From 4bd86120d41e3b01433004bf761beccb7f3a0167 Mon Sep 17 00:00:00 2001 From: Jon Huhn Date: Mon, 10 Jun 2024 06:23:03 -0500 Subject: fix(glob): handle overlapping `{}` condition elements #29236 This change fixes an issue where glob patterns like `{a,ab}` would not match `ab` because the first option `a` matches, then the end of the string is expected but `b` is found, and LPeg does not backtrack to try the next option `ab` which would match. The fix here is to also append the rest of the pattern to the generated LPeg pattern for each option. This changes a glob `{a,ab}` from being parsed as ("a" or "ab") "end of string" to ("a" "end of string" or "ab" "end of string") Here, matching against `ab` would try the first option, fail to match, then proceed to the next option, and match. The sacrifice this change makes is dropping support for nested `{}` conditions, which VSCode doesn't seem to support or test AFAICT. Fixes #28931 Co-authored-by: Sergey Slipchenko --- runtime/lua/vim/glob.lua | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) (limited to 'runtime/lua/vim/glob.lua') diff --git a/runtime/lua/vim/glob.lua b/runtime/lua/vim/glob.lua index ad4a915a94..6de2bc3e94 100644 --- a/runtime/lua/vim/glob.lua +++ b/runtime/lua/vim/glob.lua @@ -29,8 +29,10 @@ function M.to_lpeg(pattern) return patt end - local function add(acc, a) - return acc + a + local function condlist(conds, after) + return vim.iter(conds):fold(P(false), function(acc, cond) + return acc + cond * after + end) end local function mul(acc, m) @@ -63,15 +65,14 @@ function M.to_lpeg(pattern) * C(P('!') ^ -1) * Ct(Ct(C(P(1)) * P('-') * C(P(1) - P(']'))) ^ 1 * P(']')) / class, - CondList = P('{') * Cf(V('Cond') * (P(',') * V('Cond')) ^ 0, add) * P('}'), + CondList = P('{') * Ct(V('Cond') * (P(',') * V('Cond')) ^ 0) * P('}') * V('Pattern') / condlist, -- TODO: '*' inside a {} condition is interpreted literally but should probably have the same -- wildcard semantics it usually has. -- Fixing this is non-trivial because '*' should match non-greedily up to "the rest of the -- pattern" which in all other cases is the entire succeeding part of the pattern, but at the end of a {} -- condition means "everything after the {}" where several other options separated by ',' may -- exist in between that should not be matched by '*'. - Cond = Cf((V('Ques') + V('Class') + V('CondList') + (V('Literal') - S(',}'))) ^ 1, mul) - + Cc(P(0)), + Cond = Cf((V('Ques') + V('Class') + V('Literal') - S(',}')) ^ 1, mul) + Cc(P(0)), Literal = P(1) / P, End = P(-1) * Cc(P(-1)), }) -- cgit From b109b1abce8c86f80aea06948b32e35a70daa0b0 Mon Sep 17 00:00:00 2001 From: Zoltán Nyikos Date: Sat, 6 Jul 2024 11:40:08 +0200 Subject: fix(glob): avoid `subcapture nesting too deep` error (#29520) Use Cmt to evaluate Cond and Elem during match to avoid building the nested capture structure later. --- runtime/lua/vim/glob.lua | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 deletions(-) (limited to 'runtime/lua/vim/glob.lua') diff --git a/runtime/lua/vim/glob.lua b/runtime/lua/vim/glob.lua index 6de2bc3e94..22073b15c8 100644 --- a/runtime/lua/vim/glob.lua +++ b/runtime/lua/vim/glob.lua @@ -1,6 +1,6 @@ local lpeg = vim.lpeg local P, S, V, R, B = lpeg.P, lpeg.S, lpeg.V, lpeg.R, lpeg.B -local C, Cc, Ct, Cf = lpeg.C, lpeg.Cc, lpeg.Ct, lpeg.Cf +local C, Cc, Ct, Cf, Cmt = lpeg.C, lpeg.Cc, lpeg.Ct, lpeg.Cf, lpeg.Cmt local M = {} @@ -47,13 +47,22 @@ function M.to_lpeg(pattern) return (-after * P(1)) ^ 0 * after end + -- luacheck: push ignore s + local function cut(s, idx, match) + return idx, match + end + -- luacheck: pop + local p = P({ 'Pattern', Pattern = V('Elem') ^ -1 * V('End'), - Elem = Cf( - (V('DStar') + V('Star') + V('Ques') + V('Class') + V('CondList') + V('Literal')) - * (V('Elem') + V('End')), - mul + Elem = Cmt( + Cf( + (V('DStar') + V('Star') + V('Ques') + V('Class') + V('CondList') + V('Literal')) + * (V('Elem') + V('End')), + mul + ), + cut ), DStar = (B(pathsep) + -B(P(1))) * P('**') @@ -72,7 +81,7 @@ function M.to_lpeg(pattern) -- pattern" which in all other cases is the entire succeeding part of the pattern, but at the end of a {} -- condition means "everything after the {}" where several other options separated by ',' may -- exist in between that should not be matched by '*'. - Cond = Cf((V('Ques') + V('Class') + V('Literal') - S(',}')) ^ 1, mul) + Cc(P(0)), + Cond = Cmt(Cf((V('Ques') + V('Class') + V('Literal') - S(',}')) ^ 1, mul), cut) + Cc(P(0)), Literal = P(1) / P, End = P(-1) * Cc(P(-1)), }) -- cgit