Because SPSS software expects the .sps file to be readable by a human and editable by software, choice of a delimiter is really limited to visible characters that can be created with a keyboard. Localization of data requires UTF-8 coding of text, which is handled badly by SPSS in any format other than delimited.
I think that the Epidata team has chosen a good compromise.
Jamie
On Oct 28, 2013, at 8:47 AM, epidata-list@lists.umanitoba.ca wrote:
On 2013-10-28 13:24, epidata-list@lists.umanitoba.ca wrote:
Den 2013-10-28 10:39 skreiv epidata-list@lists.umanitoba.ca:
In regards to b) the main issue is that the delimiter cannot be part of the data, not even if it is enclosed in quotations. Since a text field in EpiData may contain ANY kind of data (that includes all visible UTF-8 characters), I chose the vertical bar "|" to be the on LEAST likely to be used as part of a text field.
How about using character 0x1E (30) or 0x1F (31) instead. They were *designed* for precisely this problem: http://en.wikipedia.org/wiki/Delimiter#ASCII_delimited_text
That is true, but then you would have the problem with non-displaying characters in a file/text editor - leading to strange symbyls like a square (in Windows) or a questionmark (in eg. Linux). IMHO that is a less desired alternative that using a visible, but infrequently used symbol.
Regards, Torsten Christiansen EpiData. _______________________________________________ EpiData-list mailing list EpiData-list@lists.umanitoba.ca http://lists.umanitoba.ca/mailman/listinfo/epidata-list