Add benchmarks for D's ctRegex

ethanpailes · Jan 1, 2018 · ed174df · ed174df
1 parent 49f2a3d
commit ed174df
Show file tree

Hide file tree

Showing 6 changed files with 133 additions and 2 deletions.
diff --git a/bench/Cargo.toml b/bench/Cargo.toml
@@ -49,7 +49,9 @@ re-onig = ["onig"]
 re-re2 = []
 re-dphobos = []
 re-dphobos-dmd = ["re-dphobos"]
+re-dphobos-dmd-ct = ["re-dphobos-dmd"]
 re-dphobos-ldc = ["re-dphobos"]
+re-dphobos-ldc-ct = ["re-dphobos-ldc"]
 re-rust = []
 re-rust-bytes = []
 re-tcl = []

diff --git a/bench/build.rs b/bench/build.rs
@@ -42,14 +42,19 @@ fn main() {
 
         let out_dir = env::var("OUT_DIR").unwrap();
         let out_file = &format!("-of={}/libdphobos-dmd.a", out_dir);
+        let is_compile_time = env::var("CARGO_FEATURE_RE_DPHOBOS_DMD_CT").is_ok();
+        let extra_args = if is_compile_time { vec!["-version=CtRegex"] } else { vec![] };
 
         let res = process::Command::new("dmd")
             .arg("-w")
             .arg("-lib")
             .arg("-O")
             .arg("-release")
             .arg("-inline")
+            .arg("-Isrc/ffi")
+            .args(extra_args)
             .arg("src/ffi/d_phobos.d")
+            .arg("src/ffi/d_phobos_ct.d")
             .arg(out_file)
             .output()
             .expect("unable to compile dphobos-regex (dmd)");
@@ -75,13 +80,19 @@ fn main() {
         let out_dir = env::var("OUT_DIR").unwrap();
         let out_file = &format!("-of={}/libdphobos-ldc.a", out_dir);
 
+        let is_compile_time = env::var("CARGO_FEATURE_RE_DPHOBOS_LDC_CT").is_ok();
+        let extra_args = if is_compile_time { vec!["-d-version=CtRegex"] } else { vec![] };
+
         let res = process::Command::new("ldc")
             .arg("-w")
             .arg("-lib")
             .arg("-O3")
             .arg("-release")
             .arg("-mcpu=native")
+            .arg("-Isrc/ffi")
+            .args(extra_args)
             .arg("src/ffi/d_phobos.d")
+            .arg("src/ffi/d_phobos_ct.d")
             .arg(out_file)
             .output()
             .expect("unable to compile dphobos-regex (ldc)");

diff --git a/bench/run b/bench/run
@@ -1,7 +1,7 @@
 #!/bin/bash
 
 usage() {
-  echo "Usage: $(basename $0) [dphobos-dmd | dphobos-ldc | rust | rust-bytes | pcre1 | pcre2 | re2 | onig | tcl ]" >&2
+  echo "Usage: $(basename $0) [dphobos-dmd | dphobos-ldc | dphobos-dmd-ct | dphobos-ldc-ct | rust | rust-bytes | pcre1 | pcre2 | re2 | onig | tcl ]" >&2
   exit 1
 }
 
@@ -23,6 +23,12 @@ case $which in
   dphobos-ldc)
     exec cargo bench --bench bench --features re-dphobos-ldc "$@"
     ;;
+  dphobos-dmd-ct)
+    exec cargo bench --bench bench --features re-dphobos-dmd-ct "$@"
+    ;;
+  dphobos-ldc-ct)
+    exec cargo bench --bench bench --features re-dphobos-ldc-ct "$@"
+    ;;
   rust)
     exec cargo bench --bench bench --features re-rust "$@"
     ;;

diff --git a/bench/src/ffi/d_phobos.d b/bench/src/ffi/d_phobos.d
@@ -9,6 +9,8 @@ import std.typecons;
 
 import std.stdio;
 
+import d_phobos_ct;
+
 auto rustRegexToD(string regex) {
     auto flags = "g";
     if (regex.startsWith("(?i)")) {
@@ -27,7 +29,12 @@ extern(C):
 void* d_phobos_regex_new(string s) {
     auto r = cast(Regex!char*)malloc(Regex!char.sizeof);
 
-    *r = regex(rustRegexToD(s).expand);
+    version(CtRegex) {
+        auto ctR = getCtRegex();
+        *r = ctR[s];
+    } else {
+        *r = regex(rustRegexToD(s).expand);
+    }
 
     return r;
 }

diff --git a/bench/src/ffi/d_phobos_ct.d b/bench/src/ffi/d_phobos_ct.d
@@ -0,0 +1,103 @@
+module d_phobos_ct;
+
+version(CtRegex):
+
+static immutable PATTERNS = [
+    r"y", // misc::literal
+    r".y", // misc::not_literal
+    "[abcdw]", // misc::match_class
+    "[ac]", // misc::match_class_in_range
+    r"\p{L}", // misc::match_class_unicode / sherlock::letters
+    r"^zbc(d|e)", // misc::anchored_literal_long_non_match / misc::anchored_literal_short_non_match
+    r"^.bc(d|e)", // misc::anchored_literal_short_match / misc::anchored_literal_long_match
+    r"^.bc(d|e)*$", // misc::one_pass_short
+    r".bc(d|e)*$", // misc::one_pass_short_not
+    r"^abcdefghijklmnopqrstuvwxyz.*$", // misc::one_pass_long_prefix
+    r"^.bcdefghijklmnopqrstuvwxyz.*$", // misc::one_pass_long_prefix_not
+    r"aaaaaaaaaaaaaaaaaaaaaaaaaaaaaab", // misc::long_needle1
+    r"bbbbbbbbbbbbbbbbbbbbbbbbbbbbbba", // misc::long_needle2
+    r"[r-z].*bcdefghijklmnopq", // misc::reverse_suffix_no_quadratic
+    "ABCDEFGHIJKLMNOPQRSTUVWXYZ$", // misc::easy0
+    r"A[AB]B[BC]C[CD]D[DE]E[EF]F[FG]G[GH]H[HI]I[IJ]J$", // misc::easy1
+    r"[XYZ]ABCDEFGHIJKLMNOPQRSTUVWXYZ$", // misc::medium
+    r"[ -~]*ABCDEFGHIJKLMNOPQRSTUVWXYZ$", // misc::hard
+    r"[ -~]*ABCDEFGHIJKLMNOPQRSTUVWXYZ.*", // misc::reallyhard
+    r"\w+\s+Holmes", // misc::reallyhard2
+    // This causes compile times to go from ~40s to ~9m with dmd 2.077.1
+    //r"a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa", // misc::no_exponential
+    r">[^\n]*\n|\n", // dna::find_new_lines
+    r"agggtaaa|tttaccct", // dna::variant1
+    r"[cgt]gggtaaa|tttaccc[acg]", // dna::variant2
+    r"a[act]ggtaaa|tttacc[agt]t", // dna::variant3
+    r"ag[act]gtaaa|tttac[agt]ct", // dna::variant4
+    r"agg[act]taaa|ttta[agt]cct", // dna::variant5
+    r"aggg[acg]aaa|ttt[cgt]ccct", // dna::variant6
+    r"agggt[cgt]aa|tt[acg]accct", // dna::variant7
+    r"agggta[cgt]a|t[acg]taccct", // dna::variant8
+    r"agggtaa[cgt]|[acg]ttaccct", // dna::variant9
+    r"B", // dna::subst1
+    r"D", // dna::subst2
+    r"H", // dna::subst3
+    r"K", // dna::subst4
+    r"M", // dna::subst5
+    r"N", // dna::subst6
+    r"R", // dna::subst7
+    r"S", // dna::subst8
+    r"V", // dna::subst9
+    r"W", // dna::subst10
+    r"Y", // dna::subst11
+    r"Sherlock", // sherlock::name_sherlock
+    r"Holmes", // sherlock::name_holmes
+    r"Sherlock Holmes", // sherlock::name_sherlock_holmes
+    r"(?i)Sherlock", // sherlock::name_sherlock_nocase
+    r"(?i)Holmes", // sherlock::name_holmes_nocase
+    r"(?i)Sherlock Holmes", // sherlock::name_sherlock_holmes_nocase
+    r"Sherlock\s+Holmes", // sherlock::name_whitespace
+    r"Sherlock|Street", // sherlock::name_alt1
+    r"Sherlock|Holmes", // sherlock::name_alt2
+    r"Sherlock|Holmes|Watson|Irene|Adler|John|Baker", // sherlock::name_alt3
+    r"(?i)Sherlock|Holmes|Watson|Irene|Adler|John|Baker", // sherlock::name_alt3_nocase
+    r"Sher[a-z]+|Hol[a-z]+", // sherlock::name_alt4
+    r"(?i)Sher[a-z]+|Hol[a-z]+", // sherlock::name_alt4_nocase
+    r"Sherlock|Holmes|Watson", // sherlock::name_alt5
+    r"(?i)Sherlock|Holmes|Watson", // sherlock::name_alt5_nocase
+    r"zqj", // sherlock::no_match_uncommon
+    r"aqj", // sherlock::no_match_common
+    r"aei", // sherlock::no_match_really_common
+    r"the", // sherlock::the_lower
+    r"The", // sherlock::the_upper
+    r"(?i)the", // sherlock::the_nocase
+    r"the\s+\w+", // sherlock::the_whitespace
+    r"\p{Lu}", // sherlock::letters_upper
+    r"\p{Ll}", // sherlock::letters_lower
+    r"\w+", // sherlock::words
+    r"\w+\s+Holmes", // sherlock::before_holmes
+    r"\w+\s+Holmes\s+\w+", // sherlock::before_after_holmes
+    r"Holmes.{0,25}Watson|Watson.{0,25}Holmes", // sherlock::holmes_cochar_watson
+    r"Holmes(?:\s*.+\s*){0,10}Watson|Watson(?:\s*.+\s*){0,10}Holmes", // sherlock::holmes_coword_watson
+    `["'][^"']{0,30}[?!.]["']`, // sherlock::quotes
+    r"(?m)^Sherlock Holmes|Sherlock Holmes$", // sherlock::line_boundary_sherlock_holmes
+    r"\b\w+n\b", // sherlock::word_ending_n
+    r"[a-q][^u-z]{13}x", // sherlock::repeated_class_negation
+    r"[a-zA-Z]+ing", // sherlock::ing_suffix
+    r"\s[a-zA-Z]{0,12}ing\s", // sherlock::ing_suffix_limited_space
+];
+
+public auto getCtRegex() {
+    import std.regex;
+    import std.string;
+
+    Regex!char[string] aa;
+
+    static foreach (pattern; PATTERNS) {
+        static if (pattern.startsWith("(?i)")) {
+            aa[pattern] = ctRegex!(pattern[4..$], "gi");
+        } else static if (pattern.startsWith("(?m)")) {
+            aa[pattern] = ctRegex!(pattern[4..$], "gm");
+        } else {
+            aa[pattern] = ctRegex!(pattern, "g");
+        }
+    }
+
+    return aa;
+}
diff --git a/bench/src/misc.rs b/bench/src/misc.rs
@@ -19,6 +19,8 @@ use {Regex, Text};
 #[cfg(not(feature = "re-onig"))]
 #[cfg(not(feature = "re-pcre1"))]
 #[cfg(not(feature = "re-pcre2"))]
+#[cfg(not(feature = "re-dphobos-dmd-ct"))]
+#[cfg(not(feature = "re-dphobos-ldc-ct"))]
 bench_match!(no_exponential, {
     format!(
         "{}{}",