Update todo

TODO.md

@@ -2,26 +2,107 @@
 
 ## To Do
 
-- [ ] Support GDB unit test
+- [ ] Support GDB unit test (Allowing to test C functions)
 - [ ] Support Builtin unit test
-
 - [ ] Eval filter
   - [ ] Input/Output filters
   - [ ] EvalInputFilter
   - [ ] Iter class
+- [ ] Loops and Random values in tests
+- [ ] Refactor output (cli interface)
+- [ ] Migrate to pydantic
+- [ ] Migrate from TinyKernel to RestrictedPython
 - [ ] Short test summary info
+- [ ] Implement standalone tests
 
 ## New Features
 
+### C unit tests
+
+We may use the following to capture the prototype of a function:
+
+```python
+from elftools.elf.elffile import ELFFile
+
+def extract_dwarf_info(filename):
+    with open(filename, 'rb') as f:
+        elffile = ELFFile(f)
+        if not elffile.has_dwarf_info():
+            print('No DWARF info found in', filename)
+            return
+
+        dwarf_info = elffile.get_dwarf_info()
+
+        # Parcourir les unités de compilation pour trouver les fonctions
+        for cu in dwarf_info.iter_CUs():
+            print('Found a compilation unit:', cu.get_top_DIE().attributes)
+            for die in cu.iter_DIEs():
+                if die.tag == 'DW_TAG_subprogram':
+                    name = die.attributes['DW_AT_name'].value.decode('utf-8')
+                    print('Function:', name)
+
+                    # Si l'élément DIE a des enfants (arguments de fonction)
+                    for child in die.iter_children():
+                        if child.tag == 'DW_TAG_formal_parameter':
+                            param_name = child.attributes['DW_AT_name'].value.decode('utf-8')
+                            param_type = child.attributes['DW_AT_type']
+                            print(f'Param: {param_name}, Type: {param_type}')
+
+extract_dwarf_info('foo.o')
+```
+
+```yaml
+version: 1
+tests:
+  - name: Test function add
+    function:
+      name: add
+      filename: foo.c
+      types: [int, float, str] # Would be nice to identify it automatically
+    build_cmd: gcc -flto {{filename}} runner.c # Default value
+    args: [1, 2]
+    stdout: 3
+```
+
 ### Standalone Unit test
 
 ```python
-from baygon import UnitTest
+# test_foo.py
+# Add with: 'foo.py'
+
+@baygon.stdin('2\n3\n')
+@baygon.args('2', '3')
+def test_blabla(stdout, stderr, exit_status, execution_time):
+    assert stdout == '5\n'
+    assert exit_status == 0
+
+def test_bibi(executable: Executable):
+    res = executable.run(args=['2', '3'], stdin='2\n3\n')
+    assert res.stdout == '5\n'
+    assert res.exit_status == 0
+```
+
+### Loops
+
+Using different singleton iterators. They use internal states to generate value at each call. When an iterator is used, the test is repeated for each values generated by the iterator.
+
+- `iter(min, max, step=1)`
+- `rand(min, max, n=1, seed=42, unique=True)`
+
+Variables and expressions:
+
+- `{{a=42}}` Assign value to a variable, return the variable value
+- `{{a * 2}}` Return the value of the expression
 
-class Test(UnitTest):
-    def test_version(self):
-        res = self.functions.add(2, 3)
-        self.assert_equal(res, 5)
+```yaml
+version: 1
+fixture:
+  - 'fixture.py:myfixture'
+tests:
+  - name: Test loop
+    args: [{{a=random(min=10, max=100, n=100)}}, 2]
+    stdout:
+      - contains: {{a}}
 ```
 
 ### Presentation