Re: th role mapping details - for review

Ok, good. Please let me know if it can be improved in some way.
Thanks.
Alexander.

On Mon, Dec 29, 2014 at 7:04 PM, Julie Jeongeun Kim <je00julie.kim@gmail.com
> wrote:

> Hi Alexander,
>
> At first time, when you gave me that description,
> I thought the both conditions are at the same level and I couldn't
> understand what caused to assign different role to case 4 and case 6.
> Now I can see what you mean.
> The first line is the first condition and then the second sentence is the
> second condition.
> Thanks for helping me understand.
>
> I think Firefox has a reasonable solution and appreciate that I can refer
> to that.
>
> Thanks,
> Julie
>
> 2014-12-30 0:30 GMT+09:00 Alexander Surkov <surkov.alexander@gmail.com>:
>
>> Hi, Julie. Is your wording different from the wording I gave for current
>> Firefox algorithm? It was:
>>
>> * if header is preceded by header cells then it's a column header, if it
>> is preceded by cells then it's a row header
>> * if header is followed by header cells then it's a column header, if it
>> is followed by cells then it's a row header
>>
>> On Mon, Dec 29, 2014 at 10:22 AM, Julie Jeongeun Kim <
>> je00julie.kim@gmail.com> wrote:
>>
>>> Hi Alexsander,
>>> Sorry for causing confusion.
>>> Here is correction.
>>>
>>> First we can check the previous cell.
>>> * if header is preceded by header cells then it's a column header,
>>> if it is preceded by cells then it's a row header
>>> We can cover cases except case 1), case 4), case 5) and case 7) because
>>> they have header cells at row index 0.
>>>
>>> And then,
>>> * if header has no previous cell (row index is 0) then we can check the
>>> next cell.
>>>     if header is followed by header cells then it's a column header,
>>>     if it is followed by cells then it's a row header
>>> We can cover case 1,4,5 and 7 without breaking others.
>>>
>>> What do you think?
>>>
>>> 2014-12-29 23:50 GMT+09:00 Alexander Surkov <surkov.alexander@gmail.com>
>>> :
>>>
>>>> Hi, Julie. I'm confused by terms.
>>>>
>>>> On Fri, Dec 26, 2014 at 1:04 AM, Julie Jeongeun Kim <
>>>> je00julie.kim@gmail.com> wrote:
>>>>
>>>>> Hi Alexander,
>>>>>
>>>>> First we can check the previous cell.
>>>>> * if header is followed by header cells then it's a column header,
>>>>>
>>>>
>>>> did you mean "preceded" rather than "followed" here?
>>>>
>>>>
>>>>> if it is followed by cells then it's a row header
>>>>> We can cover cases except case 4), case 5) and case 7) because they
>>>>> have header cells at row index 0.
>>>>>
>>>>> And then,
>>>>> * if header has no previous cell (row index is 0) then we can check
>>>>> the next cell.
>>>>>     if header is preceded by header cells then it's a column header,
>>>>>     if it is preceded by cells then it's a row header
>>>>>
>>>>> We can cover case 4,5 and 7 without breaking others.
>>>>> What do you think?
>>>>>
>>>>> 2014-12-24 23:33 GMT+09:00 Alexander Surkov <
>>>>> surkov.alexander@gmail.com>:
>>>>>
>>>>>> Hi, Julie. That was just an algorithm, it's not necessary perfect,
>>>>>> and probably wasn't even designed keeping in mind edge cases. I think we
>>>>>> should do what's reasonable. What do you think we should do for case 6?
>>>>>>
>>>>>> On Thu, Dec 18, 2014 at 7:32 PM, Julie Jeongeun Kim <
>>>>>> je00julie.kim@gmail.com> wrote:
>>>>>>
>>>>>>> Thanks for your detail description.
>>>>>>> Your description covers generally all cases.
>>>>>>> But I'm not clear about the case 4 and the case 6.
>>>>>>> The case 4 and the case 6 are similar.
>>>>>>>
>>>>>>> Case 4)
>>>>>>> The first th is column header role because it is preceded by header
>>>>>>> cell as you described.
>>>>>>>
>>>>>>> Case 6)
>>>>>>> The first th is row header role even though it is preceded by header
>>>>>>> cell.
>>>>>>>
>>>>>>> Did I miss something?
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Julie
>>>>>>>
>>>>>>>
>>>>>>> 2014-12-19 1:05 GMT+09:00 Alexander Surkov <
>>>>>>> surkov.alexander@gmail.com>:
>>>>>>>>
>>>>>>>> sure :)
>>>>>>>>
>>>>>>>> On Thu, Dec 18, 2014 at 9:28 AM, Joseph Scheuhammer <
>>>>>>>> clown@alum.mit.edu> wrote:
>>>>>>>>>
>>>>>>>>> On 2014-12-18 9:15 AM, Alexander Surkov wrote:
>>>>>>>>>
>>>>>>>>>> Here's correction.
>>>>>>>>>> * if header is preceded by header cells then it's a column
>>>>>>>>>> header, if it is preceded by cells then it's a raw header
>>>>>>>>>> * if header is followed by header cells then it's a column
>>>>>>>>>> header, if it is followed by cells then it's a raw header
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> You mean "row" header, not "raw" header, right?
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> ;;;;joseph.
>>>>>>>>>
>>>>>>>>> 'Array(16).join("wat" - 1) + " Batman!"'
>>>>>>>>>            - G. Bernhardt -
>>>>>>>>>
>>>>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>