From: David G. <go...@py...> - 2004-09-13 00:25:42
|
[Link, Michael (ESA)] > I'm using the handy csv-table directive. A look in the code revealed > that you use the csv-package of Python 2.3. Now it seems that I have > a encoding error somewhere in my csv data In fact, a look at the Python Library Reference reveals this: Note: This version of the csv module doesn't support Unicode input. Both Felix and I came up with the same workaround, which encodes the data as UTF-8, sends it to the CSV parser, then decodes the UTF-8 back to Unicode. It ought to work. > and I want to suppress it with > > --output-encoding-error-handler=replace If there's still a problem with your data, that won't help, because you need to check the *input*. I've added an --input-encoding-error-handler command-line option, as well as an ":encoding:" directive option for "csv-table" ("include" and "raw" directives too). These should help. The external file encoding defaults to the document's encoding. The directive uses the session's error handler (I didn't see much point in allowing separate settings). -- David Goodger <http://python.net/~goodger> |