From: Ricardo S. de A. <rsa...@li...> - 2007-09-20 03:12:58
|
On Monday 17 September 2007 11:28:08 Subrata Modak wrote: >1) We should not touch pan code for generating this output, less, we >will introduce far more bugs into pan. We can write a script which will >parse the normal LTP output file ( or log, or, failed file(s) for that >matter) and generate this output format and of course a separate output >file. Now the question arises, is, whether we should have this in: > XML ? > or HTML ? >But, whatever the case, we need to retain the original file formats and >then only parse them to generate the new format(s). Yep, please, let maintain how pan is generating the results output and just= =20 create some parsers that could write the output in another format. But my question is, do we really need to have a parser to geneate kind of a= =20 HTML file? I guess that the normal text file is already enough. >3) Ricardo=C5=9B patch (applied earlier) to provide the name of Failed file >(containing the tag name and cmdline entries of failed test-cases) does >not write error info in to the same file. May be we need to produce >error info in that file as well, although it is existing in the output >file. So, the new format can also be made applicable for the failed file >as well. >** Creation of failed filed is now made mandatory by runltp I didn't want to add the error info at the same file so we could easily ask= =20 LTP to run only the failed test cases. As we already have all the output at= =20 another file, I don't think that we should modify how we're creating this=20 file. I guess that what we have at the output file is enough to see what's=20 happening, like: <<<test_start>>> tag=3Dgf17 stime=3D1189631960 cmdline=3D"growfiles -W gf17 -b -e 1 -i 0 -L 120 -u -g 5000 -T 100 -t=20 499990 -l -C 10 -c 1000 -S 10 -f Lgf03_" contacts=3D"" analysis=3Dexit initiation_status=3D"ok" <<<test_output>>> gf17 1 PASS : Test passed <<<execution_status>>> =2D-=20 Ricardo Salveti de Araujo |