Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@src/snakemake/dag.py`:
- Around line 3197-3208: The current loop builds orderd_jobs by iterating
self.toposorted() (all jobs) which diverges from the filtered set used to
compute total_count and also uses repeated orderd_jobs.count(...) calls making
it quadratic; change the logic in the block that constructs
orderd_jobs/deduplicated_job_list so it only iterates the same filtered job set
used for total_count (e.g., the needrun/finished filtered list) instead of
self.toposorted(), and replace repeated .count(...) calls with a single pass
counter (e.g., collections.Counter) to compute counts once, then populate rows
and stats_dict from that counter; update references to orderd_jobs,
deduplicated_job_list, rows, and stats_dict accordingly so ordering and counts
match the filtered jobs.

In `@tests/tests.py`:
- Around line 2978-2981: The except block that currently does "assert False"
should instead raise an AssertionError with a clear message and chain the
original ValueError for context: catch the ValueError from names.index(expected,
idx) and raise AssertionError(f"Expected rule '{expected}' not found in stats
output") from the caught ValueError so the original traceback is preserved (look
for the try/except around names.index(expected, idx) using variables idx, names,
expected).

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@tests/tests.py`:
- Line 2952: Replace the bare open call with a context manager to ensure the
file handle is closed: change the line that assigns txt from open(stats_path,
"r", encoding="utf-8").read() to using with open(stats_path, "r",
encoding="utf-8") as f: txt = f.read() so the file is properly closed after
reading.

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@tests/tests.py`:
- Around line 2957-3007: The test test_stats_table_order_and_counts leaves the
temp output directory when run(..., cleanup=False), which can accumulate
artifacts in CI; update the test to remove the created outdir on success
(mirroring other tests) by calling the same cleanup routine used elsewhere after
the assertions (e.g., remove the directory returned by run/dpath or call the
project-specific cleanup helper), ensuring cleanup runs only after all
assertions pass and does not mask failures.