Geek Answers Handbook

Does pyodbc have any design advantages over pypyodbc?

I know pyodbc is an older project and probably more functional and reliable, but is there anything about its design (component-based compiled C code) that would make it preferable to implement a pure Python implementation like pypyodbc ?

I work a lot in ETL and am thinking of moving from the Linux / Jython / JDBC approach to the Windows / Cygwin / Python / ODBC approach.

+7
source share
1 answer

The potential benefits of pyodbc over pypyodbc written in C would be:

The potential benefits of pypyodbc over pyodbc written in Python are:

  • Less likely to contain problems with pointers to C
  • Slightly less likely memory allocation problems
  • Easier to maintain; a higher level language means fewer lines of code.
  • It is much easier to install without compilation problems requiring a separate build for individual versions of Python, the platform, etc.

Maturity Benefits:

  • Less mistakes
  • More comprehensive coverage of features
  • The best processing of corner cabinets

Maturity is largely dependent on pyodbc not buggy. In my experience, this is not true, and it had a fair amount of memory leaks , etc. And look at the comments for this answer - there is evidence that pyodbc can be quite problematic.

The author claims that pypyodbc is a reimplementation of pyodbc code in Python, and that would mean that the function coverage should be equivalent. There may be some angular cases that have been less studied in the newer code.

Disclaimer: I have not tried pypyodbc

+4
source

More articles:


All Articles