Fix callseq deduplication error, allow using Some|None=None args with no commutativity error, add ignore_basictypes_return for a ConversionPoint.from_fn

README.md

@@ -18,9 +18,11 @@
 - Поддерживает асинхронный контекст
 - Поддерживает внедрение зависимости через синхронные/асинхронные менеджеры контекста
 - Поддерживает `Union`-типы в зависимостях
+- Учитывает default-параметры
+- Позволяет выстраивать конвейеры преобразований
+- Опционально разворачивает кортежи в возвращаемых значениях
 
 #### Ограничения библиотеки:  
-- Зависимости со стандартными параметрами пока не поддерживаются
 - Выбор графа преобразований вызывает комбинаторный взрыв
 - Кэширование графов преобразований не поддерживается
 - При некоммутативности сгенерированного графа, имеется опасность неконсистентного выбора пути, поскольку порядок обхода методов, а также графа, не гарантирован
@@ -103,6 +105,49 @@ assert tst == 1
 
 ```
 
+----
+
+#### Сборка конвейеров преобразований:
+
+Пусть, имеется несколько методов-потребителей, которые необходимо вызывать последовательно:  
+
+```python
+
+from breakshaft.convertor import ConvRepo
+
+repo = ConvRepo()
+
+# Объявляем A и B, а также методы преобразований - как в прошлом примере
+
+type cons2ret = str  # избегаем использования builtin-типов, чтобы избежать простых коллизий 
+
+
+def consumer1(dep: A) -> B:
+    return B(float(42))
+
+
+def consumer2(dep: B) -> cons2ret:
+    return str(dep.b)
+
+
+def consumer3(dep: cons2ret) -> int:
+    return int(float(dep))
+
+
+pipeline = repo.create_pipeline(
+    (B,),
+    [consumer1, consumer2, consumer3],
+    force_commutative=True,
+    allow_sync=True,
+    allow_async=False,
+    force_async=False
+)
+
+dat = pipeline(B(42))
+assert dat == 42
+```
+
+
 ----
 
 #### Как получить граф преобразований:

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "breakshaft"
-version = "0.1.0.post2"
+version = "0.1.6.post5"
 description = "Library for in-time codegen for type conversion"
 authors = [
     { name = "nikto_b", email = "niktob560@yandex.ru" }

src/breakshaft/convertor.py

@@ -1,9 +1,12 @@
 from __future__ import annotations
-from typing import Optional, Callable, Unpack, TypeVarTuple, TypeVar, Awaitable, Any
+
+import collections.abc
+from typing import Optional, Callable, Unpack, TypeVarTuple, TypeVar, Awaitable, Any, Sequence, Iterable
 
 from .graph_walker import GraphWalker
 from .models import ConversionPoint, Callgraph
 from .renderer import ConvertorRenderer, InTimeGenerationConvertorRenderer
+from .util import extract_return_type, universal_qualname
 
 Tin = TypeVarTuple('Tin')
 Tout = TypeVar('Tout')
@@ -14,10 +17,14 @@ class ConvRepo:
 
     walker: GraphWalker
     renderer: ConvertorRenderer
+    store_callseq: bool
+    store_sources: bool
 
     def __init__(self,
                  graph_walker: Optional[GraphWalker] = None,
-                 renderer: Optional[ConvertorRenderer] = None, ):
+                 renderer: Optional[ConvertorRenderer] = None,
+                 store_callseq: bool = False,
+                 store_sources: bool = False):
         if graph_walker is None:
             graph_walker = GraphWalker()
         if renderer is None:
@@ -26,13 +33,60 @@ class ConvRepo:
         self._convertor_set = set()
         self.walker = graph_walker
         self.renderer = renderer
+        self.store_callseq = store_callseq
+        self.store_sources = store_sources
+
+    def create_pipeline(self,
+                        from_types: Sequence[type],
+                        fns: Sequence[Callable | Iterable[ConversionPoint] | ConversionPoint],
+                        force_commutative: bool = True,
+                        allow_async: bool = True,
+                        allow_sync: bool = True,
+                        force_async: bool = False
+                        ):
+        filtered_injectors = self.filtered_injectors(allow_async, allow_sync)
+        pipeline_callseq = []
+        orig_from_types = tuple(from_types)
+        from_types = tuple(from_types)
+
+        for fn in fns:
+            injects = None
+            if isinstance(fn, collections.abc.Iterable):
+                for f in fn:
+                    injects = f.injects
+                    break
+            elif isinstance(fn, ConversionPoint):
+                injects = fn.injects
+            else:
+                injects = extract_return_type(fn)
+
+            callseq = self.get_callseq(filtered_injectors, frozenset(from_types), fn, force_commutative)
+
+            pipeline_callseq += callseq
+
+            if injects is not None:
+                from_types += (injects,)
+
+        ret_fn = self.renderer.render(orig_from_types,
+                                      pipeline_callseq,
+                                      force_async=force_async,
+                                      store_sources=self.store_sources)
+        if self.store_callseq:
+            setattr(ret_fn, '__breakshaft_callseq__', pipeline_callseq)
+        return ret_fn
 
     @property
     def convertor_set(self):
         return self._convertor_set
 
-    def add_injector(self, func: Callable, rettype: Optional[type] = None):
+    def add_conversion_points(self, conversion_points: Iterable[ConversionPoint]):
-        self._convertor_set |= set(ConversionPoint.from_fn(func, rettype=rettype))
+        self._convertor_set |= set(conversion_points)
+
+    def add_injector(self,
+                     func: Callable,
+                     rettype: Optional[type] = None,
+                     type_remap: Optional[dict[str, type]] = None):
+        self.add_conversion_points(ConversionPoint.from_fn(func, rettype=rettype, type_remap=type_remap))
 
     def _callseq_from_callgraph(self, cg: Callgraph) -> list[ConversionPoint]:
         if len(cg.variants) == 0:
@@ -46,30 +100,27 @@ class ConvRepo:
         ret += [variant.injector]
         return ret
 
-    def get_conversion(self,
+    def filtered_injectors(self, allow_async: bool, allow_sync: bool) -> frozenset[ConversionPoint]:
-                       from_types: tuple[type[Unpack[Tin]]],
+        filtered_injectors: frozenset[ConversionPoint] = frozenset()
-                       fn: Callable[..., Tout],
+        for inj in self.convertor_set:
-                       force_commutative: bool = True,
+            if inj.is_async and not allow_async:
-                       allow_async: bool = True,
+                continue
-                       allow_sync: bool = True,
+            if not inj.is_async and not allow_sync:
-                       force_async: bool = False
+                continue
-                       ) -> Callable[[Unpack[Tin]], Tout] | Awaitable[Callable[[Unpack[Tin]], Tout]]:
+            filtered_injectors |= {inj}
-        if not allow_async or force_async:
+        return filtered_injectors
-            filtered_injectors: frozenset[ConversionPoint] = frozenset()
-            for inj in self.convertor_set:
-                if inj.is_async and not allow_async:
-                    continue
-                if not inj.is_async and not allow_sync:
-                    continue
-                filtered_injectors |= {inj}
-        else:
-            filtered_injectors = frozenset(self.convertor_set)
 
-        cg = self.walker.generate_callgraph(filtered_injectors, frozenset(from_types), fn)
+    def get_callseq(self,
+                    injectors: frozenset[ConversionPoint],
+                    from_types: frozenset[type],
+                    fn: Callable | Iterable[ConversionPoint] | ConversionPoint,
+                    force_commutative: bool) -> list[ConversionPoint]:
+
+        cg = self.walker.generate_callgraph(injectors, from_types, fn)
         if cg is None:
-            raise ValueError(f'Unable to compute conversion graph on {from_types}->{fn.__qualname__}')
+            raise ValueError(f'Unable to compute conversion graph on {from_types}->{universal_qualname(fn)}')
 
-        exploded = self.walker.explode_callgraph_branches(cg, frozenset(from_types))
+        exploded = self.walker.explode_callgraph_branches(cg, from_types)
 
         selected = self.walker.filter_exploded_callgraph_branch(exploded)
         if len(selected) == 0:
@@ -79,17 +130,51 @@ class ConvRepo:
             raise ValueError('Conversion path is not commutative')
 
         callseq = self._callseq_from_callgraph(Callgraph(frozenset([selected[0]])))
-        return self.renderer.render(from_types, callseq, force_async=force_async)
 
-    def mark_injector(self, *, rettype: Optional[type] = None):
+        if len(callseq) > 0:
+            injects = None
+            if isinstance(fn, collections.abc.Iterable):
+                for f in fn:
+                    injects = f.injects
+                    break
+            elif isinstance(fn, ConversionPoint):
+                injects = fn.injects
+            else:
+                injects = extract_return_type(fn)
+            callseq[-1] = callseq[-1].copy_with(injects=injects)
+
+        return callseq
+
+    def get_conversion(self,
+                       from_types: Sequence[type[Unpack[Tin]]],
+                       fn: Callable[..., Tout] | Iterable[ConversionPoint] | ConversionPoint,
+                       force_commutative: bool = True,
+                       allow_async: bool = True,
+                       allow_sync: bool = True,
+                       force_async: bool = False
+                       ) -> Callable[[Unpack[Tin]], Tout] | Awaitable[Callable[[Unpack[Tin]], Tout]]:
+
+        filtered_injectors = self.filtered_injectors(allow_async, allow_sync)
+        callseq = self.get_callseq(filtered_injectors, frozenset(from_types), fn, force_commutative)
+
+        ret_fn = self.renderer.render(from_types, callseq, force_async=force_async, store_sources=self.store_sources)
+        if self.store_callseq:
+            setattr(ret_fn, '__breakshaft_callseq__', callseq)
+        return ret_fn
+
+    def mark_injector(self, *, rettype: Optional[type] = None, type_remap: Optional[dict[str, type]] = None):
         def inner(func: Callable):
-            self.add_injector(func)
+            self.add_injector(func, rettype=rettype, type_remap=type_remap)
             return func
 
         return inner
 
     def fork(self, fork_with: Optional[set[ConversionPoint]] = None) -> ConvRepo:
-        return ForkedConvRepo(self, fork_with or None, self.walker, self.renderer)
+        return ForkedConvRepo(self, fork_with or None,
+                              self.walker,
+                              self.renderer,
+                              self.store_callseq,
+                              self.store_sources)
 
 
 class ForkedConvRepo(ConvRepo):
@@ -99,16 +184,16 @@ class ForkedConvRepo(ConvRepo):
                  fork_from: ConvRepo,
                  fork_with: Optional[set[ConversionPoint]] = None,
                  graph_walker: Optional[GraphWalker] = None,
-                 renderer: Optional[ConvertorRenderer] = None):
+                 renderer: Optional[ConvertorRenderer] = None,
-        super().__init__(graph_walker, renderer)
+                 store_callseq: bool = False,
+                 store_sources: bool = False,
+                 ):
+        super().__init__(graph_walker, renderer, store_callseq, store_sources)
         if fork_with is None:
             fork_with = set()
         self._convertor_set = fork_with
         self._base_repo = fork_from
 
-    def add_injector(self, func: Callable, rettype: Optional[type] = None):
-        self._convertor_set |= set(ConversionPoint.from_fn(func, rettype=rettype))
-
     @property
     def convertor_set(self):
         return self._base_repo.convertor_set | self._convertor_set

src/breakshaft/graph_walker.py

@@ -1,9 +1,11 @@
+import collections.abc
 import typing
 from types import NoneType
 from typing import Callable, Optional
 
 from .models import ConversionPoint, Callgraph, CallgraphVariant, TransformationPoint, CompositionDirection
-from .util import extract_func_argtypes, all_combinations, extract_func_argtypes_seq
+from .util import extract_func_argtypes, all_combinations, extract_func_argtypes_seq, extract_return_type, universal_qualname
+from typing import Iterable
 
 
 class GraphWalker:
@@ -12,21 +14,27 @@ class GraphWalker:
     def generate_callgraph(cls,
                            injectors: frozenset[ConversionPoint],
                            from_types: frozenset[type],
-                           consumer_fn: Callable) -> Optional[Callgraph]:
+                           consumer_fn: Callable | Iterable[ConversionPoint] | ConversionPoint) -> Optional[Callgraph]:
-
-        into_types: frozenset[type] = extract_func_argtypes(consumer_fn)
 
         branches: frozenset[Callgraph] = frozenset()
 
-        for into_type in into_types:
+        # Хак, чтобы вынудить систему поставить первым преобразованием требуемый consumer
-            cg = cls.generate_callgraph_singletype(injectors, from_types, into_type)
+        # Новый TypeAliasType каждый раз будет иметь эксклюзивный хэш, вне зависимости от содержимого
-            if cg is None:
+        # При этом, TypeAliasType также выступает в роли ключа преобразования
-                return None
+        # Это позволяет переложить обработку аргументов consumer на внутренние механизмы построения графа преобразований
-            branches |= {cg}
+        type _tmp_type_for_consumer = object
-        variant = CallgraphVariant(
+
-            ConversionPoint(consumer_fn, NoneType, tuple(extract_func_argtypes_seq(consumer_fn))),
+        if isinstance(consumer_fn, collections.abc.Iterable):
-            branches, frozenset())
+            new_consumer_injectors = set()
-        return Callgraph(frozenset({variant}))
+            for fn in consumer_fn:
+                new_consumer_injectors.add(fn.copy_with(injects=_tmp_type_for_consumer))
+            injectors |= new_consumer_injectors
+        elif isinstance(consumer_fn, ConversionPoint):
+            injectors |= set(consumer_fn.copy_with(injects=_tmp_type_for_consumer))
+        else:
+            injectors |= set(ConversionPoint.from_fn(consumer_fn, _tmp_type_for_consumer))
+
+        return cls.generate_callgraph_singletype(injectors, from_types, _tmp_type_for_consumer)
 
     @classmethod
     def generate_callgraph_singletype(cls,
@@ -72,7 +80,17 @@ class GraphWalker:
                     variant_subgraphs.add(subg)
 
                 if not dead_end:
-                    consumed = frozenset(point.requires) & from_types
+
+                    for opt in point.opt_args:
+                        subg = cls.generate_callgraph_singletype(injectors,
+                                                                 from_types,
+                                                                 opt,
+                                                                 visited_path=visited_path.copy(),
+                                                                 visited_types=visited_types.copy())
+                        if subg is not None:
+                            variant_subgraphs.add(subg)
+
+                    consumed = (frozenset(point.requires) | frozenset(point.opt_args)) & from_types
                     variant = CallgraphVariant(point, frozenset(variant_subgraphs), consumed)
                     head = head.add_subgraph_variant(variant)
 
@@ -135,7 +153,7 @@ class GraphWalker:
             if len(variants) > 1:
                 # sorting by first injector func name for creating minimal cosistancy
                 # could lead to heizenbugs due to incosistancy in path selection between calls
-                variants.sort(key=lambda x: x.injector.fn.__qualname__)
+                variants.sort(key=lambda x: universal_qualname(x.injector.fn))
             return variants
 
         if len(variants) < 2:

src/breakshaft/models.py

@@ -10,24 +10,39 @@ from typing import Callable, Optional, get_type_hints, get_origin, Generator, ge
 
 from .util import extract_func_argtypes, extract_func_argtypes_seq, is_sync_context_manager_factory, \
     is_async_context_manager_factory, \
-    all_combinations, is_context_manager_factory
+    all_combinations, is_context_manager_factory, extract_func_arg_defaults, extract_func_args, extract_func_argnames, \
+    get_tuple_types, is_basic_type_annot, universal_qualname
 
 
 @dataclass(frozen=True)
 class ConversionPoint:
     fn: Callable
     injects: type
+    rettype: type
     requires: tuple[type, ...]
+    opt_args: tuple[type, ...]
+
+    def copy_with(self, **kwargs):
+        fn = kwargs.get('fn', self.fn)
+        rettype = kwargs.get('rettype', self.rettype)
+        injects = kwargs.get('injects', self.injects)
+        requires = kwargs.get('requires', self.requires)
+        opt_args = kwargs.get('opt_args', self.opt_args)
+        return ConversionPoint(fn, injects, rettype, requires, opt_args)
 
     def __hash__(self):
         return hash((self.fn, self.injects, self.requires))
 
     def __repr__(self):
-        return f'({",".join(map(str, self.requires))}) -> {self.injects.__qualname__}: {self.fn.__qualname__}'
+        injects_name = universal_qualname(self.injects)
+        fn_name = universal_qualname(self.fn)
+
+        return f'({",".join(map(str, self.requires))}) -> {injects_name}: {fn_name}'
 
     @property
-    def fn_args(self) -> list[type]:
+    def fn_args(self) -> list[tuple[str, type]]:
-        return extract_func_argtypes_seq(self.fn)
+        funcnames = extract_func_argnames(self.fn)
+        return list(zip(funcnames, self.requires + self.opt_args))
 
     @property
     def is_ctx_manager(self) -> bool:
@@ -38,15 +53,25 @@ class ConversionPoint:
         return inspect.iscoroutinefunction(self.fn) or is_async_context_manager_factory(self.fn)
 
     @classmethod
-    def from_fn(cls, func: Callable, rettype: Optional[type] = None):
+    def from_fn(cls,
-        if rettype is None:
+                func: Callable,
+                rettype: Optional[type] = None,
+                type_remap: Optional[dict[str, type]] = None,
+                ignore_basictype_return: bool = False) -> list[ConversionPoint]:
+        if type_remap is None:
             annot = get_type_hints(func)
-            rettype = annot.get('return')
+        else:
+            annot = type_remap
+
+        fn_rettype = annot.get('return')
+        if rettype is None:
+            rettype = fn_rettype
 
         if rettype is None:
             raise ValueError(f'Function {func.__qualname__} provided as injector, but return-type is not specified')
 
         rettype_origin = get_origin(rettype)
+        fn_rettype_origin = get_origin(fn_rettype)
         cm_out_origins = [
             typing.Generator,
             typing.Iterator,
@@ -59,22 +84,55 @@ class ConversionPoint:
         ]
         if any(map(lambda x: rettype_origin is x, cm_out_origins)) and is_context_manager_factory(func):
             rettype = get_args(rettype)[0]
+        if any(map(lambda x: fn_rettype_origin is x, cm_out_origins)) and is_context_manager_factory(func):
+            fn_rettype = get_args(fn_rettype)[0]
+
+        if not ignore_basictype_return and is_basic_type_annot(rettype):
+            return []
+
+        ret = []
+
+        tuple_unwrapped = get_tuple_types(rettype)
+        # Do not unwrap elipsis, but unwrap non-empty tuples
+        if len(tuple_unwrapped) > 0 and Ellipsis not in tuple_unwrapped:
+            for t in tuple_unwrapped:
+                if not is_basic_type_annot(t):
+                    ret += ConversionPoint.from_fn(func,
+                                                   rettype=t,
+                                                   type_remap=type_remap,
+                                                   ignore_basictype_return=ignore_basictype_return)
 
         argtypes: list[list[type]] = []
-        orig_argtypes = extract_func_argtypes_seq(func)
+        orig_args = extract_func_args(func, type_remap)
-        for argtype in orig_argtypes:
+        defaults = extract_func_arg_defaults(func)
+
+        orig_argtypes = []
+        for argname, argtype in orig_args:
+            orig_argtypes.append((argtype, argname in defaults.keys()))
+
+        default_map: list[bool] = []
+        for argtype, has_default in orig_argtypes:
             if isinstance(argtype, types.UnionType) or get_origin(argtype) is Union:
                 u_types = list(get_args(argtype)) + [argtype]
             else:
                 u_types = [argtype]
+            default_map.append(has_default)
             argtypes.append(u_types)
 
         argtype_combinations = all_combinations(argtypes)
-        ret = []
-        for argtype_combination in argtype_combinations:
-            ret.append(ConversionPoint(func, rettype, tuple(argtype_combination)))
 
-        # return InjectorPoint(func, rettype, argtypes)
+        for argtype_combination in argtype_combinations:
+            req_args = []
+            opt_args = []
+            for argt, has_default in zip(argtype_combination, default_map):
+                if has_default:
+                    opt_args.append(argt)
+                else:
+                    req_args.append(argt)
+            if rettype in req_args:
+                continue
+            ret.append(ConversionPoint(func, rettype, fn_rettype, tuple(req_args), tuple(opt_args)))
+
         return ret
 
 

src/breakshaft/renderer.py

@@ -7,32 +7,76 @@ import importlib.resources
 import jinja2
 
 from .models import ConversionPoint
-from .util import hashname
+from .util import hashname, get_tuple_types, is_basic_type_annot, universal_qualname
 
 
 class ConvertorRenderer(Protocol):
     def render(self,
                from_types: Sequence[type],
                callseq: Sequence[ConversionPoint],
-               force_async: bool = False) -> Callable:
+               force_async: bool = False,
+               store_sources: bool = False) -> Callable:
         raise NotImplementedError()
 
 
+type UnwprappedTuple = tuple[tuple[UnwprappedTuple, str] | str | None, ...]
+
+
+def unwrap_tuple_type(typ: type) -> UnwprappedTuple:
+    unwrap_tuple_result = ()
+    tuple_types = get_tuple_types(typ)
+    if len(tuple_types) > 0 and Ellipsis not in tuple_types:
+        for t in tuple_types:
+            if not is_basic_type_annot(t):
+                subtuple = unwrap_tuple_type(t)
+                hn = hashname(t)
+                if len(subtuple) > 0:
+                    unwrap_tuple_result += ((subtuple, hn),)
+                else:
+                    unwrap_tuple_result += (hn,)
+            else:
+                unwrap_tuple_result += (None,)
+
+    if not any(map(lambda x: x is not None, unwrap_tuple_result)):
+        return ()
+    return unwrap_tuple_result
+
+
 @dataclass
 class ConversionRenderData:
     inj_hash: str
     funchash: str
     funcname: str
-    funcargs: list[str]
+    funcargs: list[tuple[str, str]]
     is_ctxmanager: bool
     is_async: bool
+    unwrap_tuple_result: UnwprappedTuple
+    _injection: ConversionPoint
 
     @classmethod
-    def from_inj(cls, inj: ConversionPoint):
+    def from_inj(cls, inj: ConversionPoint, provided_types: set[type]):
+        argmap = inj.fn_args
+
         fnargs = []
-        for argtype in inj.requires:
+        for arg_id, argtype in enumerate(inj.requires):
-            fnargs.append(hashname(argtype))
+            argname = argmap[arg_id][0]
-        return cls(hashname(inj.injects), hashname(inj.fn), repr(inj.fn), fnargs, inj.is_ctx_manager, inj.is_async)
+            fnargs.append((argname, hashname(argtype)))
+
+        for arg_id, argtype in enumerate(inj.opt_args, len(inj.requires)):
+            argname = argmap[arg_id][0]
+            if argtype in provided_types:
+                fnargs.append((argname, hashname(argtype)))
+
+        unwrap_tuple_result = unwrap_tuple_type(inj.rettype)
+
+        return cls(hashname(inj.rettype),
+                   hashname(inj.fn),
+                   repr(inj.fn),
+                   fnargs,
+                   inj.is_ctx_manager,
+                   inj.is_async,
+                   unwrap_tuple_result,
+                   inj)
 
 
 @dataclass
@@ -42,6 +86,49 @@ class ConversionArgRenderData:
     typehash: str
 
 
+def deduplicate_callseq(conversion_models: list[ConversionRenderData]) -> list[ConversionRenderData]:
+    deduplicated_conv_models: list[ConversionRenderData] = []
+    deduplicated_hashes = set()
+    for conv_model in conversion_models:
+        if hash((conv_model.inj_hash, conv_model.funchash)) not in deduplicated_hashes:
+            deduplicated_conv_models.append(conv_model)
+            deduplicated_hashes.add(hash((conv_model.inj_hash, conv_model.funchash)))
+            continue
+
+        argnames = list(map(lambda x: x[1], conv_model.funcargs))
+        argument_changed = False
+        found_model = False
+        for m in deduplicated_conv_models:
+            if not found_model and m.funchash == conv_model.funchash:
+                found_model = True
+
+            if found_model and m.inj_hash in argnames:
+                argument_changed = True
+                break
+        if argument_changed:
+            deduplicated_conv_models.append(conv_model)
+            deduplicated_hashes.add(hash((conv_model.inj_hash, conv_model.funchash)))
+    return deduplicated_conv_models
+
+
+def render_data_from_callseq(from_types: Sequence[type],
+                             fnmap: dict[int, Callable],
+                             callseq: Sequence[ConversionPoint]):
+    conversion_models: list[ConversionRenderData] = []
+    ret_hash = 0
+    for call_id, call in enumerate(callseq):
+
+        provided_types = set(from_types)
+        for _call in callseq[:call_id]:
+            provided_types |= {_call.injects}
+            provided_types |= set(_call.requires)
+
+        fnmap[hash(call.fn)] = call.fn
+        conv = ConversionRenderData.from_inj(call, provided_types)
+        conversion_models.append(conv)
+    return conversion_models
+
+
 class InTimeGenerationConvertorRenderer(ConvertorRenderer):
     templateLoader: jinja2.BaseLoader
     templateEnv: jinja2.Environment
@@ -60,22 +147,20 @@ class InTimeGenerationConvertorRenderer(ConvertorRenderer):
     def render(self,
                from_types: Sequence[type],
                callseq: Sequence[ConversionPoint],
-               force_async: bool = False) -> Callable:
+               force_async: bool = False,
+               store_sources: bool = False) -> Callable:
 
         fnmap = {}
-        conversion_models = []
+        conversion_models: list[ConversionRenderData] = render_data_from_callseq(from_types, fnmap, callseq)
         ret_hash = 0
         is_async = force_async
+        for call_id, call in enumerate(callseq):
+            if call.is_async:
+                is_async = True
 
-        for call in callseq:
+        conversion_models = deduplicate_callseq(conversion_models)
-            fnmap[hash(call.fn)] = call.fn
-            conv = ConversionRenderData.from_inj(call)
-            if conv not in conversion_models:
-                conversion_models.append(conv)
-                if call.is_async:
-                    is_async = True
 
-        ret_hash = hash(callseq[-1].injects)
+        ret_hash = hashname(callseq[-1].rettype)
 
         conv_args = []
         for i, from_type in enumerate(from_types):
@@ -91,7 +176,10 @@ class InTimeGenerationConvertorRenderer(ConvertorRenderer):
             is_async=is_async,
         )
         convertor_functext = '\n'.join(list(filter(lambda x: len(x.strip()), convertor_functext.split('\n'))))
+        convertor_functext = convertor_functext.replace(', )', ')').replace(',)', ')')
         exec(convertor_functext, namespace)
         unwrap_func = namespace['convertor']
+        if store_sources:
+            setattr(unwrap_func, '__breakshaft_render_src__', convertor_functext)
 
         return typing.cast(Callable, unwrap_func)

src/breakshaft/templates/convertor.jinja2

@@ -1,13 +1,34 @@
 {% set ns = namespace(indent=0) %}
+
+{% macro unwrap_tuple(tupl, unwrap_name) -%}
+{%- set out -%}
+{% if tupl | length > 0 %}
+{% for t in tupl %}
+{% if t is string %}
+_{{t}} = _{{unwrap_name}}[{{loop.index0}}]
+{% endif %}
+{% if t.__class__.__name__ == 'tuple' %}
+_{{t[1]}} = _{{unwrap_name}}[{{loop.index0}}]
+{{unwrap_tuple(t[0], t[1])}}
+{% endif %}
+{% endfor %}
+
+{% endif %}
+{%- endset %}
+{{out}}
+{%- endmacro %}
+
+
 {% if is_async %}async {% endif %}def convertor({% for arg in conv_args %}_{{arg.typehash}}: "{{arg.typename}}",{% endfor %}){% if rettype %} -> '{{rettype}}'{% endif %}:
     {% for conv in conversions %}
     {% if conv.is_ctxmanager %}
     {{ '    ' * ns.indent }}# {{conv.funcname}}
-    {{ '    ' * ns.indent }}{% if conv.is_async %}async {% endif %}with _conv_funcmap[{{ conv.funchash }}]({% for conv_arg in conv.funcargs %}_{{conv_arg}}, {% endfor %}) as _{{ conv.inj_hash }}:
+    {{ '    ' * ns.indent }}{% if conv.is_async %}async {% endif %}with _conv_funcmap[{{ conv.funchash }}]({% for conv_arg in conv.funcargs %}{{conv_arg[0]}}=_{{conv_arg[1]}}, {% endfor %}) as _{{ conv.inj_hash }}:
     {% set ns.indent = ns.indent + 1 %}
     {% else %}
     {{ '    ' * ns.indent }}# {{conv.funcname}}
-    {{ '    ' * ns.indent }}_{{conv.inj_hash}} = {% if conv.is_async %}await {% endif %}_conv_funcmap[{{conv.funchash}}]({% for conv_arg in conv.funcargs %}_{{conv_arg}}, {% endfor %})
+    {{ '    ' * ns.indent }}_{{conv.inj_hash}} = {% if conv.is_async %}await {% endif %}_conv_funcmap[{{conv.funchash}}]({% for conv_arg in conv.funcargs %}{{conv_arg[0]}}=_{{conv_arg[1]}}, {% endfor %})
     {% endif %}
+{{unwrap_tuple(conv.unwrap_tuple_result, conv.inj_hash)  | indent((ns.indent + 1) * 4)}}
     {% endfor %}
     {{ '    ' * ns.indent }}return _{{ret_hash}}

src/breakshaft/util.py

@@ -1,11 +1,31 @@
 import inspect
+import typing
 from itertools import product
-from typing import Callable, get_type_hints, TypeVar, Any
+from typing import Callable, get_type_hints, TypeVar, Any, Optional
 
 
-def extract_func_args(func: Callable) -> list[tuple[str, type]]:
+def extract_func_argnames(func: Callable) -> list[str]:
     sig = inspect.signature(func)
-    type_hints = get_type_hints(func)
+    params = sig.parameters
+
+    args_info = []
+    for name, _ in params.items():
+        args_info.append(name)
+    return args_info
+
+
+def extract_return_type(func: Callable) -> Optional[type]:
+    hints = get_type_hints(func)
+    return hints.get('return')
+
+
+def extract_func_args(func: Callable, type_hints_remap: Optional[dict[str, type]] = None) -> list[tuple[str, type]]:
+    sig = inspect.signature(func)
+    if type_hints_remap is None:
+        type_hints = get_type_hints(func)
+    else:
+        type_hints = type_hints_remap
+
     params = sig.parameters
 
     args_info = []
@@ -42,6 +62,16 @@ def extract_func_argtypes_seq(func: Callable) -> list[type]:
     return ret
 
 
+def extract_func_arg_defaults(func: Callable) -> dict[str, object]:
+    sig = inspect.signature(func)
+    defaults = {
+        name: param.default
+        for name, param in sig.parameters.items()
+        if param.default is not inspect._empty
+    }
+    return defaults
+
+
 def is_context_manager_factory(obj: object) -> bool:
     return is_sync_context_manager_factory(obj) or is_async_context_manager_factory(obj)
 
@@ -63,3 +93,62 @@ T = TypeVar('T')
 
 def all_combinations(options: list[list[T]]) -> list[list[T]]:
     return [list(comb) for comb in product(*options)]
+
+
+def get_tuple_types(type_obj: type) -> tuple:
+    ret = ()
+
+    origin = getattr(type_obj, '__origin__', None)
+    if origin is tuple:
+        args = getattr(type_obj, '__args__', ())
+        ret = args if args else ()
+
+    return ret
+
+
+def is_basic_type_annot(type_annot) -> bool:
+    basic_types = {
+        int, float, str, bool, complex,
+        list, dict, tuple, set, frozenset,
+        bytes, bytearray, memoryview,
+        type(None), object
+    }
+
+    origin = getattr(type_annot, '__origin__', None)
+    args = getattr(type_annot, '__args__', None)
+
+    if type_annot in basic_types:
+        return True
+
+    if origin is not None:
+        if origin in basic_types or origin in {list, dict, tuple, set, frozenset}:
+            if args:
+                return all(is_basic_type_annot(arg) for arg in args)
+            return True
+        return False
+
+    if origin is typing.Union:
+        return all(is_basic_type_annot(arg) for arg in args)
+
+    return False
+
+
+def universal_qualname(any: Any) -> str:
+    ret = ''
+    if hasattr(any, '__qualname__'):
+        ret = any.__qualname__
+    elif hasattr(any, '__name__'):
+        ret = any.__name__
+    else:
+        ret = str(any)
+
+    ret = (ret
+           .replace('.', '_')
+           .replace('[', '_of_')
+           .replace(']', '_of_')
+           .replace(',', '_and_')
+           .replace(' ', '_')
+           .replace('\'', '')
+           .replace('<', '')
+           .replace('>', ''))
+    return ret

src/breakshaft/util_mermaid.py

@@ -1,4 +1,4 @@
-from .models import Callgraph, TransformationPoint
+from .models import Callgraph, TransformationPoint, ConversionPoint
 from .util import hashname
 
 
@@ -68,3 +68,14 @@ def draw_callgraph_mermaid(g: Callgraph, split_duplicates=False, skip_title=Fals
         ret += 'flowchart TD\n\n'
     ret += '    %%defs:\n' + '\n'.join(d) + '\n\n    %%edges:\n' + '\n'.join(e)
     return ret
+
+
+def draw_callseq_mermaid(callseq: list[ConversionPoint]):
+    ret = ['flowchart TD\n\n']
+    ret += ['   %%defs:']
+    for cp_i, cp in enumerate(callseq):
+        ret.append(f'    e{cp_i}["{shield_mermaid_name(str(cp))}"]')
+    ret += ['', '', '   %%edges:']
+    for cp_i, cp in enumerate(callseq[:-1]):
+        ret.append(f'    e{cp_i}-->e{cp_i + 1}')
+    return '\n'.join(ret)

tests/test_basic.py

@@ -38,3 +38,33 @@ def test_basic():
     fn2 = repo.get_conversion((int,), consumer, force_commutative=True, force_async=False, allow_async=False)
     dep = fn2(123)
     assert dep == 123
+
+
+def test_union_deps():
+    repo = ConvRepo()
+
+    @repo.mark_injector()
+    def b_to_a(b: B) -> A:
+        return A(int(b.b))
+
+    @repo.mark_injector()
+    def a_to_b(a: A) -> B:
+        return B(float(a.a))
+
+    @repo.mark_injector()
+    def int_to_a(i: int) -> A:
+        return A(i)
+
+    def consumer(dep: A | B) -> int:
+        if isinstance(dep, A):
+            return dep.a
+        else:
+            return int(dep.b)
+
+    fn1 = repo.get_conversion((B,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn1(B(42.1))
+    assert dep == 42
+
+    fn2 = repo.get_conversion((int,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn2(123)
+    assert dep == 123

tests/test_default_args.py

@@ -0,0 +1,144 @@
+from dataclasses import dataclass
+
+from src.breakshaft.convertor import ConvRepo
+
+
+@dataclass
+class A:
+    a: int
+
+
+@dataclass
+class B:
+    b: float
+
+
+type optC = str
+
+
+def test_default_consumer_args():
+    repo = ConvRepo()
+
+    @repo.mark_injector()
+    def b_to_a(b: B) -> A:
+        return A(int(b.b))
+
+    @repo.mark_injector()
+    def a_to_b(a: A) -> B:
+        return B(float(a.a))
+
+    @repo.mark_injector()
+    def int_to_a(i: int) -> A:
+        return A(i)
+
+    def consumer(dep: A, opt_dep: optC = '42') -> tuple[int, str]:
+        return dep.a, opt_dep
+
+    fn1 = repo.get_conversion((B,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn1(B(42.1))
+    assert dep == (42, '42')
+
+    fn2 = repo.get_conversion((int,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn2(123)
+    assert dep == (123, '42')
+
+    fn3 = repo.get_conversion((int, optC), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn3(123, '1')
+    assert dep == (123, '1')
+
+
+def test_optional_default_none_consumer_args():
+    repo = ConvRepo()
+
+    @repo.mark_injector()
+    def b_to_a(b: B | None = None) -> A:
+        return A(int(b.b))
+
+    @repo.mark_injector()
+    def a_to_b(a: A) -> B | None:
+        return B(float(a.a))
+
+    @repo.mark_injector()
+    def int_to_a(i: int) -> A:
+        return A(i)
+
+    def consumer(dep: A, opt_dep: optC = '42') -> tuple[int, str]:
+        return dep.a, opt_dep
+
+    fn1 = repo.get_conversion((B,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn1(B(42.1))
+    assert dep == (42, '42')
+
+    fn2 = repo.get_conversion((int,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn2(123)
+    assert dep == (123, '42')
+
+    fn3 = repo.get_conversion((int, optC), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn3(123, '1')
+    assert dep == (123, '1')
+
+
+def test_default_inj_args():
+    repo = ConvRepo()
+
+    @repo.mark_injector()
+    def b_to_a(b: B) -> A:
+        return A(int(b.b))
+
+    @repo.mark_injector()
+    def a_to_b(a: A) -> B:
+        return B(float(a.a))
+
+    @repo.mark_injector()
+    def int_to_a(i: int, opt_dep: optC = '42') -> A:
+        return A(i + int(opt_dep))
+
+    def consumer(dep: A) -> int:
+        return dep.a
+
+    fn1 = repo.get_conversion((B,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn1(B(42.1))
+    assert dep == 42
+
+    fn2 = repo.get_conversion((int,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn2(123)
+    assert dep == 123 + 42
+
+    fn3 = repo.get_conversion((int, optC,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn3(123, '0')
+    assert dep == 123
+
+
+def test_default_graph_override():
+    repo = ConvRepo()
+
+    @repo.mark_injector()
+    def b_to_a(b: B) -> A:
+        return A(int(b.b))
+
+    @repo.mark_injector()
+    def a_to_b(a: A) -> B:
+        return B(float(a.a))
+
+    @repo.mark_injector()
+    def int_to_a(i: int, opt_dep: optC = '42') -> A:
+        return A(i + int(opt_dep))
+
+    @repo.mark_injector()
+    def inject_opt_dep() -> optC:
+        return '12345'
+
+    def consumer(dep: A) -> int:
+        return dep.a
+
+    fn1 = repo.get_conversion((B,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn1(B(42.1))
+    assert dep == 42
+
+    fn2 = repo.get_conversion((int,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn2(123)
+    assert dep == 123 + 12345
+
+    fn3 = repo.get_conversion((int, optC,), consumer, force_commutative=True, force_async=False, allow_async=False)
+    dep = fn3(123, '0')
+    assert dep == 123

tests/test_pipeline.py

@@ -0,0 +1,118 @@
+from dataclasses import dataclass
+
+from src.breakshaft.convertor import ConvRepo
+
+
+@dataclass
+class A:
+    a: int
+
+
+@dataclass
+class B:
+    b: float
+
+
+type optC = str
+
+
+def test_default_consumer_args():
+    repo = ConvRepo(store_sources=True)
+
+    @repo.mark_injector()
+    def b_to_a(b: B) -> A:
+        return A(int(b.b))
+
+    @repo.mark_injector()
+    def a_to_b(a: A) -> B:
+        return B(float(a.a))
+
+    @repo.mark_injector()
+    def int_to_a(i: int) -> A:
+        return A(i)
+
+    type ret1 = tuple[int, str]
+
+    def consumer1(dep: A, opt_dep: optC = '42') -> ret1:
+        return dep.a, opt_dep
+
+    def consumer2(dep: A, dep1: ret1) -> optC:
+        return str((dep.a, dep1))
+
+    p1 = repo.create_pipeline(
+        (B,),
+        [consumer1, consumer2],
+        force_commutative=True,
+        allow_sync=True,
+        allow_async=False,
+        force_async=False
+    )
+    res = p1(B(42.1))
+    assert res == "(42, (42, '42'))"
+
+    p2 = repo.create_pipeline(
+        (B,),
+        [consumer1, consumer2, consumer1],
+        force_commutative=True,
+        allow_sync=True,
+        allow_async=False,
+        force_async=False
+    )
+    res = p2(B(42.1))
+    assert res == (42, "(42, (42, '42'))")
+
+
+def test_pipeline_with_subgraph_duplicates():
+    repo = ConvRepo()
+
+    b_to_a_calls = [0]
+
+    @repo.mark_injector()
+    def b_to_a(b: B) -> A:
+        b_to_a_calls[0] += 1
+        return A(int(b.b))
+
+    @repo.mark_injector()
+    def a_to_b(a: A) -> B:
+        return B(float(a.a))
+
+    @repo.mark_injector()
+    def int_to_a(i: int) -> A:
+        return A(i)
+
+    type ret1 = tuple[int, str]
+
+    cons1_calls = [0]
+    cons2_calls = [0]
+
+    def consumer1(dep: A, opt_dep: optC = '42') -> A:
+        cons1_calls[0] += 1
+        return A(dep.a + int(opt_dep))
+
+    def consumer2(dep: A) -> optC:
+        cons2_calls[0] += 1
+        return str(dep.a)
+
+    p1 = repo.create_pipeline(
+        (B,),
+        [consumer1, consumer2, consumer1, consumer2, consumer1, consumer2, consumer1, consumer2, consumer1],
+        force_commutative=True,
+        allow_sync=True,
+        allow_async=False,
+        force_async=False
+    )
+    res = p1(B(42.1))
+    assert res.a == 42 + (42 * 31)
+    assert b_to_a_calls[0] == 1
+    assert cons1_calls[0] == 5
+    assert cons2_calls[0] == 4
+
+
+def convertor(_5891515089754: "<class 'test_pipeline.B'>"):
+    # <function test_default_consumer_args.<locals>.b_to_a at 0x7f5bb1be02c0>
+    _5891515089643 = _conv_funcmap[8751987548204](b=_5891515089754)
+    # <function test_default_consumer_args.<locals>.consumer1 at 0x7f5bb1be0c20>
+    _8751987542640 = _conv_funcmap[8751987548354](dep=_5891515089643)
+    # <function test_default_consumer_args.<locals>.consumer2 at 0x7f5bb1be0540>
+    _8751987537115 = _conv_funcmap[8751987548244](dep=_5891515089643, dep1=_8751987542640)
+    return _8751987542640

tests/test_tuple_unwrap.py

@@ -0,0 +1,84 @@
+from dataclasses import dataclass
+
+from breakshaft.models import ConversionPoint
+from src.breakshaft.convertor import ConvRepo
+
+
+@dataclass
+class A:
+    a: int
+
+
+@dataclass
+class B:
+    b: float
+
+
+@dataclass
+class C:
+    c: int
+
+
+@dataclass
+class D:
+    d: str
+
+
+def test_conv_point_tuple_unwrap():
+    def conv_into_bc(a: A) -> tuple[B, C]:
+        return B(a.a), C(a.a)
+
+    def conv_into_bcd(a: A) -> tuple[B, tuple[C, D]]:
+        return B(a.a), (C(a.a), D(str(a.a)))
+
+    def conv_into_bcda(a: A) -> tuple[B, tuple[C, tuple[D, A]]]:
+        return B(a.a), (C(a.a), (D(str(a.a)), a))
+
+    cps_bc = ConversionPoint.from_fn(conv_into_bc)
+    assert len(cps_bc) == 3  # tuple[...], B, C
+
+    cps_bcd = ConversionPoint.from_fn(conv_into_bcd)
+
+    assert len(cps_bcd) == 5  # tuple[B,...], B, tuple[C,D], C, D
+
+    cps_bcda = ConversionPoint.from_fn(conv_into_bcda)
+
+    assert len(cps_bcda) == 6  # ignores (A,...)->A
+
+
+def test_ignore_basic_types():
+    def conv_into_b_int(a: A) -> tuple[B, int]:
+        return B(a.a), a.a
+
+    cps = ConversionPoint.from_fn(conv_into_b_int)
+    assert len(cps) == 2  # tuple[...], B
+
+
+def test_codegen_tuple_unwrap():
+    repo = ConvRepo(store_sources=True)
+
+    @repo.mark_injector()
+    def conv_into_bcd(a: A) -> tuple[B, tuple[C, D]]:
+        return B(a.a), (C(a.a), D(str(a.a)))
+
+    type Z = A
+
+    @repo.mark_injector()
+    def conv_d_a(d: D) -> Z:
+        return A(int(d.d))
+
+    def consumer1(dep: D) -> int:
+        return int(dep.d)
+
+    def consumer2(dep: Z) -> int:
+        return int(dep.a)
+
+    fn1 = repo.get_conversion((A,), consumer1, force_commutative=True, force_async=False, allow_async=False)
+    assert fn1(A(1)) == 1
+
+    fn2 = repo.get_conversion((A,), consumer2, force_commutative=True, force_async=False, allow_async=False)
+    assert fn2(A(1)) == 1
+
+    pip = repo.create_pipeline((A,), [consumer1, consumer2], force_commutative=True, force_async=False, allow_async=False)
+    assert pip(A(1)) == 1
+    print(pip.__breakshaft_render_src__)

tests/test_typehints_remap.py

@@ -0,0 +1,58 @@
+from dataclasses import dataclass
+from typing import Annotated
+
+import pytest
+
+from breakshaft.models import ConversionPoint
+from src.breakshaft.convertor import ConvRepo
+
+
+@dataclass
+class A:
+    a: int
+
+
+@dataclass
+class B:
+    b: float
+
+
+def test_basic():
+    repo = ConvRepo()
+
+    @repo.mark_injector()
+    def int_to_a(i: int) -> A:
+        return A(i)
+
+    def consumer(dep: A) -> B:
+        return B(float(dep.a))
+
+    type NewA = A
+    type_remap = {'dep': NewA, 'return': B}
+
+    assert len(ConversionPoint.from_fn(consumer, type_remap=type_remap)) == 1
+
+    with pytest.raises(ValueError):
+        fn1 = repo.get_conversion((int,), ConversionPoint.from_fn(consumer, type_remap=type_remap),
+                                  force_commutative=True, force_async=False, allow_async=False)
+
+    repo.mark_injector(type_remap={'i': int, 'return': NewA})(int_to_a)
+
+    fn1 = repo.get_conversion((int,), ConversionPoint.from_fn(consumer, type_remap=type_remap),
+                              force_commutative=True, force_async=False, allow_async=False)
+
+    assert fn1(42).b == 42.0
+
+    def consumer1(dep: B) -> A:
+        return A(int(dep.b))
+
+    p1 = repo.create_pipeline(
+        (int,),
+        [ConversionPoint.from_fn(consumer, type_remap=type_remap), consumer1, consumer],
+        force_commutative=True,
+        allow_sync=True,
+        allow_async=False,
+        force_async=False
+    )
+
+    assert p1(123).b == 123.0