@jsreport/pdfjs - npm Package Compare versions

Comparing version 1.6.0 to 1.6.1

lib/mixins/processText.js

@@ -6,2 +6,3 @@ const CMapFactory = require('../pdfjs-ext/core/cmap.js').CMapFactory
 const zlib = require('zlib')
+const PDF = require('../object')
 
@@ -20,3 +21,10 @@ module.exports = (doc) => {
   for (const pageObject of pages) {
-    const streamObject = pageObject.properties.get('Contents').object.content
+    let streamObject
+    if (Array.isArray(pageObject.properties.get('Contents'))) {
+      streamObject = concatStreams(pageObject.properties.get('Contents'))
+    } else {
+      const contentsObject = pageObject.properties.get('Contents').object
+      streamObject = contentsObject.content
+    }
+
     await processStream({
@@ -295,1 +303,20 @@ doc,
 }
+
+function concatStreams (streams) {
+  const buffers = []
+  for (const content of streams) {
+    const pageStreamObject = content.object?.content?.object
+    if (pageStreamObject) {
+      const filterProp = pageStreamObject.properties.get('Filter')
+      if (filterProp && filterProp.name === 'FlateDecode') {
+        buffers.push(zlib.unzipSync(pageStreamObject.content.content))
+      }
+    }
+  }
+  const contentsObject = new PDF.Object('Contents')
+  const streamObject = new PDF.Stream(contentsObject)
+  streamObject.content = zlib.deflateSync(Buffer.concat(buffers))
+  contentsObject.prop('Length', streamObject.content.length)
+  contentsObject.prop('Filter', 'FlateDecode')
+  return streamObject
+}

package.json

 {
   "name": "@jsreport/pdfjs",
-  "version": "1.6.0",
+  "version": "1.6.1",
   "description": "",
@@ -5,0 +5,0 @@ "main": "index.js",

test/test.js

@@ -509,2 +509,21 @@ const { Document, External } = require('../')
 
+  it('processText should support page Contents as array', async () => {
+    const document = new Document()
+    const external = new External(fs.readFileSync(path.join(__dirname, 'page-contents-array.pdf')))
+    document.append(external)
+
+    let docText = ''
+    await document.processText({
+      resolver: (text) => {
+        docText += text
+      }
+    })
+    await document.asBuffer()
+
+    // the external PDF contains unsuported text commands like
+    // [(pement \351lectrique ou \351lectr)6.1 (onique usag\351 dans la limite de la quantit\351)-15 ( )]TJ
+    // in future we may support it, but at this moment it isn't important because text parsing is primarily used for embedding hidden marks with chrome
+    docText.should.containEql('amiable')
+  })
+
   it('acroform with text type', async () => {
@@ -511,0 +530,0 @@ const document = new Document()

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

2828296

increased by5.13%

Number of package files

97

increased by1.04%

Lines of code

18171

increased by0.23%

No dependency changes