ysh329 commented on issue #14505:
URL: https://github.com/apache/tvm/issues/14505#issuecomment-1510784783

   > > Hi @driazati I'm recently prepraing the release note, but I met a 
problem after create csv files. My step is following 
[readme](https://github.com/apache/tvm/blob/main/tests/scripts/release/README.md),
 it mentions below:
   > > > You can then import this CSV into a collaborative spreadsheet editor 
to distribute the work of categorizing PRs for the notes. Once done, you can 
download the resulting CSV and convert it to readable release notes.
   > > 
   > > 
   > > However, after csv files generated and make each row own its category by 
handed, but I'm confused about:
   > > I met error when executing `python3 ./make_notes.py --notes-csv 
out-large.csv`:
   > > ```shell
   > > {'oid': '8e9216013ced6b9655359a696a5bb1bf68fdd638', 'committedDate': 
'2023-04-08T12:48:48Z', 'associatedPullRequests': {'nodes': [{'number': 14533, 
'additions': 2, 'changedFiles': 1, 'de
   > > letions': 1, 'author': {'login': 'shtinsa'}, 'title': '[hexagon] Hexagon 
inference fix', 'body': '1. Resolves issue with kernels linking.\r\n2. Fixes 
warning generated by clang-15.'}]}}
   > > Opening CSV                                                              
                                                                                
                                 
   > > Traceback (most recent call last):
   > >   File "/home/ysh329/code/tvm/tests/scripts/release/make_notes.py", line 
92, in <module>
   > >     category = row["category"].strip()
   > > KeyError: 'category'
   > > ```
   > > 
   > > 
   > >     
   > >       
   > >     
   > > 
   > >       
   > >     
   > > 
   > >     
   > >   
   > > What does the excepted categoried csv file look like before executing 
`make_notes.py`? It seems csv file generated by `gather_prs.py` which is not 
compatible with `make_note.py`.
   > > Can u give me some suggestions?
   > 
   > Hey, those scripts are mostly just a guideline, we haven't really devoted 
support to making it a polished pipeline for generating release notes. For 
those to work there is a manual step of categorizing each PR listed from 
`gather_prs.py` into a CSV with the headers referenced in `make_notes.py`, so 
that's why they don't hook straight up to each other. I've attached a sample of 
the list of categorized PRs used in v0.9.0 for reference: [Release Notes v0.9.0 
- PRs with _ 150 lines 
changed.csv](https://github.com/apache/tvm/files/11246992/Release.Notes.v0.9.0.-.PRs.with._.150.lines.changed.csv).
   > 
   > Either way the scripts aren't strictly necessary, you can always make the 
release notes manually if that ends up easier.
   
   Thanks your reply. Lean a lot from your PRs. 
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to