feat!: Require encoding option to avoid inspecting chunks for UTF-8 encoding

README.md

@@ -18,7 +18,7 @@ var concat = require('concat-stream');
 var removeBOM = require('remove-bom-stream');
 
 fs.createReadStream('utf8-file-with-bom.txt')
-  .pipe(removeBOM())
+  .pipe(removeBOM('utf-8'))
   .pipe(
     concat(function (result) {
       // result won't have a BOM
@@ -28,9 +28,9 @@ fs.createReadStream('utf8-file-with-bom.txt')
 
 ## API
 
-### `removeBOM()`
+### `removeBOM(encoding)`
 
-Returns a `through2` stream that will remove a BOM, given the data is a UTF8 Buffer with a BOM at the beginning. If the data is not UTF8 or does not have a BOM, the data is not changed and this becomes a normal passthrough stream.
+Returns a `through2` stream that will remove a BOM, if the argument `encoding` is `'utf-8'` and the given data is a UTF8 Buffer with a BOM at the beginning. If the `encoding` is not `'utf-8'` or does not have a BOM, the data is not changed and this becomes a normal passthrough stream.
 
 ## License
 

index.js

@@ -1,51 +1,54 @@
 'use strict';
 
 var through = require('through2');
-var removeBom = require('remove-bom-buffer');
+var TextDecoder = require('util').TextDecoder;
 
-function removeBomStream() {
-  var state = 0; // 0:Not removed, -1:In removing, 1:Already removed
-  var buffer = Buffer.alloc(0);
+var BOM = Buffer.from([0xEF, 0xBB, 0xBF], 'utf-8');
 
-  return through(onChunk, onFlush);
+function removeBomStream(encoding) {
+  encoding = (encoding || '').toLowerCase();
+  var isUTF8 = (encoding === 'utf-8' || encoding === 'utf8');
 
-  function removeAndCleanup(data) {
-    state = 1; // Already removed
+  if (!isUTF8) {
+    return through();
+  }
 
-    buffer = null;
+  // Only used if encoding is UTF-8
+  var decoder = new TextDecoder('utf-8', { ignoreBOM: false });
 
-    return removeBom(data);
-  }
+  var state = 0; // 0:Not removed, -1:In removing, 1:Already removed
+
+  return through(onChunk);
 
-  function onChunk(data, enc, cb) {
+  function onChunk(data, _, cb) {
     if (state === 1) {
-      return cb(null, data);
+      cb(null, data);
+      return;
     }
 
-    if (state === 0 /* Not removed */ && data.length >= 7) {
-      return cb(null, removeAndCleanup(data));
-    }
+    try {
+      state = -1;
 
-    state = -1; // In removing
+      var chunk = decoder.decode(data, { stream: true });
 
-    var bufferLength = buffer.length;
-    var chunkLength = data.length;
-    var totalLength = bufferLength + chunkLength;
+      // The first time we have data after a decode, it should have already removed the BOM
+      if (chunk !== '') {
+        chunk += decoder.decode(); // end of stream mode and clear inner buffer.
 
-    buffer = Buffer.concat([buffer, data], totalLength);
+        var buffer = Buffer.from(chunk, 'utf-8');
 
-    if (totalLength >= 7) {
-      return cb(null, removeAndCleanup(buffer));
-    }
-    cb();
-  }
+        // Node<=v11, TextDecoder#decode returns a BOM if it receives a BOM separately.
+        if (BOM.compare(buffer) !== 0) {
+          state = 1;
+          cb(null, buffer);
+          return;
+        }
+      }
 
-  function onFlush(cb) {
-    if (state === 2 /* Already removed */ || !buffer) {
-      return cb();
+      cb();
+    } catch (err) {
+      cb(err);
     }
-
-    cb(null, removeAndCleanup(buffer));
   }
 }
 

package.json

@@ -22,7 +22,6 @@
     "test": "nyc mocha --async-only"
   },
   "dependencies": {
-    "remove-bom-buffer": "^3.0.0",
     "through2": "^4.0.2"
   },
   "devDependencies": {

test/index.js

@@ -24,7 +24,7 @@ describe('removeBomStream', function () {
     }
 
     pipe(
-      [fs.createReadStream(filepath), removeBomStream(), concat(assert)],
+      [fs.createReadStream(filepath), removeBomStream('utf-8'), concat(assert)],
       done
     );
   });
@@ -33,7 +33,7 @@ describe('removeBomStream', function () {
     var filepath = path.join(__dirname, './fixtures/test.txt');
     var fileContent = fs.readFileSync(filepath, 'utf-8');
 
-    var rmBom = removeBomStream();
+    var rmBom = removeBomStream('utf8');
     var output = '';
     rmBom.on('data', function (d) {
       output += d.toString();
@@ -55,7 +55,7 @@ describe('removeBomStream', function () {
     }
 
     pipe(
-      [fs.createReadStream(filepath), removeBomStream(), concat(assert)],
+      [fs.createReadStream(filepath), removeBomStream('UTF-8'), concat(assert)],
       done
     );
   });
@@ -73,7 +73,7 @@ describe('removeBomStream', function () {
       [
         fs.createReadStream(filepath),
         chunker(1),
-        removeBomStream(),
+        removeBomStream('UTF8'),
         concat(assert),
       ],
       done
@@ -92,7 +92,7 @@ describe('removeBomStream', function () {
     }
 
     pipe(
-      [fs.createReadStream(filepath), removeBomStream(), concat(assert)],
+      [fs.createReadStream(filepath), removeBomStream('UTF-8'), concat(assert)],
       done
     );
   });
@@ -101,7 +101,7 @@ describe('removeBomStream', function () {
     var filepath = path.join(__dirname, './fixtures/bom-utf8.txt');
     var fileContent = fs.readFileSync(filepath, 'utf-8');
 
-    var rmBom = removeBomStream();
+    var rmBom = removeBomStream('utf-8');
     var output = '';
     rmBom.on('data', function (d) {
       output += d.toString();
@@ -123,7 +123,7 @@ describe('removeBomStream', function () {
     }
 
     pipe(
-      [fs.createReadStream(filepath), removeBomStream(), concat(assert)],
+      [fs.createReadStream(filepath), removeBomStream('utf-16be'), concat(assert)],
       done
     );
   });
@@ -138,7 +138,7 @@ describe('removeBomStream', function () {
     }
 
     pipe(
-      [fs.createReadStream(filepath), removeBomStream(), concat(assert)],
+      [fs.createReadStream(filepath), removeBomStream('utf-16be'), concat(assert)],
       done
     );
   });
@@ -153,7 +153,7 @@ describe('removeBomStream', function () {
     }
 
     pipe(
-      [fs.createReadStream(filepath), removeBomStream(), concat(assert)],
+      [fs.createReadStream(filepath), removeBomStream('utf-16le'), concat(assert)],
       done
     );
   });