Raises:
+            ValueError: If attempting to replace content in a binary file.
+
+        Returns:
+            list[Editable]: List of affected Editable objects. For non-source files, always returns
+                an empty list since they don't have Editable sub-components.
+        """
+        if self.is_binary:
+            msg = "Cannot replace content in binary files"
+            raise ValueError(msg)
+
+        if self.ts_node is None or not isinstance(self, SourceFile):
+            if old not in self.content:
+                return 0
+
+            self._edit_byte_range(self.content.replace(old, new), 0, len(self.content_bytes), priority)
+            return 1
+        else:
+            return super().replace(old, new, count, is_regex, priority)
+
 
 TImport = TypeVar("TImport", bound="Import")
 TFunction = TypeVar("TFunction", bound="Function")

tests/unit/codegen/sdk/codebase/file/test_file.py

@@ -59,28 +59,6 @@ def test_codebase_files(tmpdir) -> None:
         assert {f for f in codebase.files(extensions=[".bin"])} == {file3}
 
 
-@pytest.mark.skip("MDX editing is broken")
-def test_codebase_edit_mdx(tmpdir) -> None:
-    """Editing MDx seems broken currently - it will just prepend to the file"""
-    with get_codebase_session(tmpdir=tmpdir, files={"file1.mdx": "# Header", "file2.tsx": "console.log('hello, world!')"}) as codebase:
-        file = codebase.get_file("file1.mdx")
-        file.edit("NEW TEXT")
-        codebase.commit()
-        file = codebase.get_file("file1.mdx")
-        assert file.content == "NEW TEXT"
-
-
-@pytest.mark.skip("MDX replacing is broken")
-def test_codebase_replace_mdx(tmpdir) -> None:
-    """Editing MDx seems broken currently - it will just prepend to the file"""
-    with get_codebase_session(tmpdir=tmpdir, files={"file1.mdx": "# Header"}) as codebase:
-        file = codebase.get_file("file1.mdx")
-        file.replace("# Header", "NEW TEXT")
-        codebase.commit()
-        file = codebase.get_file("file1.mdx")
-        assert file.content == "NEW TEXT"
-
-
 @pytest.mark.skipif(sys.platform == "darwin", reason="macOS is case-insensitive")
 def test_file_extensions_ignore_case(tmpdir) -> None:
     with get_codebase_session(tmpdir=tmpdir, files={"file1.py": "print(123)", "file2.py": "print(456)", "file3.bin": b"\x89PNG", "file4": "Hello world!"}) as codebase:

tests/unit/codegen/sdk/codebase/file/test_file_edit.py

@@ -0,0 +1,119 @@
+import pytest
+
+from codegen.sdk.codebase.factory.get_session import get_codebase_session
+from codegen.sdk.core.file import SourceFile
+
+
+def test_codebase_edit_mdx(tmpdir) -> None:
+    """Test editing MDX file content"""
+    with get_codebase_session(tmpdir=tmpdir, files={"file1.mdx": "# Header", "file2.tsx": "console.log('hello, world!')"}) as codebase:
+        file = codebase.get_file("file1.mdx")
+        file.edit("NEW TEXT")
+        codebase.commit()
+        file = codebase.get_file("file1.mdx")
+        assert file.content == "NEW TEXT"
+
+
+def test_edit_json_file(tmpdir) -> None:
+    """Test editing JSON file content"""
+    with get_codebase_session(tmpdir=tmpdir, files={"config.json": '{"key": "value", "nested": {"foo": "bar"}}'}) as codebase:
+        file = codebase.get_file("config.json")
+
+        # Test complete content replacement
+        file.edit('{"newKey": "newValue"}')
+        codebase.commit()
+        assert file.content == '{"newKey": "newValue"}'
+
+        # Test partial content replacement
+        file.edit('{"newKey": "newValue", "extra": true}')
+        codebase.commit()
+        assert file.content == '{"newKey": "newValue", "extra": true}'
+
+
+def test_edit_txt_file(tmpdir) -> None:
+    """Test editing plain text file content"""
+    with get_codebase_session(tmpdir=tmpdir, files={"data.txt": "Hello\nWorld\nTest"}) as codebase:
+        file = codebase.get_file("data.txt")
+
+        # Test single line replacement
+        file.edit("New World")
+        codebase.commit()
+        assert file.content == "New World"
+
+        # Test multiline content
+        file.edit("Line 1\nLine 2\nLine 3")
+        codebase.commit()
+        assert file.content == "Line 1\nLine 2\nLine 3"
+
+
+def test_codebase_replace_mdx(tmpdir) -> None:
+    """Test replacing content in MDX file"""
+    with get_codebase_session(tmpdir=tmpdir, files={"file1.mdx": "# Header"}) as codebase:
+        file = codebase.get_file("file1.mdx")
+        file.replace("# Header", "NEW TEXT")
+        codebase.commit()
+        file = codebase.get_file("file1.mdx")
+        assert file.content == "NEW TEXT"
+
+
+def test_replace_non_source_file(tmpdir) -> None:
+    """Test replace functionality in non-source files"""
+    with get_codebase_session(tmpdir=tmpdir, files={"doc.mdx": "# Header\nThis is a test\nMore content", "config.json": '{"test": "value", "other": "test"}'}) as codebase:
+        # Test single replacement
+        mdx_file = codebase.get_file("doc.mdx")
+        mdx_file.replace("test", "demo")
+        codebase.commit()
+        assert mdx_file.content == "# Header\nThis is a demo\nMore content"
+
+        # Test multiple occurrences
+        json_file = codebase.get_file("config.json")
+        json_file.replace("test", "demo")
+        codebase.commit()
+        assert json_file.content == '{"demo": "value", "other": "demo"}'
+
+
+def test_edit_binary_file_fails(tmpdir) -> None:
+    """Test that editing binary files raises an error"""
+    binary_content = bytes([0x89, 0x50, 0x4E, 0x47])  # PNG header
+    with get_codebase_session(tmpdir=tmpdir, files={"image.png": binary_content}) as codebase:
+        file = codebase.get_file("image.png")
+
+        with pytest.raises(ValueError, match="Cannot replace content in binary files"):
+            file.edit("new content")
+
+        with pytest.raises(ValueError, match="Cannot replace content in binary files"):
+            file.replace("old", "new")
+
+
+def test_edit_source_file_preserves_behavior(tmpdir) -> None:
+    """Test that source files still use TreeSitter-based editing"""
+    with get_codebase_session(tmpdir=tmpdir, files={"script.py": "def test():\n    print('hello')"}) as codebase:
+        file = codebase.get_file("script.py")
+
+        # Should use TreeSitter node-based editing
+        file.edit("def test():\n    print('world')")
+        codebase.commit()
+        assert file.content == "def test():\n    print('world')"
+
+        # Verify the file is still parseable as Python
+        assert isinstance(file, SourceFile)
+        assert file.get_function("test") is not None
+
+
+def test_transaction_ordering_non_source_files(tmpdir) -> None:
+    """Test that transaction ordering works for non-source files"""
+    with get_codebase_session(tmpdir=tmpdir, files={"doc.md": "# Header\nContent\nFooter"}) as codebase:
+        file = codebase.get_file("doc.md")
+
+        # Apply changes sequentially to avoid transaction conflicts
+        file.edit("# New Header\nContent\nFooter", priority=1)
+        codebase.commit()
+
+        file.edit("# New Header\nNew Content\nFooter", priority=2)
+        codebase.commit()
+
+        file.edit("# New Header\nNew Content\nNew Footer", priority=0)
+        codebase.commit()
+
+        # Verify final content
+        assert file.content == "# New Header\nNew Content\nNew Footer"