Smart pointers in TextBuffer

.clang-tidy

@@ -0,0 +1,4 @@
+---
+Checks: "*, -clang-diagnostic-*-compat, -cppcoreguidelines-init-variables, -modernize-return-braced-init-list, -misc-unused-parameters, -misc-non-private-member-variables-in-classes, -llvmlibc-*, -llvm-header-guard, -llvm-include-order, -modernize-use-trailing-return-type, -readability-avoid-const-params-in-decls, -readability-convert-member-functions-to-static, -fuchsia-default-arguments-declarations, -fuchsia-default-arguments-calls, -*-uppercase-literal-suffix, -fuchsia-overloaded-operator, -google-build-using-namespace, -google-global-names-in-headers, -google-readability-todo, -*-else-after-return, -*-braces-around-statements, -bugprone-reserved-identifier, -cert-dcl37-c, -cert-dcl51-cpp"
+HeaderFilterRegex: ".*"
+FormatStyle: none

.gitignore

@@ -6,3 +6,5 @@ build
 /browser.js
 emsdk-portable
 package-lock.json
+
+benchmark/*.csv

.travis.yml

@@ -12,7 +12,7 @@ node_js:
   - "12.14.1"
 
 before_install:
-  - export CXX="g++-4.9" CC="gcc-4.9"
+  - export CXX="clang-10.0" CC="clang++-10.0"
 
 script:
   - npm run standard
@@ -31,5 +31,5 @@ addons:
     sources:
     - ubuntu-toolchain-r-test
     packages:
-    - gcc-4.9
-    - g++-4.9
+    - clang-10.0
+    - clang++-10.0

appveyor.yml

@@ -1,4 +1,4 @@
-image: Visual Studio 2015
+image: Visual Studio 2017
 
 environment:
   matrix:

benchmark/large-text-buffer.benchmark.js

@@ -1,56 +1,53 @@
-const http = require('http')
-const fs = require('fs')
-const unzip = require('unzip')
 const { TextBuffer } = require('..')
 
-const unzipper = unzip.Parse()
-
-const getText = () => {
-  return new Promise(resolve => {
-    console.log('fetching text file...')
-    const req = http.get({
-      hostname: 'www.acleddata.com',
-      port: 80,
-      // 51 MB text file
-      path: '/wp-content/uploads/2017/01/ACLED-Version-7-All-Africa-1997-2016_csv_dyadic-file.zip',
-      agent: false
-    }, res => {
-      res
-        .pipe(unzipper)
-        .on('entry', entry => {
-          let data = '';
-          entry.on('data', chunk => data += chunk);
-          entry.on('end', () => {
-            resolve(data)
-          });
-        })
-    })
-
-    req.end()
-  })
+const fs = require('fs')
+const {promisify} = require('util')
+const readFile = promisify(fs.readFile)
+const path = require('path')
+const download = require('download')
+const {performance} = require("perf_hooks")
+
+async function getText() {
+  const filePath = path.join(__dirname, '1000000 Sales Records.csv')
+  if (!fs.existsSync(filePath)) {
+    // 122MB file
+    await download(
+      'http://eforexcel.com/wp/wp-content/uploads/2017/07/1000000%20Sales%20Records.zip',
+      __dirname,
+      {extract: true}
+    )
+  }
+  return await readFile(filePath)
 }
 
-const timer = size => `Time to find "cat" in ${size} file`
-
-getText().then(txt => {
+getText().then(async (txt) => {
   const buffer = new TextBuffer()
 
-  console.log('running findWordsWithSubsequence tests...')
+  console.log('\n running large-text-buffer tests... \n')
 
-  const sizes = [['10b', 10], ['100b', 100], ['1kb', 1000], ['1MB', 1000000], ['51MB', 100000000]]
+  const sizes = [ ['100b', 100], ['1kb', 1000], ['1MB', 1000000], ['51MB', 100000000], ['119MB', txt.length]]
 
-  const test = size => {
-    const _timer = timer(size[0])
-    buffer.setText(txt.slice(0, size[1]))
-    console.time(_timer)
-    return buffer.findWordsWithSubsequence('cat', '', 100).then(sugs => {
-      console.timeEnd(_timer)
-    })
+  const test = async (word, size) => {
+    const ti2 = performance.now()
+    await buffer.findWordsWithSubsequence(word, '', 100)
+    const tf2 = performance.now()
+    console.log(`For ${size[0]} file, time to find "${word}" was: ${' '.repeat(50-word.length-size[0].length)} ${(tf2-ti2).toFixed(5)} ms`)
   }
+  for (const size of sizes) {
 
-  return sizes.reduce((promise, size) => {
-    return promise.then(() => test(size))
-  }, Promise.resolve())
+    const bufferText = txt.slice(0, size[1])
+
+    // benchmark buffer.setText
+    const ti1 = performance.now()
+    buffer.setText(bufferText)
+    const tf1 = performance.now()
+    console.log(`For ${size[0]} file, buffer.setText took ${' '.repeat(51-size[0].length)} ${(tf1-ti1).toFixed(5)} ms`)
+
+    for (const word of ["Morocco", "Austria", "France", "Liechtenstein", "Republic of the Congo", "Antigua and Barbuda", "Japan"]) {
+      await test(word, size)
+    }
+    console.log('\n')
+  }
 }).then(() => {
-  console.log('finished')
+  console.log(' large-text-buffer finished \n')
 })

benchmark/marker-index.benchmark.js

@@ -1,6 +1,10 @@
 'use strict';
 
+console.log(' running marker-index tests... \n')
+
 const Random = require('random-seed')
+const {performance} = require("perf_hooks")
+
 const {MarkerIndex} = require('..')
 const {traverse, traversalDistance, compare} = require('../test/js/helpers/point-helpers')
 
@@ -41,12 +45,13 @@ function runBenchmark () {
 }
 
 function profileOperations (name, operations) {
-  console.time(name)
+  const ti1 = performance.now()
   for (let i = 0, n = operations.length; i < n; i++) {
     const operation = operations[i]
     markerIndex[operation[0]].apply(markerIndex, operation[1])
   }
-  console.timeEnd(name)
+  const tf1 = performance.now()
+  console.log(`${name} ${' '.repeat(80-name.length)} ${(tf1-ti1).toFixed(3)} ms`)
 }
 
 function enqueueSequentialInsert () {
@@ -118,3 +123,5 @@ function getSplice () {
 }
 
 runBenchmark()
+
+console.log(' \n marker-index finished \n')

benchmark/text-buffer.benchmark.js

@@ -1,4 +1,8 @@
+console.log(' running text-buffer tests... \n')
+
 const assert = require('assert')
+const {performance} = require("perf_hooks")
+
 const {TextBuffer} = require('..')
 
 const text = 'abc def ghi jkl\n'.repeat(1024 * 1024)
@@ -8,14 +12,15 @@ const trialCount = 10
 
 function benchmarkSearch(description, pattern, expectedPosition) {
   let name = `Search for ${description} - TextBuffer`
-  console.time(name)
+  const ti1 = performance.now()
   for (let i = 0; i < trialCount; i++) {
-    assert.deepEqual(buffer.searchSync(pattern), expectedPosition)
+    assert.deepEqual(buffer.findSync(pattern), expectedPosition)
   }
-  console.timeEnd(name)
+  const tf1 = performance.now()
+  console.log(`${name} ${' '.repeat(80-name.length)} ${(tf1-ti1).toFixed(3)} ms`)
 
   name = `Search for ${description} - lines array`
-  console.time(name)
+  const ti2 = performance.now()
   const regex = new RegExp(pattern)
   for (let i = 0; i < trialCount; i++) {
     for (let row = 0, rowCount = lines.length; row < rowCount; row++) {
@@ -32,11 +37,14 @@ function benchmarkSearch(description, pattern, expectedPosition) {
       }
     }
   }
-  console.timeEnd(name)
-  console.log()
+  const tf2 = performance.now()
+  console.log(`${name} ${' '.repeat(80-name.length)} ${(tf2-ti2).toFixed(3)} ms`)
 }
 
 benchmarkSearch('simple non-existent pattern', '\t', null)
 benchmarkSearch('complex non-existent pattern', '123|456|789', null)
 benchmarkSearch('simple existing pattern', 'jkl', {start: {row: 0, column: 12}, end: {row: 0, column: 15}})
-benchmarkSearch('complex existing pattern', 'j\\w+', {start: {row: 0, column: 12}, end: {row: 0, column: 15}})
+benchmarkSearch('complex existing pattern', 'j\\w+', {start: {row: 0, column: 12}, end: {row: 0, column: 15}})
+
+
+console.log('\n text-buffer finished \n')

binding.gyp

@@ -67,7 +67,9 @@
     ],
 
     "variables": {
-        "tests": 0
+        "tests": 0,
+        "STANDARD": 17,
+        "MACOSX_DEPLOYMENT_TARGET": "10.8"
     },
 
     "conditions": [
@@ -100,20 +102,19 @@
                 "conditions": [
                     ['OS=="mac"', {
                         'cflags': [
-                            '-mmacosx-version-min=10.8'
+                            "-mmacosx-version-min=<(MACOSX_DEPLOYMENT_TARGET)"
                         ],
                         "xcode_settings": {
                             "GCC_ENABLE_CPP_EXCEPTIONS": "YES",
-                            'MACOSX_DEPLOYMENT_TARGET': '10.8',
+                            'MACOSX_DEPLOYMENT_TARGET': '<(MACOSX_DEPLOYMENT_TARGET)',
                         }
                     }]
                 ]
             }]
         }]
     ],
-
     "target_defaults": {
-        "cflags_cc": ["-std=c++11"],
+        "cflags_cc": [ "-std=c++<(STANDARD)" ],
         "conditions": [
             ['OS=="mac"', {
                 "xcode_settings": {
@@ -129,6 +130,40 @@
                     "NOMINMAX"
                 ],
             }]
-        ]
-    }
+        ],
+        'default_configuration': 'Release',
+        'configurations': {
+          # Release Settings
+          'Release': {
+            'defines': [ 'NDEBUG' ],
+            "cflags": [ "-fno-exceptions", "-Ofast" ],
+            "cflags_cc": [ "-fno-exceptions", "-Ofast", "-std=c++<(STANDARD)" ],
+            "xcode_settings": {
+              'GCC_OPTIMIZATION_LEVEL': '3', # stop gyp from defaulting to -Os
+              "CLANG_CXX_LIBRARY": "libc++",
+              "CLANG_CXX_LANGUAGE_STANDARD": "c++<(STANDARD)",
+              'MACOSX_DEPLOYMENT_TARGET': "<(MACOSX_DEPLOYMENT_TARGET)"
+            }, # XCODE
+            "msvs_settings": {
+              "VCCLCompilerTool": {
+                'ExceptionHandling': 0,               # /EHsc
+                'MultiProcessorCompilation': 'true',
+                'RuntimeTypeInfo': 'false',
+                'Optimization': 3,              # full optimizations /O2  ==  /Og /Oi /Ot /Oy /Ob2 /GF /Gy
+                'StringPooling': 'true',        # pool string literals
+                "AdditionalOptions": [
+                  # C++ standard
+                  "/std:c++<(STANDARD)",
+
+                  # Optimizations
+                  "/O2",
+                  # "/Ob3",  # aggressive inline
+                  "/GL",   # whole Program Optimization  # /LTCG is implied with /GL.
+                  "/DNDEBUG"  # turn off asserts
+                ],
+              }
+            } # MSVC
+          }, # Release
+        }, # configurations
+    } # target-defaults
 }

package.json

@@ -12,9 +12,12 @@
     "test:node": "mocha test/js/*.js",
     "test:browser": "SUPERSTRING_USE_BROWSER_VERSION=1 mocha test/js/*.js",
     "test": "npm run test:node && npm run test:browser",
-    "benchmark": "node benchmark/marker-index.benchmark.js",
+    "benchmark": "node benchmark/text-buffer.benchmark.js && node benchmark/marker-index.benchmark.js && node benchmark/large-text-buffer.benchmark.js",
     "prepublishOnly": "git submodule update --init --recursive && npm run build:browser",
-    "standard": "standard --recursive src test"
+    "standard": "standard --recursive src test",
+    "format": "clang-format -i src/core/*.cc src/core/*.h src/bindings/*.cc src/bindings/*.h src/bindings/em/*.cc src/bindings/em/*.h",
+    "tidy": "clang-tidy src/core/*.cc src/core/*.h src/bindings/*.cc src/bindings/*.h src/bindings/em/*.cc src/bindings/em/*.h",
+    "tidy:fix": "npm run tidy -- --fix --fix-errors"
   },
   "repository": {
     "type": "git",
@@ -35,11 +38,11 @@
   },
   "devDependencies": {
     "chai": "^2.0.0",
+    "download": "^8.0.0",
     "mocha": "^2.3.4",
     "random-seed": "^0.2.0",
     "standard": "^4.5.4",
-    "temp": "^0.8.3",
-    "unzip": "^0.1.11"
+    "temp": "^0.8.3"
   },
   "standard": {
     "global": [

src/bindings/bindings.cc

@@ -1,5 +1,5 @@
 #include "marker-index-wrapper.h"
-#include "nan.h"
+#include <nan.h>
 #include "patch-wrapper.h"
 #include "range-wrapper.h"
 #include "point-wrapper.h"

src/bindings/em/auto-wrap.h

@@ -7,7 +7,8 @@
 #include <emscripten/val.h>
 #include "flat_set.h"
 #include "marker-index.h"
-#include "optional.h"
+#include <optional>
+using std::optional;
 #include "point.h"
 #include "text.h"
 

src/bindings/marker-index-wrapper.cc

@@ -1,9 +1,10 @@
 #include "marker-index-wrapper.h"
 #include <unordered_map>
 #include "marker-index.h"
-#include "nan.h"
+#include <nan.h>
 #include "noop.h"
-#include "optional.h"
+#include <optional>
+using std::optional;
 #include "point-wrapper.h"
 #include "range.h"
 

src/bindings/marker-index-wrapper.h

@@ -1,6 +1,7 @@
-#include "nan.h"
+#include <nan.h>
 #include "marker-index.h"
-#include "optional.h"
+#include <optional>
+using std::optional;
 #include "range.h"
 
 class MarkerIndexWrapper : public Nan::ObjectWrap {
@@ -36,6 +37,6 @@ class MarkerIndexWrapper : public Nan::ObjectWrap {
   static void find_ending_at(const Nan::FunctionCallbackInfo<v8::Value> &info);
   static void find_boundaries_after(const Nan::FunctionCallbackInfo<v8::Value> &info);
   static void dump(const Nan::FunctionCallbackInfo<v8::Value> &info);
-  MarkerIndexWrapper(unsigned seed);
   MarkerIndex marker_index;
+  explicit MarkerIndexWrapper(unsigned seed);
 };

src/bindings/noop.h

@@ -1,5 +1,5 @@
 #pragma once
 
-#include "nan.h"
+#include <nan.h>
 
 static void noop(const Nan::FunctionCallbackInfo<v8::Value>&) {}

src/bindings/number-conversion.h

@@ -1,8 +1,9 @@
 #ifndef SUPERSTRING_NUMBER_CONVERSION_H
 #define SUPERSTRING_NUMBER_CONVERSION_H
 
-#include "nan.h"
-#include "optional.h"
+#include <nan.h>
+#include <optional>
+using std::optional;
 
 namespace number_conversion {
   template<typename T>