test(extract): hand-built office fixtures with frozen extracted text

Three hand-laid OOXML containers, every part written out by hand and zipped
with a fixed date_time so they are byte-reproducible. No converter output
anywhere in them: a .docx written by the converter and read by the converter
proves only that the converter agrees with itself, and would stay green through
any conversion defect that is symmetric -- which is most of them.

  two-line-krav.docx    heading + label/value on one line (the docx mirror of
                        the PDF fixture)
  no-styles-krav.docx   the SAME document without word/styles.xml
  two-line-krav.xlsx    sheet name as heading + label/value on one row

THE FIXTURES FOUND A REAL DEFECT IN THE SEAM THEY WERE MEANT TO PIN. The
converter call used pypandoc's TEXT entry point, which takes an `encoding`
because it treats its source as text -- and that corrupts a zip. The xlsx
fixture failed with `Failed to unpack XLSX archive: not enough bytes` while
reading correctly from disk with the same binary. The docx of the same shape
happened to survive, which is the part worth writing down: the defect is silent
for some inputs and fatal for others, so "it worked on the file I tried" was
never evidence. Input now goes through a temporary file.

Two measurements while building, both the same shape -- structurally valid
input, silently reduced output, exit code 0, no warning:

- Without word/styles.xml the docx extracts as flat prose with no heading. A
  fixture lacking that part would pin the body and pin nothing about structure.
  Committed as a negative control that RUNS rather than a sentence in a README.
- With inline strings rather than a shared string table, the xlsx extracts with
  the sheet name intact and every cell value gone. The fixture uses a dimension
  element and a shared string table instead.

The frozen literals are pinned to a NAMED converter version, asserted beside
them: a frozen literal without one says "these bytes" without saying what
produced them.

Suite 908 -> 913. Fixtures regenerate byte-identically.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

src/llm_ingestion_okf/extract.py

@@ -26,6 +26,7 @@ from __future__ import annotations
 import csv
 import functools
 import io
+import tempfile
 import warnings
 from collections.abc import Callable, Sequence
 from html.parser import HTMLParser
@@ -250,13 +251,30 @@ def _convert_bytes(source: bytes, to: str, format: str, extra_args: Sequence[str
     would otherwise need the binary present and a real office document, which
     would make the seam's own logic untestable on a machine without the extra.
     The conversion itself is covered by the frozen-text fixtures instead.
+
+    THE INPUT GOES THROUGH A FILE, NOT THROUGH THE TEXT ENTRY POINT. Every
+    format here is a binary container, and the converter's text entry point
+    takes an `encoding` because it treats its source as text -- which corrupts
+    a zip. Measured: a hand-laid `.xlsx` that pandoc reads correctly from disk
+    fails through the text path with `Failed to unpack XLSX archive: not enough
+    bytes`. A `.docx` of the same shape happened to survive, which is what
+    makes this worth writing down: the defect is SILENT for some inputs and
+    fatal for others, so "it worked on the file I tried" is not evidence here.
+
+    The temporary directory is removed on every path, including the failure
+    one, and nothing outside it is written.
     """
     import pypandoc
 
     from ._pandoc import converter_path
 
-    with converter_path():
-        return str(pypandoc.convert_text(source, to, format=format, extra_args=list(extra_args)))
+    with tempfile.TemporaryDirectory() as staging:
+        staged = Path(staging) / f"input.{format}"
+        staged.write_bytes(source)
+        with converter_path():
+            return str(
+                pypandoc.convert_file(str(staged), to, format=format, extra_args=list(extra_args))
+            )
 
 
 def _extract_office(suffix: str, data: bytes) -> str: