i/Aether
1
0
Fork 0
mirror of https://github.com/fawney19/Aether.git synced 2026-09-03 09:50:21 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix(normalizer): 移除 Claude system 中的 billing header

Browse Source
This commit is contained in:
fawney19
2026-03-17 11:10:04 +08:00
parent dcbd7dc219
commit 5dae2a4792
3 changed files with 147 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

46
src/core/api_format/conversion/normalizers/claude.py
View File

@@ -66,6 +66,7 @@ from src.core.logger import logger
class ClaudeNormalizer(FormatNormalizer):
FORMAT_ID = "claude:chat"
_ANTHROPIC_BILLING_HEADER_PREFIX = "x-anthropic-billing-header:"
capabilities = FormatCapabilities(
supports_stream=True,
supports_error_conversion=True,
@@ -1128,7 +1129,12 @@ class ClaudeNormalizer(FormatNormalizer):
if system_value is None:
return None, dropped, None
if isinstance(system_value, str):
return (system_value or None), dropped, None
sanitized_text, removed_header_count = self._strip_anthropic_billing_headers(
system_value
)
if removed_header_count:
dropped["claude_system_billing_header_removed"] = removed_header_count
return sanitized_text, dropped, None
if isinstance(system_value, list):
texts: list[str] = []
@@ -1143,14 +1149,26 @@ class ClaudeNormalizer(FormatNormalizer):
if item.get("type") == "text":
text = item.get("text")
if text:
texts.append(str(text))
normalized_text, removed_header_count = (
self._strip_anthropic_billing_headers(text)
)
if removed_header_count:
dropped["claude_system_billing_header_removed"] = (
dropped.get("claude_system_billing_header_removed", 0)
+ removed_header_count
)
if not normalized_text:
continue
texts.append(normalized_text)
seg_extra: dict[str, Any] = {}
cc = item.get("cache_control")
if isinstance(cc, dict):
seg_extra["cache_control"] = cc
has_cache_control = True
segments.append(
InstructionSegment(role=Role.SYSTEM, text=str(text), extra=seg_extra)
InstructionSegment(
role=Role.SYSTEM, text=normalized_text, extra=seg_extra
)
)
else:
dropped_key = f"claude_system_item:{item.get('type')}"
@@ -1165,6 +1183,28 @@ class ClaudeNormalizer(FormatNormalizer):
dropped["claude_system_unsupported"] = dropped.get("claude_system_unsupported", 0) + 1
return None, dropped, None
def _strip_anthropic_billing_headers(self, text: Any) -> tuple[str | None, int]:
normalized_text = str(text)
stripped_leading = normalized_text.lstrip()
if not stripped_leading.lower().startswith(self._ANTHROPIC_BILLING_HEADER_PREFIX):
return normalized_text, 0
lines = stripped_leading.splitlines(keepends=True)
idx = 0
removed_header_count = 0
while idx < len(lines) and lines[idx].strip().lower().startswith(
self._ANTHROPIC_BILLING_HEADER_PREFIX
):
removed_header_count += 1
idx += 1
while idx < len(lines) and not lines[idx].strip():
idx += 1
sanitized_text = "".join(lines[idx:])
if not sanitized_text.strip():
return None, removed_header_count
return sanitized_text, removed_header_count
def _join_instructions(self, instructions: list[InstructionSegment]) -> str | None:
parts = [seg.text for seg in instructions if seg.text]
joined = "\n\n".join(parts)