Hello Anand, What would be the starting point to identify the parts that need more test coverage?
On Tue, Jun 21, 2016 at 8:12 PM, Anand Chitipothu <[email protected]> wrote: > On Tue, 21 Jun 2016 at 14:48 Mathieu Xhonneux <[email protected]> wrote: >> >> Hi, >> >> I've been hired by the CS departement of the Université Catholique de >> Louvain (UCL, Belgium) to port web.py to Python 3. We're using the framework >> in some of our internal tools and we do need a Py3 upgrade for them. > > > Thats a great news! And thank you for taking up this task. > >> >> I've already done some work, available here : >> https://github.com/Zashas/webpy/tree/py3 >> We saw that Anand began the migration 3 years ago and decided to pursue >> his work (making the codebase compatible with Py2 and Py3, which seems to me >> the best option, btw Django manages it successfully). I integrated the >> commits he already wrote back then, ported all the unittests to Py3 and now >> I'm busy making them pass in both versions of Python. > > > I'll review it in next couple of days and lets try to merge into master as > soon as possible. > > The current version of python maintains compatibility with Python 2.4 > onwards. I think we can drop 2.4 and 2.5 completely. I'm in favour of > dropping 2.6 as well and retain only 2.7 and Python 3.5+. > >> >> >> The ORM seems to be OK (I'm kinda hoping the tests' coverage is wide >> enough), but it originally supports some DB backends that are either no >> longer maintained, or not ported to Py3 : >> >> psycopg1 : evolved to psycopg2, I'm guessing that not much people are >> using it these days >> pysqlite2 : not ported to Py3 >> MySQLdb : not ported to Py3, but mysqlclient seems to be compatible and to >> do the job fairly enough >> DBUtils : not ported to Py3 >> >> Maybe we should consider dropping support for these ? Their presence is >> not doing any harm in the codebase, but I don't think that psycopg1 and >> pysqlite2 are still useful, and removing them could clean the code a bit. > > > I think db module is very important piece. Yes, we can drop all those legacy > postgres modules. Switching to mysqlclient for Python3 seems likes a good > idea to me. > It is desirable to have connection polling. If DBUtils doesn't work with > Python 3, lets try to find alternative. > >> >> Anyway, if anyone wants to follow my work, I'm open to any remark, >> suggestion or patch. I'm hoping to have ported everything by friday, but it >> will definitively require some testing. > > > Test coverage is really poor for web.py. Good to have that fixed as we are > making lot of changes. Any one interested to help with that? > > Anand > > -- > You received this message because you are subscribed to the Google Groups > "web.py" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at https://groups.google.com/group/webpy. > For more options, visit https://groups.google.com/d/optout. -- "The difference between theory and practice tends to be very small in theory, but in practice it is very large indeed." - Anon www.99nth.com/~krm -- You received this message because you are subscribed to the Google Groups "web.py" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/webpy. For more options, visit https://groups.google.com/d/optout.
