DLTFpT
DLTFpT copied to clipboard
MarkupSafe problem
I am getting the following message when starting the container. The prescribed fix is to downgrade the markupsafe version to 2.0.1, which will not work due to werkzeug 2.2.2 has requirement MarkupSafe>=2.1.1, but you'll have markupsafe 2.0.1 which is incompatible
File "/opt/conda/bin/jupyter-notebook", line 7, in <module>
from notebook.notebookapp import main
File "/opt/conda/lib/python3.7/site-packages/notebook/notebookapp.py", line 41, in <module>
from jinja2 import Environment, FileSystemLoader
File "/opt/conda/lib/python3.7/site-packages/jinja2/__init__.py", line 33, in <module>
from jinja2.environment import Environment, Template
File "/opt/conda/lib/python3.7/site-packages/jinja2/environment.py", line 15, in <module>
from jinja2 import nodes
File "/opt/conda/lib/python3.7/site-packages/jinja2/nodes.py", line 19, in <module>
from jinja2.utils import Markup
File "/opt/conda/lib/python3.7/site-packages/jinja2/utils.py", line 647, in <module>
from markupsafe import Markup, escape, soft_unicode
ImportError: cannot import name 'soft_unicode' from 'markupsafe' (/opt/conda/lib/python3.7/site-packages/markupsafe/__init__.py)
Host system is Ubuntu 20.04.5 LTS.
Hi @caboodle!
Confirming that you are using the Dockerfile here including the jupyter/scipy-notebook:2ce7c06a61a1 base image?
Confirmed. (FROM jupyter/scipy-notebook:2ce7c06a61a1 )
I have only cloned the project about 30 min before posting this issue.
Huh geez. That is a headscratcher, @caboodle. I'm not able to replicate the issue; the image builds perfectly for me (on both Mac OS and Ubuntu 16.04).
Probably not what you're looking to achieve here, but an alternative would be to work through the notebooks in the cloud (in Google Colab) instead of locally.
I can build the image fine, but it will not start, (using rundocker.sh)
Executing the command: jupyter notebook
Traceback (most recent call last):
File "/opt/conda/bin/jupyter-notebook", line 7, in <module>
from notebook.notebookapp import main
File "/opt/conda/lib/python3.7/site-packages/notebook/notebookapp.py", line 41, in <module>
from jinja2 import Environment, FileSystemLoader
File "/opt/conda/lib/python3.7/site-packages/jinja2/__init__.py", line 33, in <module>
from jinja2.environment import Environment, Template
File "/opt/conda/lib/python3.7/site-packages/jinja2/environment.py", line 15, in <module>
from jinja2 import nodes
File "/opt/conda/lib/python3.7/site-packages/jinja2/nodes.py", line 19, in <module>
from jinja2.utils import Markup
File "/opt/conda/lib/python3.7/site-packages/jinja2/utils.py", line 647, in <module>
from markupsafe import Markup, escape, soft_unicode
ImportError: cannot import name 'soft_unicode' from 'markupsafe' (/opt/conda/lib/python3.7/site-packages/markupsafe/__init__.py)
Apparently markupsafe is not longer supplying soft_unicode as per this issue: https://github.com/pallets/markupsafe/issues/304
Google Colab is not an option when it comes to company data, I'm afraid. For the LiveLessions it's fine though.
Conda + DataSpell should also do the trick.
I can build the image fine, but it will not start, (using rundocker.sh)
Executing the command: jupyter notebook Traceback (most recent call last): File "/opt/conda/bin/jupyter-notebook", line 7, in <module> from notebook.notebookapp import main File "/opt/conda/lib/python3.7/site-packages/notebook/notebookapp.py", line 41, in <module> from jinja2 import Environment, FileSystemLoader File "/opt/conda/lib/python3.7/site-packages/jinja2/__init__.py", line 33, in <module> from jinja2.environment import Environment, Template File "/opt/conda/lib/python3.7/site-packages/jinja2/environment.py", line 15, in <module> from jinja2 import nodes File "/opt/conda/lib/python3.7/site-packages/jinja2/nodes.py", line 19, in <module> from jinja2.utils import Markup File "/opt/conda/lib/python3.7/site-packages/jinja2/utils.py", line 647, in <module> from markupsafe import Markup, escape, soft_unicode ImportError: cannot import name 'soft_unicode' from 'markupsafe' (/opt/conda/lib/python3.7/site-packages/markupsafe/__init__.py)Apparently markupsafe is not longer supplying soft_unicode as per this issue: pallets/markupsafe#304
Google Colab is not an option when it comes to company data, I'm afraid. For the LiveLessions it's fine though.
Conda + DataSpell should also do the trick.
Hi! I'm also facing the same issue. I'm using Mac m1 and can't figure out how to deal with this issue. This is my first project.
Ok thanks @caboodle and @Mahrukhw. This seems to be a major issue. Will test on my M1 Mac and hope to resolve the issue soon.
Hi @caboodle and @Mahrukhw — this issue is related to version library dependencies that have changed over time.
There is not an easy way for me to fix it by updating the Dockerfile, so instead I've uploaded a Docker image that should work for you to DockerHub here.
If either (or better yet both!) of you could please confirm that pulling the docker image with docker pull jonkrohn/dltfpt-stack:videos works for you (instead of building the image from the Dockerfile), then I'll update the "official" install instructions for this repo to have everyone pull from DockerHub instead of building the image themselves.
Hi! I had the same problem ImportError: cannot import name 'soft_unicode' from 'markupsafe'. I'm running a Intel MacBook Pro. I downloaded the docker image (docker pull jonkrohn/dltfpt-stack:videos) and I worked just fine.
Ok awesome @brauliomsft — that's great news :D
I'll update the installation instructions in this repo shortly to focus on using the Docker image instead of the Dockerfile.
I had the same problem on a MacBook with M1 chip and with jonkrohn/dltfpt-stack:videos works fine
Ok super duper @albertoforni! Glad it worked for you too. I've been slammed lately but hope to update the install instructions in this repo with the Docker image soon.
Ok I've updated Step 8 of the macOS instructions to reflect pulling the Docker image instead of building it locally.
I'd appreciate if @albertoforni or @brauliomsft could confirm these updated instructions look solid.
@caboodle or @Mahrukhw -- if you're still experiencing this issue, could you please try the updated install instructions? Looking forward to hearing how they go for you!
Ok I've updated Step 8 of the macOS instructions to reflect pulling the Docker image instead of building it locally.
I'd appreciate if @albertoforni or @brauliomsft could confirm these updated instructions look solid.
@caboodle or @Mahrukhw -- if you're still experiencing this issue, could you please try the updated install instructions? Looking forward to hearing how they go for you!
@jonkrohn You may want to change step 9 as well. Command is still running dltfpt-stack, and the build command is removed. So I'm assuming we need to run jonkrohn/dltfpt-stack:videos. Thanks!
You are quite right, @Karthikvishw — thank you! I've updated Step 9 accordingly now and the instructions work properly 🙏