Semantic Naming & Smart Binding

SynaFlow automatically resolves parameter names to step outputs using Base Dataset Names — even when you use synonyms.

Base Dataset Name

Every step produces a Base Dataset whose name is the absolute plural form of the step name:

Step name	Base Dataset
user	users
users	users
user_list	users
person	people
fetched_securities	fetched_securities

Stripping common suffixes (_list, _set, _dict, _tuple) and pluralizing via inflect ensures consistent naming.

Smart Binding

You don't need to match step names exactly. Any synonym works:

def items() -> Generator[User, None, None]:   # step name: "items"
    ...

def transform(item: User) -> User:             # singular → binds to "items"
    ...

def collector(items_list: list[User]) -> None: # suffixed → binds to "items"
    ...

The framework matches item (singular) and items_list (suffixed) to the producer items by comparing their Base Dataset names.

Conflict Prevention

Two steps with the same Base Dataset name raise an error at build time:

# ❌ ValueError: 'user' and 'users' both map to Base Dataset 'users'
pipeline(
    steps=[
        step("user", fn=fn1),
        step("users", fn=fn2),
    ]
)

Similarly, one function with two parameters mapping to the same base name is rejected.

Naming Convention

• EACH-mode consumers naturally use the singular form (item: User)
• ALL-mode consumers naturally use the plural form (items: list[User])
• The framework handles the mapping transparently — you write what is semantically correct.