Reading log files

Distance for Windows writes the log files (the series of pages of information on the fitted models, parameter estimates etc) and graphs to the project database.

These (for CDS analysis) are stored in AnalysisCDSResults and look like this in the database:

> str(aa$AnalysisCDSResults)
'data.frame':    5 obs. of  4 variables:
    $ ID    : int  1 2 3 5 6
    $ Output: Factor w/ 5 levels
"&H3115:xœíZ[oÛ:\022~Î\001ú\037\b,ÎÂNcE’us°yÈqœnд\nšnóL[t¢F\027W’ÓºÈ?ß\031Þtñ½ÉyÈbY
±ÄáÇáÜ8âð`T”QBË(Kɧ\031þ)ÈU\004ïÒ»ƒ7\177¼ùƒ"|
__truncated__,..: 2 1 4 5 3
    $ Log   : Factor w/ 5 levels
"&HB19:xœµVQoÓ0\020~Ÿ´ÿpÚK_F:¦m°MF2m\n\025é\032š\f$\bB^r]Í\034;Ä\016£{à·sN×?&$6\032¢<X>ûóww¾Ï7ÖÒI¡ä?Ô—°½5\021W~0\024NÀH*¼7•
ÂÜI£"|
__truncated__,..: 3 1 4 5 2
    $ Plot  : Factor w/ 2 levels
"&H1E98:xœíØÁr\0337\020\004Ð{ªò\017¼§JžÁ`\006@ÎNrÕ/Ð\024\017¬bD—L\037ò÷\001i'B7˜¥ÖÇ”öFòqµ\\",..:

1 1 1 2 1

Each entry is compressed using zlib based on "freeware code by Phil Fresle, http://www.frez.co.uk".

From the source:

' The compress wrapper adds the original string size onto the
'  front of the compressed string so when it comes to
'  uncompressing you know the size of string you need to allocate.
'  The string size is stored in hex format, with a colon at the
'  end to delimit it from the rest of the compressed string.
'  (e.g., &H8E0:)  Uncompress begins by checking for the &H and
'  if it doesn't find it, assumes there is no uncompressing to be
'  done.  So, this means you can't compress strings that begin with
'  &H!

More info in Distance for Windows source here (private GH repo).

Unfortunately although R include zlib, it isn't exported for others to use see this thread. Also note that memDecompress in base R is not a solution as zlib and gzip are different standards.

For now, reading these files is not supported in readdst.