51 lines
1.7 KiB
Python
Raw Normal View History

"""Version-to-version comparison for artifacts (HUX-04).
Text artifacts get a unified diff from ``difflib``; anything that is not
valid UTF-8, or whose artifact type is binary, is described by sizes and
hashes only so the UI never tries to render bytes as text.
"""
from __future__ import annotations
import difflib
import hashlib
from typing import Any
TEXT_TYPES = frozenset({"markdown", "code", "html", "svg", "json", "csv"})
def is_text_type(artifact_type: str) -> bool:
"""True for artifact types whose content is served as UTF-8 text."""
return artifact_type in TEXT_TYPES
def decode_text(data: bytes) -> str | None:
"""UTF-8 decode, or None when the bytes are not text."""
try:
return data.decode("utf-8")
except UnicodeDecodeError:
return None
def unified(artifact_type: str, older: bytes, newer: bytes, from_n: int, to_n: int) -> dict[str, Any]:
"""Diff record ``{"from", "to", "unified"|"binary"}`` for two versions."""
record: dict[str, Any] = {"from": from_n, "to": to_n}
old_text = decode_text(older) if is_text_type(artifact_type) else None
new_text = decode_text(newer) if is_text_type(artifact_type) else None
if old_text is None or new_text is None:
record["binary"] = {
"from_bytes": len(older),
"to_bytes": len(newer),
"from_hash": "sha256:" + hashlib.sha256(older).hexdigest(),
"to_hash": "sha256:" + hashlib.sha256(newer).hexdigest(),
}
return record
lines = difflib.unified_diff(
old_text.splitlines(keepends=True),
new_text.splitlines(keepends=True),
fromfile=f"v{from_n}",
tofile=f"v{to_n}",
)
record["unified"] = "".join(lines)
return record