[TIKA-1140] Better table representation, cell spanning in Word Extractor

• Type: Improvement
• Status: Open
• Priority: Minor
• Resolution: Unresolved
• Component/s: parser
• Assignee: Unassigned
• Reporter: Denis Kildishev
• Created: 02/Jul/13 10:36
• Updated: 15/Mar/15 21:15

Description

As for current version of Word Extractor, it have access to different
features of tables, but most of them are not used. As an example of possible improvements, may be support for borders, fixed cell widths and cell spanning.
It should be noted that some of that features are already used in poi version of Html converted, so, that code can be reused in Tika.
As an example of possible solution may be patch linked as an attachment. It have some code that is based on 2007 version of doc format specification(especially, Border type and color detection), so, different improvements tends to be made to meet with older formats.
Patch already includes some changes in unit tests, that are required in accordance with changes in document structure.

Issue Links

• relates to: TIKA-1144 Changes in styling mechanism, inner table support and list support for Word Extractor
• relates to: TIKA-1178 Improve docx multiple section handling - headers and footers