long string variable split when reading in SPSS

[ofajardo]

hi

When writing a sav SPSS file with Readstat containing long string variables (long_readstat.sav), Readstat reads them correctly, however SPSS splits it (LongString2, "V2_A1", "V2_A2", "V2_A3"), this does not happen if the name of the variable is shorter (LongStr instead of LongString2). For comparison the same content but written with SPSS (long_spss.sav) it opens correctly both in SPSS and Readstat.

original reports: https://github.com/Roche/pyreadstat/issues/119 https://github.com/Roche/pyreadstat/issues/118

longstring.zip

[mtr]

I experience the same problem. Can someone please have a look at this?

[evanmiller]

Interesting. It looks like SPSS is using internal names LONGS0, LONGS1, LONGS2 rather than V2_A etc. It looks like the variable-naming algorithm needs to be updated.

[ofajardo]

solved in version 1.1.9

[ofajardo]

I was made aware that the issue is gone when reading the file with Pyreadstat, however, the issue is still there when reading the file with SPSS. Actually I forgot about this difference, now I notice that even I wrote at the first description of this issue that Readstat reads correctly but SPSS does not.