3963: markdown in the field Note to Reviewers gets quoted with backslashes after editing

paul.ma*******@alcatel-l********* (Google Code) (Is this you? Claim this profile.)
chipx86
chipx86
4110
What version are you running?

2.0.15


What's the URL of the page containing the problem?

http://reviewboard.example.com/r/2/

What steps will reproduce the problem?
1. edit the review, putting text into the "Note to reviewers" box, using markdown e.g.
#this is a highlighted comment

2. edit the review, changing the text in the note to reviewers box, perhaps adding something new

3.submit as normal

What is the expected output? 

some highlighted text "this is a highlighted comment"

What do you see instead?

plain text reading "#this is a highlighted comment"

What operating system are you using? 

Hosted on linux.


What browser?

Chrome on Linux (ubuntu desktop 15.04 x86-64)


Please provide any additional information below.
chipx86
#1 chipx86
Interesting. We got a similar comment from someone developing a custom extension just a week or two ago, but we could not reproduce it with the same extension or with Note to Reviewers. I'm unsure what the difference is...

Can you get the full payload of that review request from the API and send it here? (/api/review-requests/<id>/)
  • +NeedInfo
  • +Project-ReviewBoard
    +Component-Markdown
chipx86
#2 chipx86
Also, did you publish between steps 1 and 2? If so, I'd be interested in seeing API data from both stages.
#3 paul.ma*******@alcatel-l********* (Google Code) (Is this you? Claim this profile.)
Hi,
yes, I did publish the review with some notes-for-reviewers set, and then modified the field afterwards. 
I can subsequently edit the field and remove the backslash and click OK and then the text gets re-emboldened. But when I publish the review the markdown gets re-enquoted into backslash hash, so the text becomes normal.
chipx86
#4 chipx86
Would you be able to get the API payloads for me at each stage? That'll tell me more.
#6 paul.ma*******@alcatel-l********* (Google Code) (Is this you? Claim this profile.)
I re-read what you wrote and changed browser URL as required and downloaded the payload. I can't send that to you, it contains a load of private information.
#7 oesteban

I was thinking of reporting also this problem. Seeing it reported, I would add a new instance of it. The problem is, from the email of paul.ma, I guess we are using the same ReviewBoard server.
I've created a new request, just for testing. I write in the Notes for rewiewers field, press Ok and see the text as expected. Well, sometimes the lists (with +) get shown with a real bullet, most of the times I see the +. But this is a different story.
If I click on the field space to edit again, the markdown is already escaped (C&P from the field, in edit mode):

This a \*\*bold\*\* word, followed by a list:
\+ First
\+ Second

The initial rbt post doesn't contain any text in that field, so no problem (AFAIK). And the rest of the interaction takes place directly from the browser. I've tried to capture the communication but, as it's gzipped, I've given up :(

#8 oesteban

I'm using Firefox on Win7, by the way.

#9 oesteban

I've updated a rr, including a change to the Notes for rewiewers field. In just one screen I see both the Notes for rewiewers with the markdown escaped and, at the end, the change summary, including parts of that field, showing correctly some bold text and a link. So maybe it's stored correctly but, somehow, the problem arises when displaying (or getting the initial value for the field when opened for editing).

gmyers
#10 gmyers

Wanted to add a cross-reference here to https://hellosplat.com/s/beanbag/tickets/4110/. I believe that is essentially the same unerlying issue discussed here. I investigated that a while back and there are some notes in the ticket. I have not had time to work the issue, and probably won't for the forseeable future.

chipx86
#12 chipx86

Fixed on release-2.5.x (9d0eaa8e4ad14bde6ef3c21382f9d86698d909ca and 90119a8420b89edcf463cb6ac47651106770195f).

  • -NeedInfo
    +Fixed
  • +Release-2.0.x
    +Release-2.5.x
  • +chipx86
chipx86
#13 chipx86
  • -Release-2.0.x