Name a discovered module for where it sits under the synced directory

`fluksio sync dev` refused a one-file-per-directory layout: s1..s4 each
holding `study.py` stopped the walk with "two files would both import as
'study'". The suggested `__init__.py` per study directory fixed discovery
and broke the bare `from study import ...` a test beside it does. A file is
now named for its path under the directory being synced — `dev/s1/study.py`
imports as `s1.study`, the directories between being namespace packages —
so nothing collides and nothing has to be added. A file at the root keeps
its bare name, and the generated shim follows: `import_root` walks one
directory per dotted segment and lands on the same root.

The collision message stays for the spelling that can still reach it: two
files named on the command line, each rooted at its own directory.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019Hra4ndWMCLU5F3KjUuVAc

backend/fluksio/sdk/cli.py

@@ -136,8 +136,8 @@ def _import(root: str, dotted: str, expect: Path | None = None) -> None:
             f"  {actual}\n  {expect}\n"
             "Python keeps one module per name, and a node's generated body "
             "imports by that name, so the second would run the first's code. "
-            "Put an `__init__.py` in each directory — they become "
-            f"'<dir>.{dotted}' and stop colliding — or rename one of the files."
+            "Sync the directory they are both under — each then imports as "
+            "'<dir>.<file>' — or rename one of the files."
         )
 
 
@@ -211,7 +211,14 @@ def discover(targets: list[str], keep_going: bool = False) -> list[Flow]:
                 if entry.is_dir():
                     _import_package(entry)
                 else:
-                    _import(*_module_of(entry), entry)
+                    # Named for where it sits under the directory being synced,
+                    # so `dev/s1/study.py` imports as `s1.study` and one
+                    # `study.py` per study collides with nothing. No
+                    # `__init__.py` needed: the directories in between are
+                    # namespace packages. A file at the root keeps its bare
+                    # name, which is what it has always had.
+                    dotted = ".".join(entry.relative_to(path).with_suffix("").parts)
+                    _import(str(path), dotted, entry)
             except SyncError:
                 # A name collision is never somebody else's problem: it would
                 # put the wrong file behind a node.