We've noticed a bug in ksh93 this weekend related to reporting of line numbers in errors and $LINONO: Some of our big data production systems generate scripts to handle a specific task. Because of the size and amount of datasets at CERN, it is possible to get well beyond 2^32 lines in a generated script. We've noticed that if the script is sufficiently long (> 2^32 lines) the line numbers reset either to 1 or become random numbers
ksh -c 'integer i imax ; ((imax=2**33)) ; for ((i=0; i < imax ; i+=1024 )) ; do printf "#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n#\n# %x\n" i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i ; done' >bigscript.sh # now add error at the end echo "))l/" >>bigscript.sh If I run this I get line number 1: bigscript.sh: syntax error at line 1: `)' unexpected It would be nice to correct this and use int64_t from <stdint.h> as default datatype for line numbers. Lionel _______________________________________________ ast-developers mailing list ast-developers@research.att.com https://mailman.research.att.com/mailman/listinfo/ast-developers
