XML:Lang and attribute type as an unique key? #3
Comments
|
xydiff is written to work for any xml document whatsoever. Theoretically it would be possible to pull an xml schema off of an But that isn't what's happening here. The commit that introduced this change occurred six years ago, and I honestly can't recall my thinking. You are correct that it looks wrong. I'll see if I can't dig deeper and then possibly remove this code. Do you happen to have a test case for a bug that this causes? |
|
I do not know if this constitutes a bug, I am using this tool to detect differences in html files scrapped from the internet. Due to the nature of such files, a xml schema is rarely associated. I tried to create a test case for the bug when comparing the following to files: <!DOCTYPE html>
<html>
<head>
<title>Text Input Control</title>
</head>
<body>
<form >
First name: <input type="text" name="first_name" />
<br>
Last name: <input type="text" name="last_name" />
</form>
</body>
</html>against <!DOCTYPE html>
<html>
<head>
<title>Text Input Control</title>
</head>
<body>
<form >
Last name: <input type="text" name="last_name" />
<br>
First name: <input type="text" name="first_name" />
</form>
</body>
</html>The given output is: <?xml version="1.0" encoding="UTF-8" standalone="no" ?>
<xy:unit_delta xmlns:xy="urn:schemas-xydiff:xydelta">
<xy:t fromXidMap="(1-3;6-7;12-13;8-11|14)">
<xy:d move="yes" par="8" pos="2" xm="(7)"/>
<xy:d move="yes" par="8" pos="1" xm="(6)"/>
<xy:d par="9" pos="2" xm="(5)">
<input name="first_name" type="text"/>
</xy:d>
<xy:d par="9" pos="1" xm="(4)">
First name: </xy:d>
<xy:i move="yes" par="9" pos="1" xm="(6)"/>
<xy:i move="yes" par="9" pos="2" xm="(7)"/>
<xy:i par="8" pos="1" xm="(12)">
First name: </xy:i>
<xy:i par="8" pos="2" xm="(13)">
<input name="first_name" type="text"/>
</xy:i>
</xy:t>
</xy:unit_delta>Isn't this output of inserts and deletes a consequence of the algorithm trying to exactly match 2 seemingly different nodes because of the assumed uniqueness of attribute "type"? |
|
This can be changed by modifying the lines: You may close this issue |
Hello, i've been going over the code of this tool which seems extremely useful and I came across a part of the code which does not seem to make much sense.
https://github.com/fdintino/xydiff/blob/master/src/Diff_NodesManager.cpp#L243
Why when registering a sub tree, if a node has the attribute type then it is classified as a unique key to that node. If using this tool to diff html files for instance, we can define multiple nodes with an attribute type with the same value: "text".
A simple html form would create this situation.
Shouldn't the if merely test if an id attribute exists, that is the only guaranteed unique attribute in the whole document, assuming a valid and html compliant file.
Kind regards :)
The text was updated successfully, but these errors were encountered: