3 ms·
Your question is asking for a "diff" algorithm. For Unix/Linux systems there is the classic diff utility itself [1]. For arbitrary binary content you'd be loo
by pwg 10y ago
Your question is asking for a "diff" algorithm. For Unix/Linux systems there is the classic diff utility itself [1]. For arbitrary binary content you'd be looking at something like bdiff [2] or xdelta [3].
[1] https://en.wikipedia.org/wiki/Diff_utility https://en.wikipedia.org/wiki/Diff_utility
[2] http://bdiff.sourceforge.net/ http://bdiff.sourceforge.net/
[3] http://xdelta.org/ http://xdelta.org/
- passioncurious 10y agoAs noted in another answer-- the data is very free form and a simple diff may not help. Imagine having a million resumes (some in doc, some in pdf, some in html) and you want to dedupe them.
- pwg 10y agoWhich important detail also was not in the original question. What was in there was: I essentially want to store one variation of the content in my database, with references to altered variations of the content, but I need to be able to detect the altered data. For which a "diff" style algorithm is a possibility, just storing the "alterations". Your update edit changed your question to be one for a different problem. For this goal-post move, you'll want to look into plagiarism detection algorithms.
- passioncurious 10y agoThat's why I updated the question to prevent any more confusion. Thanks-- plagiarism detection sounds like a good avenue to check out.