Summary: | 'probable' bug in ps2epsi script | ||
---|---|---|---|
Product: | Ghostscript | Reporter: | krishan <purak> |
Component: | General | Assignee: | Ray Johnston <ray.johnston> |
Status: | RESOLVED DUPLICATE | ||
Severity: | normal | CC: | master |
Priority: | P2 | ||
Version: | 8.53 | ||
Hardware: | PC | ||
OS: | Linux | ||
Customer: | Word Size: | --- |
Description
krishan
2004-02-18 09:00:22 UTC
The sed script is correct as it stands. The regexp [^!-~] matches any characters between ! and ~ inclusive. The '-' in the regexp indicates a range and is NOT intended to match only '!', '-' and '~', but rather the entire range. This bug will be closed unless a PostScript file is attached with a specific command line that exhibts the problem. The bug is real, but is related to something different. Character ranges are locale dependent, and to make [^!-~] really work as expected you have to add LANG=C at the beginning of the script. GNU sed compiled with older glibc versions aborts on the above regexp if the locale is not C. |