Skip to content

fix: bound plain-text runs in parse_prompt_attention regex #1919

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
fszontagh wants to merge 2 commits into
base: master
Choose a base branch
from
Open

fix: bound plain-text runs in parse_prompt_attention regex #1919

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
324ff67
fix: bound plain-text runs in parse_prompt_attention regex
fszontagh Aug 27, 2026
bf552bc
Merge remote-tracking branch 'upstream/master' into fix/prompt-attent…
fszontagh Aug 27, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 5 additions & 1 deletion src/core/util.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -801,7 +801,11 @@ std::vector<std::pair<std::string, float>> parse_prompt_attention(const std::str
float round_bracket_multiplier = 1.1f;
float square_bracket_multiplier = 1 / 1.1f;

std::regex re_attention(R"(\\\(|\\\)|\\\[|\\\]|\\\\|\\|\(|\[|:([+-]?[.\d]+)\)|\)|\]|\bBREAK\b|[^\\()\[\]:B]+|:|\bB)");
// libstdc++ std::regex recurses per matched character; unbounded runs overflow
// the stack. Split runs are merged back by the equal-weight pass below.
const int max_plain_text_run = 1024;
std::regex re_attention(R"(\\\(|\\\)|\\\[|\\\]|\\\\|\\|\(|\[|:([+-]?[.\d]+)\)|\)|\]|\bBREAK\b|[^\\()\[\]:B]{1,)" +
std::to_string(max_plain_text_run) + R"(}|:|\bB)");
std::regex re_break(R"(\s*\bBREAK\b\s*)");

auto multiply_range = [&](int start_position, float multiplier) {
Expand Down
Loading