RomanNum3ral
/
hermes-agent
mirror of https://github.com/NousResearch/hermes-agent.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

polish: document newline residual, reuse span local, cheap check first 

Review follow-ups on the guard: state the accepted \n-residual in the
comment, reuse the span local in the next condition instead of
re-slicing, and short-circuit the substring checks before the regex.
This commit is contained in:
kshitij 2026-08-07 17:46:34 +05:30
parent 9377c5a539
commit 72b7305263
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
agent/redact.py
View File

@ -512,14 +512,16 @@ def _mask_control_split_tokens(text: str, mask_fn) -> str:
# A complete token at end-of-line followed by a word line
# (``ghp_<token>\nbutton [ref=e3]``) joins into one stripped-copy
# match and the mask eats ``button``. Line structure is legitimate;
# the self-matching fragment is handled by the ordinary prefix pass.
# the self-matching fragment is handled by the ordinary prefix pass
# (any remainder past the newline is left unmasked — accepted
# residual to preserve line structure).
# For NON-newline controls (ESC, ZWSP, ...) the join proceeds even
# when a fragment self-matches: those bytes never legitimately sit
# between a token and adjacent prose, and skipping there let the
# non-matching remainder of a split token leak
# (``sk-<head>\x1b<tail>`` masked only the head).
span = text[start_orig:end_orig]
if _PREFIX_RE.search(span) and ("\n" in span or "\r" in span):
if ("\n" in span or "\r" in span) and _PREFIX_RE.search(span):
continue
# Reject matches whose original span crosses a non-token char
# (e.g. ``sk_abc…\nTAVILY_API_KEY=…`` — the ``=`` is not part of a
@ -527,7 +529,7 @@ def _mask_control_split_tokens(text: str, mask_fn) -> str:
# reject when the match runs into a ``KEY=`` name: a real token value
# is followed by a newline/space/end, not ``=``.
if (all(c in _TOKEN_BODY_CHARS or _CONTROL_CHARS_RE.match(c)
for c in text[start_orig:end_orig])
for c in span)
and (end_orig >= len(text) or text[end_orig] != "=")):
matches.append((start_orig, end_orig, mask_fn(body)))
for start_orig, end_orig, replacement in reversed(matches):