[TIKA-307] Better handling of partial/truncated input data to parsers

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.4
• Component/s: parser
• Assignee: Unassigned
• Reporter: Kenneth William Krugler
• Created: 10/Oct/09 14:14
• Updated: 02/Jul/21 18:54
• Resolved: 01/Mar/15 06:46

Description

Some parsers (e.g. ZipParser) can hang if they prematurely reach the end of the input stream.

As a way of avoiding this issue, Jukka had suggested the following approach on the list:

The input stream could be wrapped into a decorator that throws a tagged IOException when the given size limit has been reached. This assumes that all parsers will correctly propagate up an IOException (at the very least).

Smarter parsers could cleanly close the emitted XHTML stream, potentially adding a metadata entry that signifies that the extracted text has been truncated.

Issue Links

• is duplicated by: TIKA-455 Zip parser stuck on truncated zip files.
• links to: GitHub Pull Request #335