Overview:

    The following pdf causes pdftohtml to output malformed xml:
    http://www.atmel.com/images/Atmel-8284-8-bit-AVR-microcontroller-ATmega169A_PA_329A_PA_3290A_PA_649A_P_6490A_P_datasheet.pdf 
    The resulting xml file has multiple similar errors, the first one on line 71641:
    <text top="180" left="71" width="101" height="15" font="11"><b>Sp<a href="Atmel-8284-8-bit-AVR-microcontroller-ATmega169A_PA_329A_PA_3290A_PA_649A_P_6490A_P_datasheet.html#876">eed [MHz] </b>(3)</a></text>
    (the closing b and a tags are not in the correct order)

Steps to Reproduce: 

    1) wget http://www.atmel.com/images/Atmel-8284-8-bit-AVR-microcontroller-ATmega169A_PA_329A_PA_3290A_PA_649A_P_6490A_P_datasheet.pdf 

    2) pdftohtml -q -i -xml Atmel-8284-8-bit-AVR-microcontroller-ATmega169A_PA_329A_PA_3290A_PA_649A_P_6490A_P_datasheet.pdf output.xml

Actual Results: 

    malformed xml

Expected Results: 

    well-formed xml. And I'm not quite sure if the link is placed on the correct piece of text. In the pdf only the text "(3)" is clickable and none of it is bold.

Build Date & Hardware: 

    Built on 2016-10-18 from source (0.48.0) on Ubunty 14.04 LTS

Additional Builds and Platforms: 

    Also occurred in the version of pdftohtml that was installed using apt-get (0.28 if I recall correctly)


Cheers,


Daniel