Re: [courier-users] Courier::Filter 0.16
Sam Varshavchik wrote: I think that ctlfile should be opened for append mode. ...capital idea. Am I wrong to assume that all of the calls to open() in the openctl() function should have the same flags? --- courier/submit2.C.orig Wed Jan 26 18:38:43 2005 +++ courier/submit2.C Wed Jan 26 18:42:17 2005 @@ -309,7 +309,7 @@ current_submit_file=this; int nfd=open(filename, -O_WRONLY | O_TRUNC | O_CREAT | O_EXCL, +O_WRONLY | O_TRUNC | O_CREAT | O_EXCL | O_APPEND, PERMISSION); if (nfd = 0) @@ -349,7 +349,7 @@ if (w == -1 errno == ECHILD) break; int nfd=open(filename, -O_WRONLY | O_TRUNC | O_CREAT, +O_WRONLY | O_TRUNC | O_CREAT | O_EXCL | O_APPEND, PERMISSION); if (nfd = 0) @@ -362,7 +362,7 @@ filename=namefile(C, num_control_files_created); int nfd=open(filename, -O_WRONLY | O_TRUNC | O_CREAT, +O_WRONLY | O_TRUNC | O_CREAT | O_EXCL | O_APPEND, PERMISSION); if (nfd = 0)
Re: [courier-users] Courier::Filter 0.16
Sam Varshavchik wrote: It's always safe to append more stuff to the control file. Whether it's safe to rewrite it from scratch is something I need to think about. At the time submit calls run_filter, that's not true. ctlfile is open and the write pointer will indicate the end of the file. If a filter adds data to the control file, that pointer does not change, so anything that submit writes to the file will overwrite the records that were added by the filter. I've tried calling closectl() and openctl(), but that causes submit to create a new control file, which is messy (and may crash submit; submitting a new message didn't work and I didn't spend a lot of time tracking down the problem I'd created). ctlfile doesn't seem to support a seekp() method that can jump to the end of the file. Do you have any suggestions regarding what sort of change you'd be most willing to accept which would allow me to make changes to the ctlfile from a filter program? --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Gordon Messmer wrote:
ctlfile doesn't seem to support a seekp() method that can jump to the
end of the file.
Attached is a patch which adds and uses such a method. I can now modify
the control file from a courierfilter without corrupting the file.
diff -ru courier-0.48.2.orig/afx/afx.h courier-0.48.2/afx/afx.h
--- courier-0.48.2.orig/afx/afx.h 2001-08-07 17:53:30.0 -0700
+++ courier-0.48.2/afx/afx.h 2005-01-26 11:18:46.019909831 -0800
@@ -375,6 +375,7 @@
void close();
void seekg(std::streampos);
void seekp(std::streampos);
+ void seekp(std::streamoff off, std::ios::seekdir dir);
int handle() { return (fd); }
void handle(int f);
int sync();
@@ -392,6 +393,7 @@
void close() { fd_.close(); }
void seekg(std::streampos p) { fd_.seekg(p); }
void seekp(std::streampos p) { fd_.seekp(p); }
+ void seekp(std::streamoff off, std::ios::seekdir dir) { fd_.seekp(off, dir); }
int fd() { return (fd_.handle()); }
void fd(int f) { fd_.handle(f); }
void sync() { fd_.sync(); }
@@ -405,6 +407,7 @@
void close() { fd_.close(); }
void seekg(std::streampos p) { fd_.seekg(p); }
void seekp(std::streampos p) { fd_.seekp(p); }
+ void seekp(std::streamoff off, std::ios::seekdir dir) { fd_.seekp(off, dir); }
int fd() { return (fd_.handle()); }
void fd(int f) { fd_.handle(f); }
void sync() { fd_.sync(); }
@@ -418,6 +421,7 @@
void close() { fd_.close(); }
void seekg(std::streampos p) { fd_.seekg(p); }
void seekp(std::streampos p) { fd_.seekp(p); }
+ void seekp(std::streamoff off, std::ios::seekdir dir) { fd_.seekp(off, dir); }
int fd() { return (fd_.handle()); }
void fd(int f) { fd_.handle(f); }
void sync() { fd_.sync(); }
diff -ru courier-0.48.2.orig/afx/afxpipe.C courier-0.48.2/afx/afxpipe.C
--- courier-0.48.2.orig/afx/afxpipe.C 2001-08-05 13:00:33.0 -0700
+++ courier-0.48.2/afx/afxpipe.C 2005-01-26 11:19:40.829108003 -0800
@@ -56,6 +56,17 @@
setg(buffer, buffer+sizeof(buffer), buffer+sizeof(buffer));
}
+void afxpipestreambuf::seekp(std::streamoff off, std::ios::seekdir dir)
+{
+ if (sync())
+ return;
+
+ if (lseek(fd, off, dir) 0 iosp)
+ iosp-setstate(std::ios::badbit);
+ else
+ setg(buffer, buffer+sizeof(buffer), buffer+sizeof(buffer));
+}
+
int afxpipestreambuf::sync()
{
if (fd 0)
diff -ru courier-0.48.2.orig/courier/submit2.C courier-0.48.2/courier/submit2.C
--- courier-0.48.2.orig/courier/submit2.C 2005-01-26 10:03:49.318271000 -0800
+++ courier-0.48.2/courier/submit2.C 2005-01-26 11:24:10.878229077 -0800
@@ -945,6 +945,9 @@
SubmitFile::get_msgid_for_filtering, voidp))
return (1);
+ /* seek to end of file in case file was modified. */
+ ctlfile.seekp(0, std::ios::end);
+
if (!mime || strcmp(mime, none))
{
if (mime strcmp(mime, 7bit) == 0)
Re: [courier-users] Courier::Filter 0.16
Gordon Messmer writes: Sam Varshavchik wrote: It's always safe to append more stuff to the control file. Whether it's safe to rewrite it from scratch is something I need to think about. At the time submit calls run_filter, that's not true. ctlfile is open and the write pointer will indicate the end of the file. If a filter adds data to the control file, that pointer does not change, so anything that submit writes to the file will overwrite the records that were added by the filter. I think that ctlfile should be opened for append mode. pgpkZRb6Q29IC.pgp Description: PGP signature
RE: [courier-users] Courier::Filter 0.16
Mark Bucciarelli wrote: [taking off list, as I think I have borked my setup.] No, unfortunately this is a general issue with IO::InnerFile. See below. Julian Mehnle wrote: Try 0.11 as soon as it appears on CPAN (which should be within the next few hours). That got me futher, but I'm still having some trouble. Jan 24 08:27:48 dev courierfilter: Starting pureperlfilter Jan 24 08:28:08 dev courieresmtpd: started,ip=[::: ...] Jan 24 08:28:09 dev courierfilter: Courier::Filter::Module::ClamAVd: MIME::Parser: can't seek: Bad file descriptor at /usr/local/share/perl/5.8.4/MIME/Parser.pm line 790, GEN3 line 38. Jan 24 08:28:09 dev courierfilter: Rejected message sent from [EMAIL PROTECTED] to [EMAIL PROTECTED] through::: ... (host-...) Jan 24 08:28:09 dev courierfilter: Reason: Mail filters temporarily unavailable. Running install Courier::Filter completes cleanly in CPAN. The problem is that IO::InnerFile, which is used by MIME::Parser, does not return true on success. Old versions of MIME::Parser (up to 5.413) did not care about that. ( Detailed explanation: The MIME-tools and IO-stringy packages were initially written and maintained by the same author who now has not been maintaining his CPAN packages for over three years. MIME-tools has since been taken over by a new maintainer, and starting with version 5.414, MIME::Parser checks the result of the seek() call and thus always fails when working with IO::InnerFile. I filed a bug report on IO::InnerFile[1] a while ago, but nothing has happened so far. I am trying to resolve the issue with MIME-tools' new maintainer and IO-stringy's current maintainer. ) Now what all this essentially means is that either a version of MIME-tools/MIME::Parser = 5.413 must be installed on the system, or the installed version of IO::InnerFile must be patched manually until a fixed version is officially available. You can get the patch here[2]. References: 1. http://rt.cpan.org/NoAuth/Bug.html?id=8605 2. http://files.mehnle.net/software/perl/io-innerfile-seek-return-value.diff --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
RE: [courier-users] Courier::Filter 0.16
Mark Bucciarelli [EMAIL PROTECTED] wrote: On Tuesday 18 January 2005 15:03, Julian Mehnle wrote: * Added the ClamAVd filter module for malware filtering using a ClamAV `clamd` dameon. Looks like this modules required ClamAV::Client? I was not able to retrieve it using CPAN. Below is the tail end of the cpan install ClamAV::Client output--any pointers? Removing previously used /root/.cpan/build/ClamAV-Client-0.10 Package seems to come without Makefile.PL. (The test -f /root/.cpan/build/ClamAV-Client-0.10/Makefile.PL returned false.) Writing one on our own (setting NAME to ClamAVClient) [...] Oops, I accidentally did not include Makefile.PL in the distribution; don't know how that could have happened. Try 0.11 as soon as it appears on CPAN (which should be within the next few hours). Thanks for the hint. --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
On Mit 19.01.2005 21:17, Sam Varshavchik wrote: It's always safe to append more stuff to the control file. Whether it's safe to rewrite it from scratch is something I need to think about. Cool, thanx for your great work on courier and i'am looking, very positive, forward to have these feature in the far future ;-)) al ;-) --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Gordon Messmer wrote: Sam Varshavchik wrote: It's always safe to append more stuff to the control file. Whether it's safe to rewrite it from scratch is something I need to think about. I thought it was, too. However, when I started actually testing that theory, I found that it's not the case at the time that a courierfilter runs. If a courierfilter appends data to a control file, Courier may then add a COMCTLFILE_8BIT record. The write pointer for the ctlfile hasn't changed, so that record overwrites a portion of what data the courierfilter added. submit2.C does not flush the control file before filtering. As a workaround, to modify the control file I do something like fp = fopen(cfc-fname, r+); fseek(fp, 0, SEEK_END); fprintf(fp, %254s\n, ); before adding I or S records as needed. I reckon about 32 bytes per record is enough (possible records may be U, V, w, E, p, W, and A, in addition to 8.) --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
RE: [courier-users] Courier::Filter 0.16
Scott [EMAIL PROTECTED] wrote: Julian Mehnle wrote: Courier::Filter 0.16 has been released on CPAN[1]. The most important changes are: * Added the ClamAVd filter module for malware filtering using a ClamAV `clamd` dameon. Before I dive into this new release, does that mean that I can quit using amavis and instead use courier perl-filter to call clamd instead? Yes. Frederik Dannemare [EMAIL PROTECTED] wrote: Not that this answers your question, but you could also quit using amavis and instead head for ClamCour [...] The point of using Courier::Filter is that it is very flexible and supports arbitrary filter modules and complex filter module configurations, not just ClamAV scanning. --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Hi, On Mit 19.01.2005 12:59, Julian Mehnle wrote: The point of using Courier::Filter is that it is very flexible and supports arbitrary filter modules and complex filter module configurations, not just ClamAV scanning. But at the moment the quequefile can't still be modified by any filter, isn't it? al ;-) --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
RE: [courier-users] Courier::Filter 0.16
Alexander Lazic [EMAIL PROTECTED] wrote: On Mit 19.01.2005 12:59, Julian Mehnle wrote: The point of using Courier::Filter is that it is very flexible and supports arbitrary filter modules and complex filter module configurations, not just ClamAV scanning. But at the moment the quequefile can't still be modified by any filter, isn't it? True. But that means ClamCour and Amavis can't do it either. Courierfilters can do SMTP-time message rejection only. --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
On Mit 19.01.2005 16:49, Julian Mehnle wrote: True. But that means ClamCour and Amavis can't do it either. Courierfilters can do SMTP-time message rejection only. Yeah :-( I like courier but these is a really bad limitation, imho :-( al ;-) --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Alexander Lazic said: On Mit 19.01.2005 16:49, Julian Mehnle wrote: True. But that means ClamCour and Amavis can't do it either. Courierfilters can do SMTP-time message rejection only. Yeah :-( I like courier but these is a really bad limitation, imho :-( To be fair, very few mail servers I know of or have used have this feature. So I'd instead just call it a really cool feature of Postfix that Sam might one day have the time to add. :) I was thinking the other day that it'd be incredibly awesome if the delivery time filtering worked much like maildrop filtering, allowing for a simple scripting language to control the acceptance/modification/denial of SMTP mail. Still, Courier is pretty terrific though. Jay -- Jay Lee Network / Systems Administrator Information Technology Dept. Philadelphia Biblical University -- --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Jay Lee wrote: To be fair, very few mail servers I know of or have used have this feature. At least Postfix and Sendmail have it. It's not that exotic. It should be fairly simple to add to Courier. About all that needs to be done, AFAIK, is add a seek() to line 947 of submit2.C, or close the control files before run_filter and reopen them afterward. It doesn't look like the afx lib in use has a suitable seekp() function, so the latter may be easier than the former. --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Gordon Messmer writes: Jay Lee wrote: To be fair, very few mail servers I know of or have used have this feature. At least Postfix and Sendmail have it. It's not that exotic. It should be fairly simple to add to Courier. About all that needs to be done, AFAIK, is add a seek() to line 947 of submit2.C, or close the control files before run_filter and reopen them afterward. It doesn't look like the afx lib in use has a suitable seekp() function, so the latter may be easier than the former. No -- the problem that prevents this from working is that by that time the message's MIME structure is already parsed. If, suddenly the message's contents change, the pre-parsed offsets of various MIME entities in the message will result in a corrupted message, should Courier choose to rewrite it. pgp5cu7LYKUqj.pgp Description: PGP signature
RE: [courier-users] Courier::Filter 0.16
Sam Varshavchik [EMAIL PROTECTED] wrote: No -- the problem that prevents this from working is that by that time the message's MIME structure is already parsed. If, suddenly the message's contents change, the pre-parsed offsets of various MIME entities in the message will result in a corrupted message, should Courier choose to rewrite it. Can't Courier just determine whether the message file was modified, and if so, re-read and re-parse the message? That should avoid any performance problems for backward compatible behavior (i.e. for non-modifying filters). --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Sam Varshavchik wrote: Gordon Messmer writes: About all that needs to be done, AFAIK, is add a seek() to line 947 of submit2.C, or close the control files before run_filter and reopen them afterward. It doesn't look like the afx lib in use has a suitable seekp() function, so the latter may be easier than the former. No -- the problem that prevents this from working is that by that time the message's MIME structure is already parsed. If, suddenly the message's contents change, the pre-parsed offsets of various MIME entities in the message will result in a corrupted message, should Courier choose to rewrite it. For the control file? --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Gordon Messmer writes: Sam Varshavchik wrote: Gordon Messmer writes: About all that needs to be done, AFAIK, is add a seek() to line 947 of submit2.C, or close the control files before run_filter and reopen them afterward. It doesn't look like the afx lib in use has a suitable seekp() function, so the latter may be easier than the former. No -- the problem that prevents this from working is that by that time the message's MIME structure is already parsed. If, suddenly the message's contents change, the pre-parsed offsets of various MIME entities in the message will result in a corrupted message, should Courier choose to rewrite it. For the control file? No, the message file. pgpXQo9gv6XKW.pgp Description: PGP signature
RE: [courier-users] Courier::Filter 0.16
Can't Courier just determine whether the message file was modified, and if so, re-read and re-parse the message? That should avoid any performance problems for backward compatible behavior (i.e. for non-modifying filters). [Mitch says:] That's been kicked around before - it would seem to be a lot more efficient than resubmitting all the time, but the fact that it hasn't happened yet makes me wonder how much of a bear that option must be... If it's not that BAD, but just a low priority , maybe Sam could provide a few helpful hints to people who want to take that on - would there be an objection to including the feature if it were made? m/ --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
RE: [courier-users] Courier::Filter 0.16
Sam Varshavchik [EMAIL PROTECTED] wrote: No -- the problem that prevents this from working is that by that time the message's MIME structure is already parsed. If, suddenly the message's contents change, the pre-parsed offsets of various MIME entities in the message will result in a corrupted message, should Courier choose to rewrite it. Can't Courier just determine whether the message file was modified, and if so, re-read and re-parse the message? That should avoid any performance problems for backward compatible behavior (i.e. for non-modifying filters). [Mitch says:] Another hack I played with a while back (which may still work on current versions was to edit a PREEXISTING header - thereby preserving the structure... that is sometimes enough for people - you make sure all messages are created with a header X-My-Header: XXX and then as needed you rewrite it to X-My-Header: No_ or X-My-Header: Yes - note they are the same length, so it doesn't screw up the structure byte offsets... Just in case that helps... m/ --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Mitch (WebCob) writes: Can't Courier just determine whether the message file was modified, and if so, re-read and re-parse the message? That should avoid any performance problems for backward compatible behavior (i.e. for non-modifying filters). [Mitch says:] That's been kicked around before - it would seem to be a lot more efficient than resubmitting all the time, but the fact that it hasn't happened yet makes me wonder how much of a bear that option must be... If it's not that BAD, but just a low priority , maybe Sam could provide a few helpful hints to people who want to take that on - would there be an objection to including the feature if it were made? No -- no objections. pgpw4Mi9Z9FDx.pgp Description: PGP signature
Re: [courier-users] Courier::Filter 0.16
Sam Varshavchik wrote: Gordon Messmer writes: Sam Varshavchik wrote: No -- the problem that prevents this from working is that by that time the message's MIME structure is already parsed. If, suddenly the message's contents change, the pre-parsed offsets of various MIME entities in the message will result in a corrupted message, should Courier choose to rewrite it. For the control file? No, the message file. Yeah, but I was talking about the changes required to allow filters to modify the control file, which also allows for interesting possibilities. --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Gordon Messmer writes: Sam Varshavchik wrote: Gordon Messmer writes: Sam Varshavchik wrote: No -- the problem that prevents this from working is that by that time the message's MIME structure is already parsed. If, suddenly the message's contents change, the pre-parsed offsets of various MIME entities in the message will result in a corrupted message, should Courier choose to rewrite it. For the control file? No, the message file. Yeah, but I was talking about the changes required to allow filters to modify the control file, which also allows for interesting possibilities. It's always safe to append more stuff to the control file. Whether it's safe to rewrite it from scratch is something I need to think about. pgpz3Tr6l5xrk.pgp Description: PGP signature
Re: [courier-users] Courier::Filter 0.16
Sam Varshavchik wrote: It's always safe to append more stuff to the control file. Whether it's safe to rewrite it from scratch is something I need to think about. I thought it was, too. However, when I started actually testing that theory, I found that it's not the case at the time that a courierfilter runs. If a courierfilter appends data to a control file, Courier may then add a COMCTLFILE_8BIT record. The write pointer for the ctlfile hasn't changed, so that record overwrites a portion of what data the courierfilter added. --- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag--drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
Julian Mehnle wrote: Courier::Filter 0.16 has been released on CPAN[1]. The most important changes are: * Added the ClamAVd filter module for malware filtering using a ClamAV `clamd` dameon. Before I dive into this new release, does that mean that I can quit using amavis and instead use courier perl-filter to call clamd instead? -Scott --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
Re: [courier-users] Courier::Filter 0.16
On Wednesday 19 January 2005 03:56, Scott wrote: Julian Mehnle wrote: Courier::Filter 0.16 has been released on CPAN[1]. The most important changes are: * Added the ClamAVd filter module for malware filtering using a ClamAV `clamd` dameon. Before I dive into this new release, does that mean that I can quit using amavis and instead use courier perl-filter to call clamd instead? Not that this answers your question, but you could also quit using amavis and instead head for ClamCour http://www.becrux.com/index.php?page=projectsname=clamcour. I discovered ClamCour myself only a couple of weeks ago and I have been very happy with it as a replacement for amavisd-new. On my Pentium 166MHz with only 32 MB RAM it really makes a difference as ClamCour is more light on CPU and memory usage. -- Frederik Dannemare | mailto:[EMAIL PROTECTED] http://qa.debian.org/developer.php?login=Frederik+Dannemare http://frederik.dannemare.net | http://www.linuxworlddomination.dk --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt ___ courier-users mailing list courier-users@lists.sourceforge.net Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
