[wtr-general] Re: Any fast ways to parse a table

2013-08-02 Thread Chuck van der Linden
I doubt it would beat the speed of nokogiri, but I have to wonder, if you were to assemble the row contents as a string or regex, lets call it 'expected', and then simply look for the row via .row(:text => expected). how fast would that be? On Monday, July 15, 2013 8:51:09 AM UTC-7, Super Kev

[wtr-general] Re: Any fast ways to parse a table

2013-07-15 Thread Super Kevy
Awesome just the advise I'm looking for, thanks for contributing. On Friday, July 12, 2013 9:31:20 PM UTC-5, Justin Ko wrote: > > The "CheckTableUsingRowArray" approach does not seem faster to me. The > location of the timer excludes the Watir portion of the code, which is > costly. In partic

[wtr-general] Re: Any fast ways to parse a table

2013-07-12 Thread Justin Ko
The "CheckTableUsingRowArray" approach does not seem faster to me. The location of the timer excludes the Watir portion of the code, which is costly. In particular, the following line adds about 28 seconds for me: myTableArray = obj.to_a #Array the table An alternative approach is to use the N

[wtr-general] Re: Any fast ways to parse a table

2013-07-12 Thread Super Kevy
OK. I ran a sample and have some benchmarks for a simple table look up In the sample below I have a simple table with 30 row of 8 columns. The test measure the time to find the matching row (the sample is row 30) There are two methods. One uses if comparisons of the tables columns within the ro

[wtr-general] Re: Any fast ways to parse a table

2013-06-12 Thread Super Kevy
Good advice. Maybe I can strut up some samples and get some time measurements. Then later post it for review. On Jun 11, 9:47 am, Oscar Rieken wrote: > collect each row as a hash colum as the key, then compare that hash to the > expected data hash, personally without more info of why you want to

Re: [wtr-general] Re: Any fast ways to parse a table

2013-06-11 Thread Oscar Rieken
collect each row as a hash colum as the key, then compare that hash to the expected data hash, personally without more info of why you want to do this i cant tell you an easier way On Tue, Jun 11, 2013 at 10:42 AM, Dan wrote: > Ah, I missed the part where you're trying to match data in multip

[wtr-general] Re: Any fast ways to parse a table

2013-06-11 Thread Dan
Ah, I missed the part where you're trying to match data in multiple columns. On Tuesday, June 11, 2013 9:32:03 AM UTC-4, Super Kevy wrote: > > The clicking is no problem. Its matching the 7 columns of data to > make sure I've got the correct row. > The matching is what is taking so long. > > C

[wtr-general] Re: Any fast ways to parse a table

2013-06-11 Thread Super Kevy
The clicking is no problem. Its matching the 7 columns of data to make sure I've got the correct row. The matching is what is taking so long. Currently I push a table row into an array and match with .include? which beats a bunch of if A=B comparisons. Would it be faster to push the table into a

[wtr-general] Re: Any fast ways to parse a table

2013-06-10 Thread Dan
I use something like the below. Find the row you're looking for and then click on the column and element within that column. You might be able to get away with less, but this works pretty well. b.tr(:text => /the row I'm looking for/).td(:index => 5).img.click On Friday, June 7, 2013 4:36:21