Error in Equality on complex values

[GoogleCodeExporter]

What steps will reproduce the problem?
What is the expected output? What do you see instead?

=> Detailed in the attachment.

What version of the product are you using? On what operating system?

=> 0.8.6

Please provide any additional information below :

According to a discussion with Mike, it seems that when comparing values which 
it doesn’t recognize (such as nested values), the Asterix system will compare 
the binary representation, which happen to be equal.

Original issue reported on code.google.com by jules.te...@gmail.com on 19 Dec 2014 at 8:46

Attachments:

• equality_paradox.md

[GoogleCodeExporter]

I don't know what an md file is - can you please cut/paste a minimized 
reproducer (self-contained) as text into an issue comment?  (Or attach a text 
file at least?)  Thanks!  We tried to reproduce this from memory late this 
afternoon and kept getting the "right" response (a feature not supported error).

Original comment by dtab...@gmail.com on 20 Dec 2014 at 8:22

[GoogleCodeExporter]

1) I don't know what an md file is
An md file is nothing more than a markdown file, such that the contents are 
pretty-formatted when viewed with a markdown viewer, and are readable as text 
when the file is opened in the regular viewer. 

2) can you please cut/paste a minimized reproducer (self-contained) as text
A input.txt and output.txt are provided.

Original comment by jules.te...@gmail.com on 22 Dec 2014 at 4:46

Attachments:

• input.txt
• output.txt

[GoogleCodeExporter]

Interesting...  Thx!

Original comment by dtab...@gmail.com on 24 Dec 2014 at 7:12

[GoogleCodeExporter]

Are you sure the provided output is correct?
My concern here is that it contains records with "co": [ 0.5d ]  while none of 
the records inserted into the dataset have co=0.5d.

Did you mean "co": [ 0.5d , 0.7 ] ?? 

Original comment by bamou...@gmail.com on 24 Dec 2014 at 9:29

[GoogleCodeExporter]

sorry for my late response, I had lost track of this issue. 

Indeed the output you have seen is incorrect. Here is the actual output :

Original comment by jules.te...@gmail.com on 21 Jan 2015 at 7:05

Attachments:

• output.txt