For possible solutions and dicussions, go to #56
Reflowable means text can be automatically adapted to different output device, especially different widths. See Wikipedia Page
PDF is mainly designed for printing, where it is not necessary for text to be reflowable. Some PDF files with tags are reflowable, which is not common, and is not widely supported by PDF viewers. However HTML originated from reflowable plain text, although nowadays HTML pages become more and more complicated, large paragraphs of text are still reflowable -- unless for the purpose of obscure or encryption.
This page discusses about the difficulties to let pdf2htmlEX produce reflowable text in HTML from a PDF.
Each line is a box. If two lines l1 and l2 share the same left and right edges, l2 is exactly underneath l1 (touching vertically) and there is only one font appearing in both of them, we can append the text in l2 to l1, and increase the height of l1 by the height of l2, after that we remove l2.
For text in English, if l1 ends with a '-', we may need to remove it, otherwise we may need to insert a ' ' there.
In HTML files produced by pdf2htmlEX, The position of a line is determined by the position of the top-left corner and the height of the line. Especially a correct height is necessary, because of the difference of the coordinate systems between PDF (origin point of characters) and HTML (bounding boxes).
The height of a line is determined by the tallest character in the lines regarding font families and sizes:
So the height will be quite likely to change if the text is reflowed.
Underlined text in PDF may be actually normal text with a correctly positioned horizontal bar. Reflowing text is likely to break this combination. The cause of this is that some information cannot be recognized by pdf2htmlEX.
Other similar things includes code blocks, math formulas (with symbols like square root or fractions with a horizontal bar), poems...
However these may not be easily recognized.