From 45ac5e34fcefa38552804b124fcae3c20ef22e41 Mon Sep 17 00:00:00 2001
From: Serg Gini <kornburn@gmail.com>
Date: Mon, 5 Jun 2023 20:51:47 +0300
Subject: [PATCH 1/2] Optimized solution based on R base

Uses PCRE engine and Bytes for text data, instead of UTF
---
 Dockerfile         |  4 ++++
 r/README.md        |  8 ++++++++
 r/benchmark.R      | 28 ++++++++++++++++++++++++++++
 run-benchmarks.php |  1 +
 4 files changed, 41 insertions(+)
 create mode 100644 r/README.md
 create mode 100644 r/benchmark.R

diff --git a/Dockerfile b/Dockerfile
index 1af642f..357b17d 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -132,6 +132,10 @@ RUN wget -q https://downloads.python.org/pypy/pypy3.6-v7.3.3-linux64.tar.bz2 -O
     ln -s /opt/pypy3/bin/pypy3 /usr/local/bin/pypy3 && \
     rm pypy3.6-v7.3.3-linux64.tar.bz2
 
+## R
+RUN apt-get install -yq --no-install-recommends \
+        r-base r-cran-readr
+
 ## Ruby
 RUN apt-get install -yq --no-install-recommends \
         ruby-full
diff --git a/r/README.md b/r/README.md
new file mode 100644
index 0000000..4a9dd44
--- /dev/null
+++ b/r/README.md
@@ -0,0 +1,8 @@
+# R Regex Benchmark
+
+## How to run
+
+```sh
+# R
+RScript --vanilla benchmark.R <filename>
+```
diff --git a/r/benchmark.R b/r/benchmark.R
new file mode 100644
index 0000000..49591fd
--- /dev/null
+++ b/r/benchmark.R
@@ -0,0 +1,28 @@
+#!/usr/bin/env Rscript
+library(readr)
+
+measure <- function(data, pattern) {
+    start <- Sys.time()
+    res <- length(regmatches(data, gregexpr(pattern, data, perl = TRUE, useBytes = TRUE))[[1]])
+    end <- Sys.time()
+    s <- end - start
+    out <- paste0(format(as.numeric(s)*1e3, digits = 3), " - ", res)
+    print(out)
+}
+
+args = commandArgs(trailingOnly=TRUE)
+
+if (length(args)==0) {
+  stop("At least one argument must be supplied (input file).n", call.=FALSE)
+}
+
+file_str <- read_file(args[1])
+
+# Email
+measure(file_str, r"{[\w\\.+-]+@[\w\.-]+\.[\w\.-]+}")
+
+# URI
+measure(file_str, r"{[\w]+://[^/\s?#]+[^\s?#]+(?:\?[^\s#]*)?(?:#[^\s]*)?}")
+
+# IPv4
+measure(file_str, r"{(?:(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9])\.){3}(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9])}") 
diff --git a/run-benchmarks.php b/run-benchmarks.php
index 86756b0..7e80cc3 100644
--- a/run-benchmarks.php
+++ b/run-benchmarks.php
@@ -46,6 +46,7 @@
     'Python 3'     => 'python3.6 python/benchmark.py',
     'Python PyPy2' => 'pypy2 python/benchmark.py',
     'Python PyPy3' => 'pypy3 python/benchmark.py',
+    'R'            => 'RScript --vanilla r/benchmark.R',
     'Ruby'         => 'ruby ruby/benchmark.rb',
     'Rust'         => 'rust/target/release/benchmark',
 ];

From 875de7e5c9dfa73b5bd7642c9cfc99a13ae79611 Mon Sep 17 00:00:00 2001
From: Serg Gini <kornburn@gmail.com>
Date: Mon, 5 Jun 2023 20:58:21 +0300
Subject: [PATCH 2/2] Fixed typo

---
 r/README.md        | 2 +-
 run-benchmarks.php | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/r/README.md b/r/README.md
index 4a9dd44..77520cd 100644
--- a/r/README.md
+++ b/r/README.md
@@ -4,5 +4,5 @@
 
 ```sh
 # R
-RScript --vanilla benchmark.R <filename>
+Rscript --vanilla benchmark.R <filename>
 ```
diff --git a/run-benchmarks.php b/run-benchmarks.php
index 7e80cc3..78e35b0 100644
--- a/run-benchmarks.php
+++ b/run-benchmarks.php
@@ -46,7 +46,7 @@
     'Python 3'     => 'python3.6 python/benchmark.py',
     'Python PyPy2' => 'pypy2 python/benchmark.py',
     'Python PyPy3' => 'pypy3 python/benchmark.py',
-    'R'            => 'RScript --vanilla r/benchmark.R',
+    'R'            => 'Rscript --vanilla r/benchmark.R',
     'Ruby'         => 'ruby ruby/benchmark.rb',
     'Rust'         => 'rust/target/release/benchmark',
 ];