fix: avoid 'parse empty input' log for streaming tool-call parsing

- json-partial: when SAX error is at position 0, return false without
  calling json::parse or logging 'Failed to parse up to error' (covers
  all formats: GLM 4.5, GPT-OSS, Granite, Hermes, etc.)
- chat-parser-xml-toolcall: only call try_consume_json when remainder
  looks like start of JSON value (", {, [, digit, -, true/false/null);
  otherwise treat as plain text/partial (vLLM-style, avoids SAX
  error-at-position-0 for e.g. 'exp' before 'explore')

Co-authored-by: Cursor <cursoragent@cursor.com>
This commit is contained in:
Gunther Schulz 2026-02-14 01:20:26 +01:00
parent f9b571dc96
commit 329a0015a8
2 changed files with 47 additions and 0 deletions

View File

@ -498,6 +498,48 @@ inline bool parse_xml_tool_calls(common_chat_msg_parser & builder, const struct
}
auto val_start = builder.pos();
// vLLM-style: only try to parse value when there is content; empty/whitespace or non-JSON start = incomplete (avoids "parse empty input" log)
{
const auto & inp = builder.input();
const size_t rem_len = (val_start < inp.size()) ? (inp.size() - val_start) : 0;
std::string_view rest_sv(inp.data() + val_start, rem_len);
if (rest_sv.empty() || all_space(rest_sv)) {
gen_partial_args([&](auto & rest, auto & needle) { arguments[key] = (form.trim_raw_argval ? string_strip(rest) : rest) + needle; });
throw common_chat_msg_partial_exception(
"Expected " + gbnf_format_literal(form.val_end) +
" after " + gbnf_format_literal(form.key_val_sep) +
(form.key_val_sep2 ? " " + gbnf_format_literal(*form.key_val_sep2) : "")
);
}
// Only call try_consume_json when remainder looks like start of a JSON value (avoids SAX error-at-position-0 → "empty input" log)
size_t pos = 0;
while (pos < rest_sv.size() && std::isspace(static_cast<unsigned char>(rest_sv[pos]))) { ++pos; }
if (pos >= rest_sv.size()) {
gen_partial_args([&](auto & rest, auto & needle) { arguments[key] = (form.trim_raw_argval ? string_strip(rest) : rest) + needle; });
throw common_chat_msg_partial_exception(
"Expected " + gbnf_format_literal(form.val_end) +
" after " + gbnf_format_literal(form.key_val_sep) +
(form.key_val_sep2 ? " " + gbnf_format_literal(*form.key_val_sep2) : "")
);
}
std::string_view rest_trim = rest_sv.substr(pos);
char c = rest_trim[0];
bool looks_like_json = (c == '"' || c == '{' || c == '[' || (c >= '0' && c <= '9') || c == '-');
if (!looks_like_json) {
if (c == 't') looks_like_json = (rest_trim.size() <= 4 && std::string_view("true").substr(0, rest_trim.size()) == rest_trim);
else if (c == 'f') looks_like_json = (rest_trim.size() <= 5 && std::string_view("false").substr(0, rest_trim.size()) == rest_trim);
else if (c == 'n') looks_like_json = (rest_trim.size() <= 4 && std::string_view("null").substr(0, rest_trim.size()) == rest_trim);
}
if (!looks_like_json) {
gen_partial_args([&](auto & rest, auto & needle) { arguments[key] = (form.trim_raw_argval ? string_strip(rest) : rest) + needle; });
throw common_chat_msg_partial_exception(
"Expected " + gbnf_format_literal(form.val_end) +
" after " + gbnf_format_literal(form.key_val_sep) +
(form.key_val_sep2 ? " " + gbnf_format_literal(*form.key_val_sep2) : "")
);
}
}
// Test if arg_val is a partial JSON
std::optional<common_json> value_json = std::nullopt;
if (!form.raw_argval || !*form.raw_argval) {

View File

@ -120,6 +120,11 @@ bool common_json_parse(
auto temptative_end = it + err_loc.position;
// LOG_DBG("Error at position %zu (is_end = %s): %s\n", err_loc.position, temptative_end == end ? "true" : "false", err_loc.exception_message.c_str());
// Avoid parsing and logging "empty input" when error is at position 0 (e.g. streaming partial/invalid JSON)
if (temptative_end == it) {
return false;
}
auto input = std::string(it, temptative_end);
try {
out.json = json::parse(input);