Convert docx to XHTML introduces invalid characters

by **david.zhaowl** » Mon Jun 29, 2015 2:09 am

Hi Jason,

I'm trying to convert part of a docx document into XHTML format. It has been working fine for me but sometimes when I start the application on different linux machine with under same JDK version, the conversion introduces some invalid characters for the bullets. I attached a photo to display the error. And here's my code:

Code: Select all: /* * Convert the description in table cell back into html code to be saved into database */ private String convertTcToXhtml(Tc tc, WordprocessingMLPackage emptyImportMLPackage) throws Docx4JException { List<Object> paragraphs = getAllElementFromObject(tc, P.class); if (paragraphs == null || paragraphs.size() == 0) { return null; } StringBuilder sb = new StringBuilder(); /* clear all content of the clone wordMLPackage */ emptyImportMLPackage.getMainDocumentPart().getContent() .removeAll(emptyImportMLPackage.getMainDocumentPart().getContent()); removeHeaderFooterFromWord(emptyImportMLPackage); /* Add content of the description to new wordMLPackage. */ emptyImportMLPackage.getMainDocumentPart().getContent().addAll(paragraphs); HTMLSettings htmlSettings = Docx4J.createHTMLSettings(); htmlSettings.setWmlPackage(emptyImportMLPackage); OutputStream os = new ByteArrayOutputStream(); Docx4jProperties.setProperty("docx4j.Convert.Out.HTML.OutputMethodXML", true); Docx4J.toHTML(htmlSettings, os, Docx4J.FLAG_EXPORT_PREFER_XSL); String xhtmlString = ((ByteArrayOutputStream) os).toString(); sb.append(xhtmlString); /* delete useless header */ sb.delete(0, sb.indexOf("<style>")); sb.insert(0, "<html><head>"); LOGGER.debug("clean converTcToXhml : " + sb.toString()); /* replace java bullet dot with html bullet dot. */ String descString = sb.toString().replaceAll("\u2022", "•"); return descString; }

I have a line

Code: Select all: String descString = sb.toString().replaceAll("\u2022", "•");

to replace the bullet dot with html bullet dot code, otherwise the bullet will become a double quote within the editor. The syntax of the editor is not open to me and belongs to the vendor. Not sure if you have met this problem before.

I've been running under JDK1.8.0_20, version of docx4j-ImportXHTML and docx4j is 3.2.0.

Thanks in advance for any help.

I found that actually the invalid characters are already there after line

Code: Select all: String xhtmlString = ((ByteArrayOutputStream) os).toString();

. Here is the log I see:

2015-06-28_12:36:13.816 INFO c.c.p.a.w.w.WordProcessor - clean converTcToXhml : <html><head><style></style><script type="text/javascript"></script></head><body>



<div class="document">

Test Case Description

??? Test Objective

??? To verify that the insertion and removals of the STM-0J to and from the 6500 slots are as expected.

??? Also to verify that the STM-0J Circuit Pack is properly keyed so that it cannot be inserted into the 6500 slots upside down or into slots where allocated for SPs in 14 and 32-slot shelves.

 

??? Test Configuration

??? STM-0J Circuit Pack is available for inspection and removals.

??

??? Test Procedure 

??? Turn the power the targeted 6500 shelf off so that no power related issues occur during this test.

??? Attempt to insert the STM-0J Circuit Pack into any empty supported slot with neighboring circuit packs already inserted. Verify that:

??? The circuit pack slide smoothly into the slot without rubbing against the neighboring circuit packs or the shelf wall (if it is slot 1).

??? That excessive force is not required to seat the card and close the latches.

??? Attempt to remove the STM-0J Circuit Pack from the slot. Verify that:

??? Verify that the latches can be opened without excessive force and that the circuit pack can be removed without rubbing against other circuit packs or the shelf wall (if it is slot 1).

??? Verify that no physical damage has occurred to the backplane of the shelf or to the backplane connector of the circuit pack.

??? Attempt to remove the STM-0J Circuit Pack from the slot. Verify that:

??? Verify that the latches can be opened without excessive force and that the circuit pack can be removed without rubbing against other circuit packs or the shelf wall (if it is slot 1).

??? Verify that no physical damage has occurred to the backplane of the shelf or to the backplane connector of the circuit pack.

??? Repeat the above steps for all of the supported slots on the 6500 shelf (If it is not done in every slot, please record the tested slots and record it in the Quality Center results).

??? Choose one of the supported slots on the 6500 shelf and attempt to insert the STM-0J Circuit Pack upside down. 

??? Verify that the keying on the STM-0J Circuit Pack and 6500 shelf slot does not allow the circuit pack to be seated upside down.

??? Verify that no physical damage has occurred to the backplane of the shelf or to the backplane connector of the circuit pack due to the attempted insertion.

??? Inspect the STM-0J Circuit Pack and verify that it???s slot-keyed to prevent being seated into slots 15 and 16 (41 & 42 for 32 slot shelf). Attempt to insert the STM-0J Circuit Pack into an empty slot which is designated for the SP card (15 or 16). Try to insert the card both right-side up and upside down.

??? Verify that the keying on the STM-0J Circuit Pack and 6500 shelf slot does not allow the circuit pack to be seated in the SP slot. 

??? Verify that no physical damage has occurred to the backplane of the shelf or to the backplane connector of the circuit pack due to the attempted insertion

??

??? Expected Results

Verify STM-0J CP insertion and removal is as expected.

??</div>

<div class="footnotes">

??</div>



</body></html>

In the log you can see the "??" are the invalid characters. Any clue how would that happen?

I uploaded the docx file that I was trying to import, within which it's the "Description" content in the form. Thanks.

Convert docx to XHTML introduces invalid characters

Convert docx to XHTML introduces invalid characters

Who is online