mysterious bug in gawk

[Tony Gozdz]

Hi,

Recently, searching for text/data files with a NULL character(s), I found a 
strange bug in gawk (gawk-3.1.6-1-bin.zip, run under WinXP Pro).  Here is a 
listing that demonstrates the problem: as I expand the list of files to search 
through, gawk suddenly crashes; it is as if the file name suddenly got cut... 
There is nothing strange in the files where the program crashes (no NULLs).

I'm not sure how to get around this issue and what causes it.

$ gawk '/\0/ {print FILENAME; nextfile}' TW8287*
$ gawk '/\0/ {print FILENAME; nextfile}' TW828*
TW8286FW08040170002_E8_6m60.065

$ gawk '/\0/ {print FILENAME; nextfile}' TW82*
TW8254EL080208322620_after_OD_35ART.010
TW8254EL080208322659_after_OD_35ART.009
TW8255EL080208414904_after_OD_35ART.012
TW8255EL080208414982_after_OD_35ART.011
TW8286FW08040170002_E8_6m60.065

$ gawk '/\0/ {print FILENAME; nextfile}' TW8*
TW8019EW08010090001_Conoco_SSDE_QF2.054
TW8019EW08010090003_Conoco_SSDE_QF2.055
TW8019EW08010090004_Conoco_SSDE_QF2.057
TW8019EW08010090005_Conoco_SSDE_QF2.058
TW8019EW08010090006_Conoco_SSDE_QF2.059
TW8019EW08010090008_Conoco_SSDE_QF2.060
TW8019EW08010090009_Conoco_SSDE_QF2.061
TW8019EW08010090010_Conoco_SSDE_QF2.062
TW8022EW08010120004_SSDE_1SN_QF2.064
TW8022EW08010120007_SSDE_1SN_QF2.065
TW8022EW08010120010_SSDE_1SN_QF2.063
TW8067FB07120191201_10d00VRT.060
TW8067FB07120192301_10d00VRT.061
TW8067FB07120192349_10d00VRT.062
TW8067FB07120192350_10d00VRT.064
TW8067FB07120192668_10d00VRT.063
TW8070FB0611007_1_QF.083
TW8070FB0611008_2_QF.086
TW8070FB0611012_1_QF.077
TW8070FB0611012_2_QF.078
TW8070FB0611071_rail_10d00VRT.041
TW8090FW08020170003_MCP_QF1.073
TW8090FW08020170004_MCP_QF1.074
TW8090FW08020170005_MCP_QF1.075
TW8090FW08020170006_MCP_QF1.076
TW8090FW08020170007_MCP_10d42V60C.113
TW8090FW08020170008_MCP_QF1.077
TW8090FW08020170009_MCP_10d42V60C.114
TW8090FW08020170009_MCP_QF1.078
TW8090FW08020170010_MCP_QF1.079
TW8254EL080208322620_after_OD_35ART.010
TW8254EL080208322659_after_OD_35ART.009
TW8255EL080208414904_after_OD_35ART.012
TW8255EL080208414982_after_OD_35ART.011
TW8286FW08040170002_E8_6m60.065
gawk: cmd. line:2: (FILENAME=TW8287FW08040180004_E9_for1.048 FNR=474) fatal: 
cannot open file `TW8287FW0804018000
_E9_2w6' for reading (No such file or directory)
$

Here is a listing of files in that area:

TW8287FW08040180001_E9_for.050
TW8287FW08040180001_E9_for1.045
TW8287FW08040180002_E9_for.051
TW8287FW08040180002_E9_for1.046
TW8287FW08040180003_E9_for.052
TW8287FW08040180003_E9_for1.047
TW8287FW08040180004_E9_2w60.044
TW8287FW08040180004_E9_3m60.063
TW8287FW08040180004_E9_6m60.010
TW8287FW08040180004_E9_QF.029
TW8287FW08040180004_E9_for.053
TW8287FW08040180004_E9_for1.048   <== crash location/file
TW8287FW08040180005_E9_2w60.045
TW8287FW08040180005_E9_3m60.064
TW8287FW08040180005_E9_6m60.011
TW8287FW08040180005_E9_QF.030
TW8287FW08040180005_E9_for.054
TW8287FW08040180005_E9_for1.049
TW8287FW08040180006_E9_for.055
TW8287FW08040180006_E9_for1.050
TW8287FW08040180007_E9_for.056
TW8287FW08040180007_E9_for1.051
TW8287FW08040180008_E9_for.057
TW8287FW08040180008_E9_for1.052

--Tony
------------------------
Antoni S. Gozdz, PhD
Principal Scientist
A123Systems
Arsenal on the Charles | 321 Arsenal Street | Watertown MA 02472
P 617 778-5706 | C/BB 857 383-0038 | F 617 924-5749
www.a123systems.com<http://www.a123systems.com>