diff --git a/README.md b/README.md
index eceaf2e..221ab33 100644
--- a/README.md
+++ b/README.md
@@ -1,3 +1,11 @@
 # lazy-import
 
-python with no imports
\ No newline at end of file
+python with no imports
+
+hell ya it's got side effects
+
+## TODO
+
+- oh frick we actually need to conert to a NodeTransformer so that 
+  we can control the rendering of child nodes (aka we can recursively
+  get the names in the namespace rather than having an entry-only visitor)
\ No newline at end of file
diff --git a/pdm.lock b/pdm.lock
new file mode 100644
index 0000000..3c1efd5
--- /dev/null
+++ b/pdm.lock
@@ -0,0 +1,11 @@
+# This file is @generated by PDM.
+# It is not intended for manual editing.
+
+[metadata]
+groups = ["default"]
+strategy = ["inherit_metadata"]
+lock_version = "4.5.0"
+content_hash = "sha256:6e3746c7427353cd94cfd6c1a5a6309fdfd871714e3b1213ded2befc6ef522a0"
+
+[[metadata.targets]]
+requires_python = ">=3.10"
diff --git a/pyproject.toml b/pyproject.toml
new file mode 100644
index 0000000..9e58a6c
--- /dev/null
+++ b/pyproject.toml
@@ -0,0 +1,19 @@
+[project]
+name = "lazy-import"
+version = "0.1.0"
+description = "python with no imports"
+authors = [
+    {name = "sneakers-the-rat", email = "sneakers-the-rat@protonmail.com"},
+]
+dependencies = []
+requires-python = ">=3.10"
+readme = "README.md"
+license = {text = "EUPL-1.2"}
+
+[build-system]
+requires = ["pdm-backend"]
+build-backend = "pdm.backend"
+
+
+[tool.pdm]
+distribution = true
diff --git a/src/lazy_import/__init__.py b/src/lazy_import/__init__.py
new file mode 100644
index 0000000..8b969c5
--- /dev/null
+++ b/src/lazy_import/__init__.py
@@ -0,0 +1,4 @@
+import sys
+
+def getattr(name):
+    pass
\ No newline at end of file
diff --git a/src/lazy_import/ast.py b/src/lazy_import/ast.py
new file mode 100644
index 0000000..b78106f
--- /dev/null
+++ b/src/lazy_import/ast.py
@@ -0,0 +1,123 @@
+import ast
+from collections import ChainMap
+from dataclasses import dataclass, field
+from pathlib import Path
+
+@dataclass
+class Name:
+
+    module: str
+    name: str | None = None
+    aliases: set = field(default_factory=set)
+
+    def __contains__(self, item: str):
+        return item in self.aliases or (
+            item == self.module
+            if self.name is None else
+            item == self.name
+        )
+
+    @classmethod
+    def from_ast_name(cls, name: ast.Name) -> 'Name':
+        return cls.from_str(name.id)
+
+    @classmethod
+    def from_str(cls, name: str) -> 'Name':
+        if len(name_parts := name.rsplit(".")) > 1:
+            return Name(module=name_parts[0], name=name_parts[1])
+        else:
+            return Name(module=name)
+
+@dataclass
+class NameCollection:
+
+    names: list[Name] = field(default_factory=set)
+
+    def add(self, new_name: ast.Name):
+        for name in self.names:
+            if new_name.id in name:
+                name.aliases.add(new_name.id)
+                return
+        self.names.append(Name.from_ast_name(new_name))
+
+class NameVisitor(ast.NodeVisitor):
+    """
+    Extract names to import and assign from an importless python module
+    """
+
+    def __init__(self):
+        self.real_names = ChainMap()
+        self.fake_names = NameCollection()
+
+    def pop_ctx(self):
+        self.real_names = self.real_names.parents
+
+    def push_ctx(self):
+        self.real_names = self.real_names.new_child()
+
+    def visit_Import(self, node: ast.Import | ast.ImportFrom) -> None:
+        """Add to names"""
+        for alias in node.names:
+            name = alias.asname if alias.asname else alias.name
+            self.real_names[name] = node
+
+    def visit_ImportFrom(self, node: ast.ImportFrom) -> None:
+        """Add to names"""
+        self.visit_Import(node)
+
+
+    def visit_Name(self, node: ast.Name):
+        """Either add to real names or fake names depending on ctx"""
+        if node.ctx == ast.Store():
+            self.real_names[node.id] = node
+        elif node.ctx == ast.Load() and node.id not in self.real_names:
+            self.fake_names.add(node)
+        elif node.ctx == ast.Del() and node.id in self.real_names:
+            del self.real_names[node.id]
+        else:  # pragma: no cover
+            if node.ctx not in (ast.Del(), ast.Store(), ast.Load()):
+                raise ValueError('How did this happen!? wrong node ctx type?')
+
+    def visit_Attribute(self, node: ast.Attribute):
+        attr_name = flatten_attribute(node)
+        if node.ctx == ast.Load():
+            if attr_name not in self.real_names:
+                self.fake_names.add(Name.from_str(attr_name))
+        elif node.ctx == ast.Store():
+            self.real_names[attr_name] = node
+        elif node.ctx == ast.Del() and attr_name in self.real_names:
+            del self.real_names[attr_name]
+        else:  # pragma: no cover
+            if node.ctx not in (ast.Del(), ast.Store(), ast.Load()):
+                raise ValueError('How did this happen!? wrong node ctx type?')
+
+    def visit_FunctionDef(self, node):
+        """push context"""
+
+        self.push_ctx()
+        args = node.args
+        for arg in args.args:
+            self.real_names[arg.arg] = arg
+        if args.vararg:
+            self.real_names[args.vararg.arg] = args.vararg
+        if args.kwarg:
+            self.real_names[args.kwarg.arg] = args.kwarg
+
+    def visit_AsyncFunctionDef(self, node):
+        """push context"""
+        self.push_ctx()
+
+    def visit_ClassDef(self, node):
+        """push context"""
+        self.push_ctx()
+
+    def visit_Return(self, node):
+        """pop context"""
+        self.pop_ctx()
+
+
+def flatten_attribute(attr: ast.Attribute) -> str:
+    if isinstance(attr.value, ast.Attribute):
+        return '.'.join([flatten_attribute(attr.value), attr.attr])
+    elif isinstance(attr.value, ast.Name):
+        return '.'.join([attr.value.id, attr.attr])
diff --git a/src/lazy_import/importer.py b/src/lazy_import/importer.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/__init__.py b/tests/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/conftest.py b/tests/conftest.py
new file mode 100644
index 0000000..250228f
--- /dev/null
+++ b/tests/conftest.py
@@ -0,0 +1,3 @@
+import lazy_import
+
+__all__ = ['lazy_import']
\ No newline at end of file
diff --git a/tests/data/__init__.py b/tests/data/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/data/input_file.py b/tests/data/input_file.py
new file mode 100644
index 0000000..92a8510
--- /dev/null
+++ b/tests/data/input_file.py
@@ -0,0 +1,51 @@
+import lazy_import
+import collections.abc
+import json as jay_son
+from typing import List
+from collections.abc import Callable
+from collections.abc import ChainMap as cm
+
+mod_variable = 10
+
+def imports_are_lazy():
+    ast
+    typing.List
+    re.match('hell ya', 'hell ya')
+    # one day you should be able to do this
+    # numpy as np
+
+    def a_function(a: typing.Iterable) -> importlib.abc.Finder:
+        a = 1
+
+    class AClass():
+        zz = 1
+        yy = array.array()
+
+        def __init__(self):
+            _ = base64.b64decode(f"{binascii.hexlify(b'abc')}")
+
+def regular_names_still_work():
+    # assert mod_variable == 10
+    x = 20
+    # assert x == 20
+
+    def afunc(y, z=1, *args, **kwargs):
+        z = 10
+        assert z == 10
+        assert y == 30
+
+    class AClass():
+        z = 1
+
+
+    afunc(30)
+
+def test_names_are_lazy():
+    """
+    you can just use the last unique segment
+    """
+    _ = numpy.random.random(100)
+    _ = random
+
+    assert random is numpy.random.random
+
diff --git a/tests/test_ast.py b/tests/test_ast.py
new file mode 100644
index 0000000..c1df1e3
--- /dev/null
+++ b/tests/test_ast.py
@@ -0,0 +1 @@
+def test
\ No newline at end of file
diff --git a/tests/test_lazy_import.py b/tests/test_lazy_import.py
new file mode 100644
index 0000000..f017afa
--- /dev/null
+++ b/tests/test_lazy_import.py
@@ -0,0 +1,33 @@
+import lazy_import
+from typing import List
+import collections.abc
+from collections.abc import Callable
+from collections.abc import ChainMap as cm
+
+mod_variable = 10
+
+def test_imports_are_lazy():
+    re.match('hell ya', 'hell ya')
+    typing.List
+
+def test_regular_names_still_work():
+    assert mod_variable == 10
+    x = 20
+    assert x == 20
+
+    def afunc(y):
+        z = 10
+        assert z == 10
+        assert y == 30
+
+    afunc(30)
+
+def test_names_are_lazy():
+    """
+    you can just use the last unique segment
+    """
+    _ = numpy.random.random(100)
+    _ = random
+
+    assert random is numpy.random.random
+