@bonniernews/atlas-html-stream - npm Package Compare versions

Comparing version 1.2.2 to 2.0.0

package.json

 {
   "name": "@bonniernews/atlas-html-stream",
-  "version": "1.2.2",
+  "version": "2.0.0",
   "description": "A super fast html-parser stream that outputs tag, text and closing nodes.",
   "scripts": {
     "test-server": "mocha --colors --watch --recursive",
-    "test": "mocha --colors --recursive"
+    "test": "mocha --colors --recursive",
+    "posttest": "eslint . --cache"
   },
@@ -26,4 +27,5 @@ "main": "./src/HtmlParser.js",
   "devDependencies": {
-    "chai": "^4.1.2",
-    "mocha": "^5.2.0"
+    "chai": "^4.3.4",
+    "eslint": "^8.4.1",
+    "mocha": "^9.1.3"
   },
@@ -33,5 +35,2 @@ "engines": {
   },
-  "dependencies": {
-    "@bonniernews/atlas-seq-matcher": "^1.0.3"
-  },
   "homepage": "https://github.com/BonnierNews/atlas-html-stream#readme",
@@ -42,3 +41,6 @@ "bugs": "https://github.com/BonnierNews/atlas-html-stream/issues",
     "url": "https://github.com/BonnierNews/atlas-html-stream.git"
-  }
+  },
+  "files": [
+    "src"
+  ]
 }

README.md

@@ -1,3 +0,6 @@
-# atlas-html-stream
+atlas-html-stream
+=================
 
+[![Built latest](https://github.com/BonnierNews/atlas-html-stream/actions/workflows/build-latest.yaml/badge.svg)](https://github.com/BonnierNews/atlas-html-stream/actions/workflows/build-latest.yaml)
+
 Fork of [atlas-html-stream](https://github.com/atlassubbed/atlas-html-stream) which is a super fast html-parser stream that outputs tag,
@@ -7,3 +10,3 @@ text and closing nodes. See `CHANGELOG.md` for changes introduced in the fork.
 ---
-## install 
+## install
 
@@ -16,3 +19,3 @@ ```
 
-I didn't like htmlparser2's streaming API and I wanted an html parser that collpased text-node whitespace by default. I also wanted to see if I could write a faster parser. 
+I didn't like htmlparser2's streaming API and I wanted an html parser that collpased text-node whitespace by default. I also wanted to see if I could write a faster parser.
 
@@ -41,3 +44,3 @@ ## performance
 
-## examples 
+## examples
 
@@ -130,5 +133,5 @@ Using this parser is really easy since it `extends` the stream interface -- all you need to do is pipe or write HTML to it and listen to `"data"` events:
 const commentNodes = parseHtml(`
-  <!-- 
-    this is 
-    a comment 
+  <!--
+    this is
+    a comment
   -->
@@ -150,5 +153,5 @@ `)
 
-      <b> 
+      <b>
         Hola
-          </b>  
+          </b>
 `)
@@ -177,3 +180,3 @@ nodes.forEach(n => console.log(n));
 
-I'd like to make this thing even faster. The parsing itself takes about `3.5 ms/file` (using [htmlparser-benchmark](https://github.com/AndreasMadsen/htmlparser-benchmark)) on my machine. Pushing nodes as `data` events to our stream adds around 40% more processing time, which is why the benchmark above shows around `4.9 ms/file` -- this can't be avoided, because we *want* the streaming interface. 
+I'd like to make this thing even faster. The parsing itself takes about `3.5 ms/file` (using [htmlparser-benchmark](https://github.com/AndreasMadsen/htmlparser-benchmark)) on my machine. Pushing nodes as `data` events to our stream adds around 40% more processing time, which is why the benchmark above shows around `4.9 ms/file` -- this can't be avoided, because we *want* the streaming interface.
 
@@ -180,0 +183,0 @@ The `SeqMatcher` slows down this parser (checking comment, script and style nodes); there might be a faster way to handle these special nodes.

src/HtmlParser.js

@@ -1,123 +0,186 @@
-const SeqMatcher = require("@bonniernews/atlas-seq-matcher");
 const { Transform } = require("stream");
 const { TEXT, NODE, NAME, KEY, VALUE, SCRIPT, STYLE, COMMENT } = require("./states");
 
+class SeqMatcher {
+  constructor(str) {
+    this.str = str;
+    this.max = str.length - 1;
+    this.pos = 0;
+  }
+  found(code) {
+    if (code !== this.str.charCodeAt(this.pos)) return !!(this.pos = 0);
+    if (this.pos === this.max) return !(this.pos = 0);
+    return !++this.pos;
+  }
+  reset() {
+    return !(this.pos = 0);
+  }
+}
+
 module.exports = class HtmlParser extends Transform {
   constructor({preserveWS} = {preserveWS: false}){
     super({readableObjectMode: true})
-    const endScript = SeqMatcher("</script>")
-    const endStyle = SeqMatcher("</style>")
-    const beginComment = SeqMatcher("!--")
-    const endComment = SeqMatcher("-->")
-    let cache, name, key, text, data, state, curPos, minPos;
-    let isClose, isSelfClose, hasEqual, valStartChar;
-    this.reset = () => {
-      endStyle.reset(), endScript.reset(), endComment.reset(), beginComment.reset();
-      state = TEXT, data = {}, text = [], name = cache = key = "";
-      curPos = minPos = 0;
-      isClose = isSelfClose = hasEqual = valStartChar = null;
-    }
-    this.reset();
-    this.cache = chunk => (cache+=chunk).length;
-    const getCache = (start, end) => cache.substr(start, end-start);
-    const flushKey = (v, i) => (key = data[key || cache.substr(v, i-v)] = "")
-    const flushVal = (v, i) => (data[key] = cache.substr(v, i-v), key = "", valStartChar = hasEqual = null)
-    const flushNode = () => {
-      if (!isClose) this.push({name, data})
-      if (isSelfClose || isClose) this.push({name})
-      const s = name === "script" ? SCRIPT : name === "style" ? STYLE : name === "!--" ? COMMENT : TEXT
-      data = {}, name = "", isClose = isSelfClose = null
-      return s
-    }
-    const flushSpecialNode = (v, i, name) => {
-      const text = cache.substr(v, i-v)
-      text && this.push({text}), this.push({name})
-      return TEXT
-    }
-    const flushText = (v, i) => {
-      if (v < i) {
-        text.push(cache.substr(v, i-v))
-        this.push({text:text.join(" ")})
-        text = []
-      } else if (text.length) {
-        this.push({text:text.join(" ")})
-        text = []
-      }
-    }
-    this.flushText = () => flushText(minPos, curPos)
-    this.parse = cacheLen => {
-      let i = curPos, v = minPos, s = state, c
-      while (i < cacheLen){
-        c = cache.charCodeAt(i)
-        if (s === TEXT){
-          if (!preserveWS && (c === 32 || c >= 9 && c <= 13)) // ws
-            v < i && text.push(getCache(v, i)), v = i + 1
-          else if (c === 60) // <
-            flushText(v, i), s = NODE, v = i + 1
-        } else if (s === NODE){
-          if (c === 62) // >
-            key && flushKey(), s = flushNode(), v = i + 1
-          else if (c === 47 && !hasEqual) // /
-            isClose = !(isSelfClose = !!name)
-          else if (c !== 32 && (c < 9 || c > 13)){ // !ws
-            if (!name) // name start
-              beginComment.found(c), v = i, s = NAME
-            else if (!key) // key start
-              v = i, s = KEY
-            else if (c === 61) // =
-              hasEqual = true
-            else if (!hasEqual) // next key
-              flushKey(), v = i, s = KEY
-            else if (c === 34 || c === 39) // ', "
-              v = i + 1, valStartChar = c, s = VALUE
-            else // un-quoted val
-              v = i, s = VALUE
-          }
-        } else if (s === NAME){
-          if (beginComment.found(c)) // start comment
-            name = getCache(v, i + 1), s = flushNode(), v = i + 1
-          else if (c === 32 || c >= 9 && c <= 13) // ws
-            name = getCache(v, i), s = NODE, v = i + 1
-          else if (c === 47) // /
-            isSelfClose = true, name = getCache(v, i), s = NODE, v = i + 1
-          else if (c === 62) // >
-            name = getCache(v, i), s = flushNode(), v = i + 1
-        } else if (s === KEY){
-          if (c === 32 || c >= 9 && c <= 13) // ws
-            key = getCache(v, i), s = NODE, v = i + 1
-          else if (c === 61) // =
-            hasEqual = true, key = getCache(v, i), s = NODE, v = i + 1
-          else if (c === 47) // /
-            isSelfClose = true, key = getCache(v, i), s = NODE, v = i + 1
-          else if (c === 62) // >
-            flushKey(v,i), s = flushNode(), v = i + 1
-        } else if (s === VALUE){
-          if (valStartChar != null){
-            if (c === valStartChar) // found end quote
-              flushVal(v,i), s = NODE, v = i + 1
-          } else if (c === 32 || c >= 9 && c <= 13) // ws
-            flushVal(v,i), s = NODE, v = i + 1
-          else if (c === 62) // >
-            flushVal(v,i), s = flushNode(), v = i + 1
-        } else if (s === COMMENT && endComment.found(c))
-          s = flushSpecialNode(v, i-2, "!--"), v = i + 1
-        else if (s === SCRIPT && endScript.found(c))
-          s = flushSpecialNode(v, i-8, "script"), v = i + 1
-        else if (s === STYLE && endStyle.found(c))
-          s = flushSpecialNode(v, i-7, "style"), v = i + 1
-        i = i + 1
-      }
-      cache = cache.substr(v), curPos = i - v, minPos = 0, state = s;
-    }
+    this.preserveWS = preserveWS;
+    this.endScript = new SeqMatcher("</script>")
+    this.endStyle = new SeqMatcher("</style>")
+    this.beginComment = new SeqMatcher("!--")
+    this.endComment = new SeqMatcher("-->")
+
+    this.curPos = 0;
+    this.minPos = 0;
+    this.state = TEXT;
+
+    this.cache = "";
+    this.name = "";
+    this.key = "";
+    this.text = [];
+    this.data = {};
+
+    this.flags = {
+      isClose: null,
+      isSelfClose: null,
+      hasEqual: null,
+      valStartChar: null,
+    };
   }
   _transform(chunk, encoding, done){
     if (chunk === null) return this.end();
-    this.parse(this.cache(chunk));
+    this.parse(this.addToCache(chunk));
     done(null)
   }
   _flush(done){
-    this.flushText();
+    this.flushAllText();
     this.reset();
     done(null)
   }
+  parse(cacheLen) {
+    let i = this.curPos, v = this.minPos, s = this.state, c
+    while (i < cacheLen){
+      c = this.cache.charCodeAt(i)
+      if (s === TEXT){
+        if (!this.preserveWS && (c === 32 || c >= 9 && c <= 13)) // ws
+          v < i && this.text.push(this.getCache(v, i)), v = i + 1
+        else if (c === 60) // <
+          this.flushText(v, i), s = NODE, v = i + 1
+      } else if (s === NODE){
+        if (c === 62) // >
+          this.key && this.flushKey(), s = this.flushNode(), v = i + 1
+        else if (c === 47 && !this.flags.hasEqual) // /
+          this.flags.isClose = !(this.flags.isSelfClose = !!this.name)
+        else if (c !== 32 && (c < 9 || c > 13)){ // !ws
+          if (!this.name) // name start
+            this.beginComment.found(c), v = i, s = NAME
+          else if (!this.key) // key start
+            v = i, s = KEY
+          else if (c === 61) // =
+            this.flags.hasEqual = true
+          else if (!this.flags.hasEqual) // next key
+            this.flushKey(), v = i, s = KEY
+          else if (c === 34 || c === 39) // ', "
+            v = i + 1, this.flags.valStartChar = c, s = VALUE
+          else // un-quoted val
+            v = i, s = VALUE
+        }
+      } else if (s === NAME){
+        if (this.beginComment.found(c)) // start comment
+          this.name = this.getCache(v, i + 1), s = this.flushNode(), v = i + 1
+        else if (c === 32 || c >= 9 && c <= 13) // ws
+          this.name = this.getCache(v, i), s = NODE, v = i + 1
+        else if (c === 47) // /
+          this.flags.isSelfClose = true, this.name = this.getCache(v, i), s = NODE, v = i + 1
+        else if (c === 62) // >
+          this.name = this.getCache(v, i), s = this.flushNode(), v = i + 1
+      } else if (s === KEY){
+        if (c === 32 || c >= 9 && c <= 13) // ws
+          this.key = this.getCache(v, i), s = NODE, v = i + 1
+        else if (c === 61) // =
+          this.flags.hasEqual = true, this.key = this.getCache(v, i), s = NODE, v = i + 1
+        else if (c === 47) // /
+          this.flags.isSelfClose = true, this.key = this.getCache(v, i), s = NODE, v = i + 1
+        else if (c === 62) // >
+          this.flushKey(v,i), s = this.flushNode(), v = i + 1
+      } else if (s === VALUE){
+        if (this.flags.valStartChar != null){
+          if (c === this.flags.valStartChar) // found end quote
+            this.flushVal(v,i), s = NODE, v = i + 1
+        } else if (c === 32 || c >= 9 && c <= 13) // ws
+          this.flushVal(v,i), s = NODE, v = i + 1
+        else if (c === 62) // >
+          this.flushVal(v,i), s = this.flushNode(), v = i + 1
+      } else if (s === COMMENT && this.endComment.found(c))
+        s = this.flushSpecialNode(v, i-2, "!--"), v = i + 1
+      else if (s === SCRIPT && this.endScript.found(c))
+        s = this.flushSpecialNode(v, i-8, "script"), v = i + 1
+      else if (s === STYLE && this.endStyle.found(c))
+        s = this.flushSpecialNode(v, i-7, "style"), v = i + 1
+      i = i + 1
+    }
+    this.cache = this.cache.substr(v);
+    this.curPos = i - v;
+    this.minPos = 0;
+    this.state = s;
+  }
+  reset() {
+    this.endStyle.reset();
+    this.endScript.reset();
+    this.endComment.reset();
+    this.beginComment.reset();
+
+    this.state = TEXT;
+    this.cache = "";
+    this.name = "";
+    this.key = "";
+    this.text = [];
+    this.data = {};
+
+    this.curPos = 0;
+    this.minPos = 0;
+
+    this.flags = {
+      isClose: null,
+      isSelfClose: null,
+      hasEqual: null,
+      valStartChar: null,
+    };
+  }
+  getCache(start, end) {
+    return this.cache.substr(start, end-start);
+  }
+  addToCache(chunk) {
+    return (this.cache += chunk).length;
+  }
+  flushKey(v, i) {
+    return (this.key = this.data[this.key || this.cache.substr(v, i-v)] = "");
+  }
+  flushVal(v, i) {
+    return (this.data[this.key] = this.cache.substr(v, i-v), this.key = "", this.flags.valStartChar = this.flags.hasEqual = null);
+  }
+  flushNode() {
+    const name = this.name;
+    if (!this.flags.isClose) this.push({name, data: this.data})
+    if (this.flags.isSelfClose || this.flags.isClose) this.push({name})
+    const s = name === "script" ? SCRIPT : name === "style" ? STYLE : name === "!--" ? COMMENT : TEXT
+    this.data = {};
+    this.name = "";
+    this.flags.isClose = this.flags.isSelfClose = null
+    return s
+  }
+  flushSpecialNode(v, i, name) {
+    const text = this.cache.substr(v, i-v)
+    text && this.push({text}), this.push({name})
+    return TEXT
+  }
+  flushText(v, i) {
+    if (v < i) {
+      this.text.push(this.cache.substr(v, i-v))
+      this.push({text: this.text.join(" ")})
+      this.text = []
+    } else if (this.text.length) {
+      this.push({text: this.text.join(" ")})
+      this.text = []
+    }
+  }
+  flushAllText() {
+    return this.flushText(this.minPos, this.curPos);
+  }
 }

New alerts

Major refactor

Supply chain risk

Package has recently undergone a major refactor. It may be unstable or indicate significant internal changes. Use caution when updating to versions that include significant changes.

Found 1 instance in 1 package

New author

Supply chain risk

A new npm collaborator published a version of the package for the first time. New collaborators are usually benign additions to a project, but do indicate a change to the security surface area of a package.

Found 1 instance in 1 package

Fixed alerts

Filesystem access

Supply chain risk

Accesses the file system, and could potentially read sensitive data.

Found 1 instance in 1 package

Improved metrics

Dependency count

0

decreased by-100%

Number of lines in readme file

200

increased by1.52%

Number of low supply chain risk alerts

0

decreased by-100%

Worsened metrics

Total package byte prevSize

16812

decreased by-51.03%

Dev dependency count

3

increased by50%

Number of package files

5

decreased by-84.37%

Lines of code

187

decreased by-65.75%

Number of medium supply chain risk alerts

2

increased byInfinity%

Dependency changes

• - Removed@bonniernews/atlas-seq-matcher@^1.0.3

• - Removed@bonniernews/atlas-seq-matcher@1.0.4(transitive)