Kjell Tore Guttormsen
cda4a6a497
TDD step 2: markdown-table rendering with the load-bearing escape order (backslash, then pipe, then newlines to spaces with CRLF as one unit), SQL value-to-text conversion (NULL empty, plain-decimal integers, shortest round-trip floats, explicit bool/BLOB rejection — never silent coercion), and the verbatim fenced block for http bodies. Table shape and fenced-block form match the reference implementation byte for byte ahead of the shared §11 golden fixtures. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QeqhJpYQyghASjiJo5EhGg
115 lines
3.4 KiB
Python
115 lines
3.4 KiB
Python
"""Body rendering as pure functions (ingest-spec §5).
|
|
|
|
Byte-compatible with the reference implementation (portfolio-optimiser):
|
|
markdown table with `| --- |` separator, escape order `\\` then `|` then
|
|
newlines→space, SQL value conversion with fail-on-unknown-type, and the
|
|
verbatim fenced block for http bodies.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import pytest
|
|
|
|
from llm_ingestion_okf.errors import IngestError, RenderError
|
|
from llm_ingestion_okf.render import render_fenced_block, render_table, sql_value_to_text
|
|
|
|
# --- error hierarchy ---
|
|
|
|
|
|
def test_render_error_is_ingest_error() -> None:
|
|
assert issubclass(RenderError, IngestError)
|
|
|
|
|
|
# --- render_table ---
|
|
|
|
|
|
def test_basic_table_bytes() -> None:
|
|
assert (
|
|
render_table(["a", "b"], [["1", "x"], ["2", "y"]])
|
|
== "| a | b |\n| --- | --- |\n| 1 | x |\n| 2 | y |\n"
|
|
)
|
|
|
|
|
|
def test_header_only_table() -> None:
|
|
assert render_table(["a"], []) == "| a |\n| --- |\n"
|
|
|
|
|
|
def test_table_is_lf_only_with_one_trailing_newline() -> None:
|
|
rendered = render_table(["a"], [["x"]])
|
|
assert "\r" not in rendered
|
|
assert rendered.endswith("|\n")
|
|
assert not rendered.endswith("\n\n")
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
("raw", "escaped"),
|
|
[
|
|
("a|b", "a\\|b"), # pipe → \|
|
|
("a\\b", "a\\\\b"), # backslash → \\
|
|
("|", "\\|"), # escape ORDER: backslash first, else this becomes \\|
|
|
("\\|", "\\\\\\|"), # combined: raw \| → \\ + \|
|
|
("a\nb", "a b"), # LF → single space
|
|
("a\rb", "a b"), # CR → single space
|
|
("a\r\nb", "a b"), # CRLF is ONE unit → one space
|
|
],
|
|
)
|
|
def test_cell_escaping(raw: str, escaped: str) -> None:
|
|
assert render_table(["h"], [[raw]]) == f"| h |\n| --- |\n| {escaped} |\n"
|
|
|
|
|
|
def test_header_cells_escaped_identically() -> None:
|
|
assert render_table(["a|b"], []) == "| a\\|b |\n| --- |\n"
|
|
|
|
|
|
# --- sql_value_to_text ---
|
|
|
|
|
|
def test_sql_null_is_empty_string() -> None:
|
|
assert sql_value_to_text(None) == ""
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
("value", "text"), [(0, "0"), (42, "42"), (-7, "-7"), (10**20, "100000000000000000000")]
|
|
)
|
|
def test_sql_integers_plain_decimal(value: int, text: str) -> None:
|
|
assert sql_value_to_text(value) == text
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
("value", "text"),
|
|
[(1.5, "1.5"), (0.1, "0.1"), (-2.25, "-2.25"), (1e300, "1e+300")],
|
|
)
|
|
def test_sql_floats_shortest_round_trip(value: float, text: str) -> None:
|
|
assert sql_value_to_text(value) == text
|
|
|
|
|
|
def test_sql_text_verbatim_unescaped() -> None:
|
|
# Table escaping is render_table's job, not the value conversion's.
|
|
assert sql_value_to_text("a|b\\c") == "a|b\\c"
|
|
|
|
|
|
@pytest.mark.parametrize("value", [True, False, b"blob", [1], {"a": 1}, object()])
|
|
def test_sql_unsupported_types_fail(value: object) -> None:
|
|
"""Never silent coercion (spec §5) — bool guarded explicitly (int subclass)."""
|
|
with pytest.raises(RenderError):
|
|
sql_value_to_text(value)
|
|
|
|
|
|
# --- render_fenced_block (http bodies) ---
|
|
|
|
|
|
def test_fenced_block_verbatim() -> None:
|
|
assert render_fenced_block("hello") == "```\nhello\n```\n"
|
|
|
|
|
|
def test_fenced_block_strips_trailing_newlines_only() -> None:
|
|
assert render_fenced_block("hello\n\n") == "```\nhello\n```\n"
|
|
assert render_fenced_block("a\n\nb\n") == "```\na\n\nb\n```\n"
|
|
|
|
|
|
def test_fenced_block_not_table_escaped() -> None:
|
|
assert render_fenced_block("a|b\\c") == "```\na|b\\c\n```\n"
|
|
|
|
|
|
def test_fenced_block_empty_body() -> None:
|
|
assert render_fenced_block("") == "```\n\n```\n"
|