This directory was copied by Don Wells on 10/5/94 from directory ftp://nssdca.gsfc.nasa.gov/fits/errtest/ The files were created by Barry Schlesinger at the NOST FITS Office. -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- (Last update 3 Jan 1994 1500 GMT; XTRACARD, SIZEPROB, alphabetizing) This directory consists of a series of files that can be used to test the ability of a FITS reader to handle errors in the headers of the files that it is trying to read. There is one good test file, containing a brief header along with data. The data are taken off a plot in an Astrophysical Journal paper, cited in the header. All files except one contain the same data. Sometimes, the header contains enough information to read the data; sometimes it doesn't. But the data are there to mimic an attempt at a FITS file that failed because of header errors. All files are of the form *.TST. The contents of the files are as follows: (names now in alphabetical order, except for GOOD) GOOD - The FITS file in conformance, to show how it should appear. BITPIXQ2 - BITPIX has the indecipherable value Q2. BITPIX13 - Here BITPIX has a numerical value, but not one permitted under the FITS rules. DIFFERENT - This file isn't even close to FITS; it's just text. It will show what your software if you apply it to something that isn't even supposed to be FITS. DUPKEY - The NAXIS2 keyword is repeated. If the software keeps the first, it can read the data array; if it uses the second it will lose half. ENDEQUAL - Suggested by Don Wells. In this file, there is an = in column 9 of the END "card", as if there were a value. This error has been known to crash some implementations in the past. EXTRAXIS - NAXIS is 2 but NAXIS3 is present. At present, this case is not forbidden, so your reader should be able to read the file. Many advise against having more axes than the value of NAXIS. FALSEND - This file is in fact a perfectly good FITS file. However, it contains the string 'END ' in the middle of a comment. A good FITS reader will see that this string is not in the proper place and ignore it. A careless FITS reader might interpret it as the end of the header. LATENAXIS - Another case of keywords out of order. Here NAXIS follows the NAXISn. MISSAXIS - Here NAXIS is 2 but NAXIS2 is missing, so the file is not readable. NOEQUAL - The required = is missing from column 9 after the BITPIX NOSIMPLE - This file looks like a FITS file, but it doesn't begin with SIMPLE. Your software should tell you. QIN10 - Column 10 is not blank following the NAXIS keyword. SHUFFLE - The keywords are all there, but they're out of order. Your software doesn't have to be able to find them, but it should give you some idea there's a problem. SIMPLEF - In this case SIMPLE = F. While the accompanying file here is standard FITS, normally, if SIMPLE = F, the file will differ in some way from standard FITS. Your reader should warn you up front. SIMPLEW - The file begins with SIMPLE , but the value is wrong. Your software should catch the value. SIMPLE2ND - Here, some text accidently got in front of the SIMPLE keyword. Can your software handle it? SIZEPROB - The data array size specified in the header (12) does not correspond to the actual array size. Does your software find a problem? One plausible result would be a finding of no fill values after the data array is supposed to have ended. XTRACARD - There is a card image after the END card image, where there should be only fill. Does your software notice?