From 14c4e9ccc6112077a51b4eacbcdca8f0702446e2 Mon Sep 17 00:00:00 2001
From: Michal Moskal <michal@moskal.me>
Date: Tue, 3 Sep 2024 19:14:58 -0700
Subject: [PATCH] bump derivre; add logging

---
 parser/Cargo.toml             | 2 +-
 parser/src/earley/regexvec.rs | 8 +++++++-
 rust/Cargo.lock               | 2 +-
 3 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/parser/Cargo.toml b/parser/Cargo.toml
index 709a2a6d..ac240cfd 100644
--- a/parser/Cargo.toml
+++ b/parser/Cargo.toml
@@ -5,7 +5,7 @@ edition = "2021"
 
 [dependencies]
 toktrie = { git = "https://github.com/microsoft/toktrie", rev = "7cb20cd594428f5397f6159115d5e070e8c0935f" }
-derivre = { git = "https://github.com/microsoft/derivre", rev = "9bf0f4842b012b963812a6099acf2cb86406adb2" }
+derivre = { git = "https://github.com/microsoft/derivre", rev = "677799cfa3abb1a33928d40f81972b3a9c965e4f" }
 serde = { version = "1.0.192", features = ["derive"] }
 serde_json = "1.0.108"
 anyhow = "1.0.75"
diff --git a/parser/src/earley/regexvec.rs b/parser/src/earley/regexvec.rs
index 53f98990..4cebfe96 100644
--- a/parser/src/earley/regexvec.rs
+++ b/parser/src/earley/regexvec.rs
@@ -132,6 +132,8 @@ impl RegexVec {
         lexeme_idx: usize,
         mut budget: u64,
     ) -> Result<bool> {
+        let budget0 = budget;
+        let t0 = instant::Instant::now();
         assert!(self.subsume_possible(state));
         let small = self.rx_list[lexeme_idx];
         self.set_fuel(u64::MAX);
@@ -142,13 +144,17 @@ impl RegexVec {
             let c0 = self.exprs.cost();
             let not_contained =
                 self.relevance
-                    .is_relevant_limited(&mut self.exprs, check, budget)?;
+                    .is_non_empty_limited(&mut self.exprs, check, budget)?;
             if !not_contained {
                 return Ok(true);
             }
             let cost = self.exprs.cost() - c0;
             budget = budget.saturating_sub(cost);
         }
+        let cost = budget0 - budget;
+        if false && cost > 10 {
+            println!("check_subsume: {}/{} {:?}", cost, budget0, t0.elapsed());
+        }
         Ok(false)
     }
 
diff --git a/rust/Cargo.lock b/rust/Cargo.lock
index f1e00519..ebe3f049 100644
--- a/rust/Cargo.lock
+++ b/rust/Cargo.lock
@@ -65,7 +65,7 @@ checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
 [[package]]
 name = "derivre"
 version = "0.1.0"
-source = "git+https://github.com/microsoft/derivre?rev=9bf0f4842b012b963812a6099acf2cb86406adb2#9bf0f4842b012b963812a6099acf2cb86406adb2"
+source = "git+https://github.com/microsoft/derivre?rev=677799cfa3abb1a33928d40f81972b3a9c965e4f#677799cfa3abb1a33928d40f81972b3a9c965e4f"
 dependencies = [
  "ahash",
  "anyhow",