I recently downloaded the GEDCOM Torture Test files. Looking at the HEAD record source corporation address, it seems to be in error. According to the 5.5 spec, the ADDR and ADR1 tags should contain the same information, as should the first CONT and ADR2 tags. (See page 37, ADDRESS_LINE, ADDRESS_LINE1, and ADDRESS_LINE2). The spec implies that ADR1 and ADR2 can be used *instead* of ADDR and CONT. The spec says nothing about using them simultaneously or about a data clash.
On Sun, 26 Jun 2011 05:58:33 -0700, Zev wrote: > I recently downloaded the GEDCOM Torture Test files. Looking at the HEAD > record source corporation address, it seems to be in error. > > According to the 5.5 spec, the ADDR and ADR1 tags should contain the > same information, as should the first CONT and ADR2 tags. (See page 37, > ADDRESS_LINE, ADDRESS_LINE1, and ADDRESS_LINE2). The spec implies that > ADR1 and ADR2 can be used *instead* of ADDR and CONT. The spec says > nothing about using them simultaneously or about a data clash. Perhaps it's intentional??? After all, the file *is* intended to test a program's ability to properly import gedcom files. If the program doesn't flag the error, it'd fail the test, right? Todd
Actually, no. The torture test files are designed to present a completely accurate GEDCOM file containing every possible tag. It's too easy to create an incorrect one. Zev On Jun 26, 9:22 am, Todd Carnes <[email protected]> wrote: > On Sun, 26 Jun 2011 05:58:33 -0700, Zev wrote: > > I recently downloaded the GEDCOM Torture Test files. Looking at the HEAD > > record source corporation address, it seems to be in error. > > > According to the 5.5 spec, the ADDR and ADR1 tags should contain the > > same information, as should the first CONT and ADR2 tags. (See page 37, > > ADDRESS_LINE, ADDRESS_LINE1, and ADDRESS_LINE2). The spec implies that > > ADR1 and ADR2 can be used *instead* of ADDR and CONT. The spec says > > nothing about using them simultaneously or about a data clash. > > Perhaps it's intentional??? > > After all, the file *is* intended to test a program's ability to properly > import gedcom files. If the program doesn't flag the error, it'd fail the > test, right? > > Todd