fixes #78, forcing all types to lower case

dbashford · Mar 28, 2016 · e719671 · e719671
1 parent 549061b
commit e719671
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 2 deletions.
diff --git a/lib/extract.js b/lib/extract.js
@@ -10,7 +10,7 @@ var fs = require( 'fs' )
   , totalExtractors = 0
   , satisfiedExtractors = 0
   , hasInitialized = false
-  
+
   , WHITELIST_PRESERVE_LINEBREAKS = /[^A-Za-z\x80-\xFF 0-9 \u2018\u2019\u201C|\u201D\u2026 \u00C0-\u1FFF \u2C00-\uD7FF \.,\?""!@#\$%\^&\*\(\)-_=\+;:<>\/\\\|\}\{\[\]`~'-\w\n\r]*/g
   , WHITELIST_STRIP_LINEBREAKS = /[^A-Za-z\x80-\xFF 0-9 \u2018\u2019\u201C|\u201D\u2026 \u00C0-\u1FFF \u2C00-\uD7FF \.,\?""!@#\$%\^&\*\(\)-_=\+;:<>\/\\\|\}\{\[\]`~'-\w]*/g
   , SINGLE_QUOTES = /[\u2018|\u2019]/g
@@ -22,6 +22,7 @@ var registerExtractor = function( extractor ) {
   if ( extractor.types ) {
     extractor.types.forEach( function( type ) {
       if ( typeof type === "string" ) {
+        type = type.toLowerCase();
         typeExtractors[type] = extractor.extract;
       } else {
         if ( type instanceof RegExp ) {
@@ -35,7 +36,7 @@ var registerExtractor = function( extractor ) {
 var registerFailedExtractor = function( extractor, failedMessage ) {
   if ( extractor.types ) {
     extractor.types.forEach( function( type ) {
-      failedExtractorTypes[type] = failedMessage;
+      failedExtractorTypes[type.toLowerCase()] = failedMessage;
     });
   }
 };
@@ -100,6 +101,7 @@ var initializeExtractors = function(options) {
 };
 
 var findExtractor = function( type, filePath ) {
+  type = type.toLowerCase();
   if ( typeExtractors[type] ) {
     return typeExtractors[type];
   } else {

diff --git a/test/general_test.js b/test/general_test.js
@@ -43,6 +43,17 @@ describe('textract', function() {
 
   });
 
+  it('can handle types of varying cases', function(done) {
+    var filePath = path.join(__dirname, 'files', 'new docx(1).docx');
+    fromFileWithMimeAndPath('appLication/vnd.openXMLformats-Officedocument.WordProcessingml.Document', filePath, function(error, text) {
+      expect(error).to.be.null;
+      expect(text).to.be.a('string');
+      expect(text.substring(0, 38)).to.eql( "This is a test Just so you know: Lorem" );
+      done();
+    });
+  });
+
+
   it('can handle a text file with parens', function(done) {
     var filePath = path.join(__dirname, 'files', 'new doc(1).txt');
     fromFileWithPath(filePath, function( error, text ) {