Bug 70507
| Summary: | RE in BUG section of re_format(7) in obsolete notation | ||
|---|---|---|---|
| Product: | Documentation | Reporter: | Marian Cerny <jojo> |
| Component: | Books & Articles | Assignee: | freebsd-doc (Nobody) <doc> |
| Status: | Closed FIXED | ||
| Severity: | Affects Only Me | ||
| Priority: | Normal | ||
| Version: | Latest | ||
| Hardware: | Any | ||
| OS: | Any | ||
On 2004-08-16 01:30, Marian Cerny <jojo@matfyz.cz> wrote: > Shouldn't this paragraph in BUGS section in manual page of re_format: > > Back references are a dreadful botch, posing major problems for > efficient implementations. They are also somewhat vaguely > defined (does `a\(\(b\)*\2\)*d' match `abbbd'?). > Avoid using them. > > be > > Back references are a dreadful botch, posing major problems for > efficient implementations. They are also somewhat vaguely > defined (does `a((b)*\2)*d' match `abbbd'?). > Avoid using them. > > because `a\(\(b\)*\2\)*d' is in obsolete notation? Or does this bug > concern only the obsolete REs? You're probably right that we should change the syntax to look like a modern RE. The basic RE syntax is still used by many utils in the base system though. This is probably why the regexp has remained as you see it now. Daniel, Ruslan and David... what do you think? Is this change ok? - Giorgos Responsible Changed From-To: freebsd-doc->keramida I'll see what I can do about this one. Responsible Changed From-To: keramida->freebsd-doc Put this back in the free pool. I don't think I can handle it appropriately. Giorgos Keramidas wrote: > On 2004-08-16 01:30, Marian Cerny <jojo@matfyz.cz> wrote: > >>Shouldn't this paragraph in BUGS section in manual page of re_format: >> >> Back references are a dreadful botch, posing major problems for >> efficient implementations. They are also somewhat vaguely >> defined (does `a\(\(b\)*\2\)*d' match `abbbd'?). >> Avoid using them. >> >>be >> >> Back references are a dreadful botch, posing major problems for >> efficient implementations. They are also somewhat vaguely >> defined (does `a((b)*\2)*d' match `abbbd'?). >> Avoid using them. >> >>because `a\(\(b\)*\2\)*d' is in obsolete notation? Or does this bug >>concern only the obsolete REs? > > > You're probably right that we should change the syntax to look like a > modern RE. The basic RE syntax is still used by many utils in the base > system though. This is probably why the regexp has remained as you see > it now. > > Daniel, Ruslan and David... what do you think? Is this change ok? Old, OLD messages... This was lost in a number of spams I'm happing to be clearing right now. Thing about back references is... they didn't work with Extended Regex, only with basic Regex, which is the obsolete notation. So I'm guessing the rewritten example wouldn't work, because back references is not supported with that syntax. So, if this change was done, could someone check if back references are actually supported in extended regex (the modern syntax), and, if not, undone this change? :-) -- Daniel C. Sobral (8-DCS) dcs@newsguy.com dcs@freebsd.org capo@the.great.underground.bsdconpiracy.org In related news Microsoft Windows users are now covered under the Americans with Disabilities Act. On 2005-07-09 23:01, "Daniel C. Sobral" <dcs@newsguy.com> wrote: >Giorgos Keramidas wrote: >>On 2004-08-16 01:30, Marian Cerny <jojo@matfyz.cz> wrote: >>> Shouldn't this paragraph in BUGS section in manual page of re_format: >>> >>> Back references are a dreadful botch, posing major problems for >>> efficient implementations. They are also somewhat vaguely >>> defined (does `a\(\(b\)*\2\)*d' match `abbbd'?). >>> Avoid using them. >>> >>> be >>> >>> Back references are a dreadful botch, posing major problems for >>> efficient implementations. They are also somewhat vaguely >>> defined (does `a((b)*\2)*d' match `abbbd'?). >>> Avoid using them. >>> >>> because `a\(\(b\)*\2\)*d' is in obsolete notation? Or does this bug >>> concern only the obsolete REs? >> >> You're probably right that we should change the syntax to look like a >> modern RE. The basic RE syntax is still used by many utils in the base >> system though. This is probably why the regexp has remained as you see >> it now. >> >> Daniel, Ruslan and David... what do you think? Is this change ok? > > Old, OLD messages... This was lost in a number of spams I'm happing to > be clearing right now. Thing about back references is... they didn't > work with Extended Regex, only with basic Regex, which is the obsolete > notation. > > So I'm guessing the rewritten example wouldn't work, because back > references is not supported with that syntax. So, if this change was > done, could someone check if back references are actually supported in > extended regex (the modern syntax), and, if not, undone this change? :-) Nothing was changed, since I wasn't sure of what to do. Thanks for the clarification :-) Does this mean we can close this PR now? On 2005-07-10 15:05 +0300, Giorgos Keramidas wrote: > On 2005-07-09 23:01, "Daniel C. Sobral" <dcs@newsguy.com> wrote: > >Giorgos Keramidas wrote: > >>On 2004-08-16 01:30, Marian Cerny <jojo@matfyz.cz> wrote: > >>> Shouldn't this paragraph in BUGS section in manual page of re_format: > >>> > >>> Back references are a dreadful botch, posing major problems for > >>> efficient implementations. They are also somewhat vaguely > >>> defined (does `a\(\(b\)*\2\)*d' match `abbbd'?). > >>> Avoid using them. > >>> > >>> be > >>> > >>> Back references are a dreadful botch, posing major problems for > >>> efficient implementations. They are also somewhat vaguely > >>> defined (does `a((b)*\2)*d' match `abbbd'?). > >>> Avoid using them. > >>> > >>> because `a\(\(b\)*\2\)*d' is in obsolete notation? Or does this bug > >>> concern only the obsolete REs? > >> > >> You're probably right that we should change the syntax to look like a > >> modern RE. The basic RE syntax is still used by many utils in the base > >> system though. This is probably why the regexp has remained as you see > >> it now. > >> > >> Daniel, Ruslan and David... what do you think? Is this change ok? > > > > Old, OLD messages... This was lost in a number of spams I'm happing to > > be clearing right now. Thing about back references is... they didn't > > work with Extended Regex, only with basic Regex, which is the obsolete > > notation. > > > > So I'm guessing the rewritten example wouldn't work, because back > > references is not supported with that syntax. So, if this change was > > done, could someone check if back references are actually supported in > > extended regex (the modern syntax), and, if not, undone this change? :-) > > Nothing was changed, since I wasn't sure of what to do. > > Thanks for the clarification :-) Now, when I am reading the manpage once again, i noticed that back references are not supported in modern REs. But implementation in FreeBSD (egrep) supports them and the modified example works. > Does this mean we can close this PR now? Yes, if back references are not supported by POSIX in modern RE, the PR can be IMHO closed. -- Marian Cerny <jojo@matfyz.cz> Jabber: jojo@njs.netlab.cz [ UNIX is user friendly. It's just selective about who its friends are. ] State Changed From-To: open->closed Submitted says it's ok to close this PR. |
Shouldn't this paragraph in BUGS section in manual page of re_format: Back references are a dreadful botch, posing major problems for efficient implementations. They are also somewhat vaguely defined (does `a\(\(b\)*\2\)*d' match `abbbd'?). Avoid using them. be Back references are a dreadful botch, posing major problems for efficient implementations. They are also somewhat vaguely defined (does `a((b)*\2)*d' match `abbbd'?). Avoid using them. because `a\(\(b\)*\2\)*d' is in obsolete notation? Or does this bug concern only the obsolete REs? How-To-Repeat: man 7 re_format