-
-
Notifications
You must be signed in to change notification settings - Fork 515
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Few failing internet doctests in mma_free_integrator #25501
Comments
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
comment:4
It seems that Ma#######ca is now hiding carefully its results, so we can no longer parse them.. |
comment:5
From the starting point provided in the ticket description:
one can check what this function is doing:
and follow its content step by step to trace the error:
Since the next line is the one that gives the error:
we can explore
Now if we use Wolfram|Alpha to collect the page describing the integral:
we see that the display of the solution happens in JavaScript. I see no way to parse the html page for the computed integral. |
comment:6
Yes, the web site tries very hard to hide the result, so that no machine can extract it. |
comment:7
The API access for Wolfram|Alpha requires signing in with a Wolfram ID. It can be used for free, limited to 2000 queries per month. |
New commits:
|
Branch: u/slabbe/25501 |
Commit: |
Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:
|
comment:10
Oups, my first push was containing commits from another ticket. I remade the branch on top of the most recent beta. Sorry for the noise. |
Author: Sébastien Labbé |
comment:12
Attachment: wolfram.py.gz I was able to make the Wolfram API work without login. I basically traced what the Wolfram page does (it's full of javascript) using Firefox dev tools, and produced the smallest Python program I could to reproduce it. At the moment it only prints the result, which is big-ish JSON with several results. Example use:
Some random notes: it seems cookies are important, the "Referer" header was important, there is a proxy_code which is important but it's not clear if it can be reused. Obviously Wolfram may change this interface at any time. I have not checked if the params of the requests are the same as the official API. EDIT: I uploaded a second program that uses the synchronous interface and drops some of the (seemingly) useless parameters of the requests. Some parameters are documented in the Full Reference of the Wolfram API. |
Attachment: wolfram2.py.gz |
comment:13
I just updated the branch with a fix based on the suggestion made by gh-pamaury. I stil need to add some doctests... New commits:
|
Changed branch from u/slabbe/25501 to u/slabbe/25501-fix |
Branch pushed to git repo; I updated commit sha1. New commits:
|
Changed keywords from none to thursdaysbdx |
Changed branch from u/slabbe/25501-fix to u/slabbe/25501 |
comment:23
After fixing Python 3, I needed to adapt the doctests for Python 2 again... Rebased on 8.5.rc0. Needs review. |
comment:24
oh, well. Let it be.. |
Reviewer: Frédéric Chapoton |
comment:25
There is a internet-using doctest thats not marked as such:
|
Branch pushed to git repo; I updated commit sha1. New commits:
|
comment:29
ok |
Changed branch from u/slabbe/25501 to |
Changed commit from |
As reported on sage-release 8.3.beta3,
gives
It can be reproduced with:
CC: @slel
Component: doctest coverage
Keywords: thursdaysbdx
Author: Sébastien Labbé, Amaury Pouly
Branch:
f083393
Reviewer: Frédéric Chapoton
Issue created by migration from https://trac.sagemath.org/ticket/25501
The text was updated successfully, but these errors were encountered: