Fix hang when parsing unclosed comment #49
Conversation
ochafik
left a comment
ochafik
left a comment
There was a problem hiding this comment.
Thanks @cebtenzzre !
Trying to understand why tests are all red before this can go in. Some windows failures are clearly unrelated (hendrikmuhs/ccache-action#287), others, not sure yet.
include/minja/minja.hpp
Outdated
| } | ||
| } else if (std::regex_search(it, end, match, non_text_open_regex)) { | ||
| if (!match.position()) { | ||
| assert(match[0] == "{#"); |
There was a problem hiding this comment.
I'd rather avoid asserts for now, could you throw something here instead?
There was a problem hiding this comment.
Fixed. I think some kind of assert macro should be added at some point, though clearly out of scope for this PR.
| auto left = parseStringConcat(); | ||
| if (!left) throw std::runtime_error("Expected left side of 'logical compare' expression"); | ||
|
|
||
| static std::regex compare_tok(R"(==|!=|<=?|>=?|in\b|is\b|not[\r\n\s]+in\b)"); |
cebtenzzre
force-pushed
the
fix-opencomment-hang
branch
from
e97bb24 to
96ebff4
Compare
|
Is there any reason this repo uses plain merges instead of squash merges for PRs? A linear commit history like llama.cpp has is simpler and keeps fixup commits in PRs out of the main branch. |
Main reason has been most of my prs contained sometimes unrelated changes so far (just like this one has the welcome |
|
Thanks again @cebtenzzre ! |
2 participants
Continuation of #37
Since c73b53f, minja hangs when parsing the template
{#, e.g.:This program should exit with status code 1, but instead it hangs.
I found a fix that works even with the nonconforming Microsoft STL, and added a test.