ENH:read_html() handles tables with multiple header rows #13434

[viswaraavi]

• closes read_html() doesn't handle tables with multiple header rows  #13434
• read_html() handles tables with multiple header rows

Now it produces the following output for the issue
Unnamed: 0 Age Party
0 Hillary 68 D
1 Bernie 74 D
2 Donald 69 R

[TomAugspurger]

Could you add tests for the new behavior, a release note, and a sentence (and maybe example) to the documentation? And the travis failure was for a style check. git diff upstream/master | flake8 --diff.

[codecov-io]

Current coverage is 84.33% (diff: 33.33%)

Merging #13485 into master will decrease coverage by 0.90%

@@             master     #13485   diff @@
==========================================
  Files           140        138     -2   
  Lines         50563      51084   +521   
  Methods           0          0          
  Messages          0          0          
  Branches          0          0          
==========================================
- Hits          43103      43083    -20   
- Misses         7460       8001   +541   
  Partials          0          0          

Powered by Codecov. Last update 1e61aed...8629537

[viswaraavi]

I think I first overlooked the issue. In the second commit I made a change such that we select the first row of the table that contains th tags in thead of the table. And this should make previous test cases to pass. I will add some new test cases to test the modified behaviour and also i will change the documentation

[viswaraavi]

I see that there are no tests added for beautifulsoup parser. is there any reason for that ?. And I think so that is the reason for the decrease in code coverage

[jreback]

your test should be called by each parser unless you put it in a specific class (which only exercises specific parsers)

[jreback]

can you rebase / update?

[jreback]

needs tests.

please reopen / comment if you are continuing.