BUG: Fixed to_html with index=False and max_rows

[TomAugspurger]

closes #14998

Previously raised an IndexError by assuming that
len(fmt_values) was always the same length as len(self.data).

[codecov-io]

Current coverage is 84.78% (diff: 100%)

Merging #14999 into master will increase coverage by <.01%

@@             master     #14999   diff @@
==========================================
  Files           145        145          
  Lines         51090      51090          
  Methods           0          0          
  Messages          0          0          
  Branches          0          0          
==========================================
+ Hits          43315      43316     +1   
+ Misses         7775       7774     -1   
  Partials          0          0          

Powered by Codecov. Last update 7f0eefc...2bd8629

[jorisvandenbossche]

You now tested with max_rows=1, but am I correct that with max_rows=2 you would not get the ... row? Or if so maybe include that in the test?

[TomAugspurger]

but am I correct that with max_rows=2 you would not get the ... row?

Looks like we have some issues with the interaction between index and max_rows

I think the third one is buggy, index=False should not affect whether or not the ... is drawn. Do that as a followup?

[jorisvandenbossche]

I can as a follow-up, but possibly you can also solve it at once here. If you use the _write_regular_rows whether index is false or not, but let _write_regular_rows adapt based on that: eg extend values with index er not: https://github.com/pandas-dev/pandas/pull/14999/files#diff-893f461e63cbb3ced9ab2c53328416a0L1216 and then pass an argument to self.write_tr to specify there should be a th or not (https://github.com/pandas-dev/pandas/pull/14999/files#diff-893f461e63cbb3ced9ab2c53328416a0L1222).
But not sure how many side effects this would have that makes it more complicated. So as a follow-up is fine as well.

[jreback]

thanks @TomAugspurger

[jorisvandenbossche]

Opened #15019 for the remaining issue with the truncation