From peterv at propagandism.org Mon Dec 13 22:07:45 2004 From: peterv at propagandism.org (Peter Van der Beken) Date: Mon Dec 13 22:07:48 2004 Subject: [Expat-discuss] XML_ParserCreateNS Message-ID: <41BE04A1.2030104@propagandism.org> Quick question about XML_ParserCreateNS, expat.h has this as description: /* Constructs a new parser and namespace processor. Element type names and attribute names that belong to a namespace will be expanded; unprefixed attribute names are never expanded; unprefixed element type names are expanded only if there is a default namespace. The expanded name is the concatenation of the namespace URI, the namespace separator character, and the local part of the name. If the namespace separator is '\0' then the namespace URI and the local part will be concatenated without any separator. When a namespace is not declared, the name and prefix will be passed through without expansion. */ The last sentence confuses me, is that only in the case of the namespace separator being '\0'? I definitely got XML_ERROR_UNBOUND_PREFIX for an undeclared prefix and I don't see where that's special-cased for '\0'. Thanks, Peter From karl at waclawek.net Tue Dec 14 02:18:06 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 02:18:05 2004 Subject: [Expat-discuss] XML_ParserCreateNS In-Reply-To: <41BE04A1.2030104@propagandism.org> References: <41BE04A1.2030104@propagandism.org> Message-ID: <41BE3F4E.3060209@waclawek.net> Peter Van der Beken wrote: > Quick question about XML_ParserCreateNS, expat.h has this as description: > > /* Constructs a new parser and namespace processor. Element type > names and attribute names that belong to a namespace will be > expanded; unprefixed attribute names are never expanded; unprefixed > element type names are expanded only if there is a default > namespace. The expanded name is the concatenation of the namespace > URI, the namespace separator character, and the local part of the > name. If the namespace separator is '\0' then the namespace URI > and the local part will be concatenated without any separator. > When a namespace is not declared, the name and prefix will be > passed through without expansion. > */ > > The last sentence confuses me, is that only in the case of the namespace > separator being '\0'? I definitely got XML_ERROR_UNBOUND_PREFIX for an > undeclared prefix and I don't see where that's special-cased for '\0'. About '\0' as namespace separator: Good observation. I looked in xmlparse.c for all occurrences of namespaceSeparator, and found a somewhat confusing picture. It seems that at some point there was an intention to omit the separator if it is '\0', but it has not been followed through consistently, especially not for the element name in a non-empty tag. However, later a feature called ns_triplets was added, which adds the prefix to the end of the name - again separated with the namespaceSeparator. In my opinion, having all three concatenated without any spearator makes little sense. I am inclined to change the code to simply use whatever namespace separator is supplied, even if it is '\0'. Can you think of any use case for having the separator omitted? Fred, if you read this - do you remember why this feature was provided? About the last sentence: > When a namespace is not declared, the name and prefix will be > passed through without expansion. This makes no sense to me either - when namespace processing is turned on - as the parser would then not conform to the namespaces specification, if I remember correctly. So, I suggest two steps: - modify namespace separator code as above and adjust comments/docs - remove the sentence above from the comments/docs Karl From weigelt at metux.de Tue Dec 14 02:53:53 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 02:53:56 2004 Subject: [Expat-discuss] Expat 1.95.8 fails on XML with newline In-Reply-To: <001b01c4d163$583c70a0$9e539696@torstar.net> References: <41A33F33.9040103@hypershell.com> <001b01c4d163$583c70a0$9e539696@torstar.net> Message-ID: <20041214015353.GA27907@nibiru.borg.metux.de> * Karl Waclawek wrote: > AFAIK, starting an XML document with whitespace is an error, > i.e. your document is not well-formed. Nevertheless its bad behaviour of expat and shall be fixed. BTW: is my install-fix ($DESTDIR prefix) now included in the latest release ? cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From weigelt at metux.de Tue Dec 14 03:10:07 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 03:10:11 2004 Subject: [Expat-discuss] Create XML file In-Reply-To: <002c01c4bbe6$844bd150$4900a8c0@gmitech> References: <002c01c4bbe6$844bd150$4900a8c0@gmitech> Message-ID: <20041214021006.GA28636@nibiru.borg.metux.de> * Micky MeNeZeS wrote: > Hi all, > > Is it possible to create XML using expat ? > I mean can we write into xml file which is blank or already > having some structured data. Expat is only a parser -> it only *reads* xml. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From weigelt at metux.de Tue Dec 14 03:21:12 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 03:21:16 2004 Subject: [Expat-discuss] [patch] fixing broken install rules In-Reply-To: <20040724184608.GA22105@nibiru.metux.de> References: <20040724184608.GA22105@nibiru.metux.de> Message-ID: <20041214022112.GB28636@nibiru.borg.metux.de> * Enrico Weigelt wrote: Hi folks, > I've reported a bug in the makefile which lead to corrupt > installations when specifying another install-root > (make DESTDIR=foo install) and sent a patch to fix it. It's now some month's ago since I've posted this patch and really no response to it. Can I assume that repairing fundamental bugs and allowing packaging of expat is not in of interest here ? Do I need to maintain my own branch ? Here's the patch again. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- -------------- next part -------------- diff -ruN expat-1.95.8.orig/Makefile.in expat-1.95.8/Makefile.in --- expat-1.95.8.orig/Makefile.in 2004-12-03 16:48:45.000000000 +0100 +++ expat-1.95.8/Makefile.in 2004-12-03 16:50:03.000000000 +0100 @@ -73,22 +73,22 @@ tests/runtests install: xmlwf/xmlwf installlib - $(mkinstalldirs) $(bindir) $(man1dir) - $(LIBTOOL) --mode=install $(INSTALL_PROGRAM) xmlwf/xmlwf $(bindir)/xmlwf - $(INSTALL_DATA) $(MANFILE) $(man1dir) + $(mkinstalldirs) $(DESTDIR)$(bindir) $(DESTDIR)$(man1dir) + $(LIBTOOL) --mode=install $(INSTALL_PROGRAM) xmlwf/xmlwf $(DESTDIR)$(bindir)/xmlwf + $(INSTALL_DATA) $(MANFILE) $(DESTDIR)$(man1dir) installlib: $(LIBRARY) $(APIHEADER) - $(mkinstalldirs) $(libdir) $(includedir) - $(LIBTOOL) --mode=install $(INSTALL) $(LIBRARY) $(libdir)/$(LIBRARY) - $(INSTALL_DATA) $(APIHEADER) $(includedir) + $(mkinstalldirs) $(DESTDIR)$(libdir) $(DESTDIR)$(includedir) + $(LIBTOOL) --mode=install $(INSTALL) $(LIBRARY) $(DESTDIR)$(libdir)/$(LIBRARY) + $(INSTALL_DATA) $(APIHEADER) $(DESTDIR)$(includedir) uninstall: uninstalllib - $(LIBTOOL) --mode=uninstall rm -f $(bindir)/xmlwf - rm -f $(man1dir)/xmlwf.1 + $(LIBTOOL) --mode=uninstall rm -f $(DESTDIR)$(bindir)/xmlwf + rm -f $(DESTDIR)$(man1dir)/xmlwf.1 uninstalllib: - $(LIBTOOL) --mode=uninstall rm -f $(libdir)/$(LIBRARY) - rm -f $(includedir)/$(APIHEADER) + $(LIBTOOL) --mode=uninstall rm -f $(DESTDIR)$(libdir)/$(LIBRARY) + rm -f $(DESTDIR)$(includedir)/$(APIHEADER) # for VPATH builds (invoked by configure) mkdir-init: From weigelt at metux.de Tue Dec 14 03:26:44 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 03:26:46 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <87brdy4uhj.fsf@dod.no> References: <001301c4b451$665825a0$0200a8c0@karlglen188> <87k6tpjhhu.fsf@dod.no> <001801c4b48a$3fbffdd0$097ca8c0@karlglen188> <87brf1j9dp.fsf@dod.no> <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> Message-ID: <20041214022643.GC28636@nibiru.borg.metux.de> Probably I missed something, but if *the same* API function exist with different behaviour (ie one time seeing char* as 8bit-string and another time as 16bit string) is a fundamental misdesign. If an package should provide a *reliable* quality, such things are an absolute *NO* *NO* ! What stands against providing either *one* API (8bit *XOR* 16bit) and putting write frontend of the other variant or provide two different sets of functions (ie xml8_xxx, xml16_xxx) ? cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From weigelt at metux.de Tue Dec 14 03:30:22 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 03:30:30 2004 Subject: [Expat-discuss] canonical download URLs ? Message-ID: <20041214023021.GD28636@nibiru.borg.metux.de> Hi folks, I'm using an completely automated buildfarm for building my distros, but still I have to download expat manually, which really really suxx. Is there an canonical uri scheme for downloading sources of an specific release ? cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Tue Dec 14 03:31:47 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 03:31:45 2004 Subject: [Expat-discuss] Expat 1.95.8 fails on XML with newline In-Reply-To: <20041214015353.GA27907@nibiru.borg.metux.de> References: <41A33F33.9040103@hypershell.com> <001b01c4d163$583c70a0$9e539696@torstar.net> <20041214015353.GA27907@nibiru.borg.metux.de> Message-ID: <41BE5093.9040203@waclawek.net> Enrico Weigelt wrote: > * Karl Waclawek wrote: > >>AFAIK, starting an XML document with whitespace is an error, >>i.e. your document is not well-formed. > > > Nevertheless its bad behaviour of expat and shall be fixed. For saying this, the wrath of the XML gods will come upon you. ;-) > > BTW: is my install-fix ($DESTDIR prefix) now included in the > latest release ? I leave the answer to Fred, as he deals with build issues better than me. Karl From karl at waclawek.net Tue Dec 14 03:39:46 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 03:39:45 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <20041214022643.GC28636@nibiru.borg.metux.de> References: <001301c4b451$665825a0$0200a8c0@karlglen188> <87k6tpjhhu.fsf@dod.no> <001801c4b48a$3fbffdd0$097ca8c0@karlglen188> <87brf1j9dp.fsf@dod.no> <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> Message-ID: <41BE5272.3030508@waclawek.net> Enrico Weigelt wrote: > Probably I missed something, but if *the same* API function exist > with different behaviour (ie one time seeing char* as 8bit-string > and another time as 16bit string) is a fundamental misdesign. They are not the same API functions, as the signatures are different due to the different type of XML_Char. This overloading of names is quite common. Secondly, they don't even co-exist in the same library instance, and the libraries for the two types of XML_Char have different names. > If an package should provide a *reliable* quality, such things > are an absolute *NO* *NO* ! > > What stands against providing either *one* API (8bit *XOR* 16bit) > and putting write frontend of the other variant or provide two > different sets of functions (ie xml8_xxx, xml16_xxx) ? For some situations it would be nice to have both sets of functions in the same library instance. This would mean reworking a lot of Expat. Since Expat is a volunteer drive effort, why could it not be you who contributes such a re-design? Karl From weigelt at metux.de Tue Dec 14 04:08:17 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 04:08:21 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <41BE5093.9040203@waclawek.net> References: <41A33F33.9040103@hypershell.com> <001b01c4d163$583c70a0$9e539696@torstar.net> <20041214015353.GA27907@nibiru.borg.metux.de> <41BE5093.9040203@waclawek.net> Message-ID: <20041214030817.GE28636@nibiru.borg.metux.de> * Karl Waclawek wrote: > >Nevertheless its bad behaviour of expat and shall be fixed. > > For saying this, the wrath of the XML gods will come upon you. ;-) I really dont care about such fundamentalists. XML is a text format, and those normally have dozens over newlines. I dont see where kindly accepting some additional newlines makes any trouble. Expat is per definition an non-validating parser and thus is not intendet for syntax checking. The only point where *I* am as fanatic as GWB is the point that Interfaces have to be *deterministic* and *constant*. Once interfaces are established, they dont have to be changed anymore, not even in thousand years, so that someone seeing this interface can really rely on it. If some wants to change an interface, he has to establish a new one and probably throw the old away if its not needed anylonger. Thats one of the primary rules for ensuring software quality. > >BTW: is my install-fix ($DESTDIR prefix) now included in the > >latest release ? > > I leave the answer to Fred, as he deals with build issues > better than me. Well, I had a short look into the recent code and can answer it by myself: the make rules are still broken - sysroot-installs are still impossible yet. Crosscompiling seems to work on equal host and build. But on certain sites it refuses to build shared libraries. Probably its the same as on apache: configure is too stupid to detect a well installed gnu ld as such, because it was created with broken autoconf. In fact autoconf isnt really broken - its an misdesign, but that's another story ... Also its not possible to disable static libraries and commands. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Tue Dec 14 04:14:02 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 04:13:59 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <20041214030817.GE28636@nibiru.borg.metux.de> References: <41A33F33.9040103@hypershell.com> <001b01c4d163$583c70a0$9e539696@torstar.net> <20041214015353.GA27907@nibiru.borg.metux.de> <41BE5093.9040203@waclawek.net> <20041214030817.GE28636@nibiru.borg.metux.de> Message-ID: <41BE5A7A.30908@waclawek.net> Enrico Weigelt wrote: > * Karl Waclawek wrote: > > > >>>Nevertheless its bad behaviour of expat and shall be fixed. >> >>For saying this, the wrath of the XML gods will come upon you. ;-) > > I really dont care about such fundamentalists. > XML is a text format, and those normally have dozens over newlines. > I dont see where kindly accepting some additional newlines makes > any trouble. Expat is per definition an non-validating parser and > thus is not intendet for syntax checking. This has nothing to do with validation. Well-formednes is fundamental, even an XML document without any DTD must conform to these rules. There is simply no way around it. You will find no major XML parser that will accept such a document. Karl From weigelt at metux.de Tue Dec 14 04:17:47 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 04:17:50 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <41BE5272.3030508@waclawek.net> References: <001801c4b48a$3fbffdd0$097ca8c0@karlglen188> <87brf1j9dp.fsf@dod.no> <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> <41BE5272.3030508@waclawek.net> Message-ID: <20041214031746.GF28636@nibiru.borg.metux.de> * Karl Waclawek wrote: > Enrico Weigelt wrote: > >Probably I missed something, but if *the same* API function exist > >with different behaviour (ie one time seeing char* as 8bit-string > >and another time as 16bit string) is a fundamental misdesign. > > They are not the same API functions, as the signatures are > different due to the different type of XML_Char. > This overloading of names is quite common. Is the linker (ld.so) able to handle this overloading properly ? > Secondly, they don't even co-exist in the same library instance, > and the libraries for the two types of XML_Char have different names. Well, thats how it has to be. But then where's the problem ? > For some situations it would be nice to have both sets of > functions in the same library instance. This would mean reworking > a lot of Expat. Since Expat is a volunteer drive effort, why could > it not be you who contributes such a re-design? No problem, if my work wouldn't fall into the trashcan of history due missing interest of other people ... At the moment I'm not very hopeful at this point, since the folks here didn't get a single trivial but important patch in for a couple of month, neither gave any response to my posting. It really suxx if I have to fix trivial things again and again and people even dont recognize my contributions. This makes distro maintenance an ugly job. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Tue Dec 14 04:49:24 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 04:49:23 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <20041214031746.GF28636@nibiru.borg.metux.de> References: <001801c4b48a$3fbffdd0$097ca8c0@karlglen188> <87brf1j9dp.fsf@dod.no> <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> <41BE5272.3030508@waclawek.net> <20041214031746.GF28636@nibiru.borg.metux.de> Message-ID: <41BE62C4.6060802@waclawek.net> Enrico Weigelt wrote: > * Karl Waclawek wrote: > > > >>Enrico Weigelt wrote: >> >>>Probably I missed something, but if *the same* API function exist >>>with different behaviour (ie one time seeing char* as 8bit-string >>>and another time as 16bit string) is a fundamental misdesign. >> >>They are not the same API functions, as the signatures are >>different due to the different type of XML_Char. >>This overloading of names is quite common. > > > Is the linker (ld.so) able to handle this overloading properly ? I don't know about Linux - but there was a thread about it a while ago. In Windows - where I usually work - this is no problem. > > > >>Secondly, they don't even co-exist in the same library instance, >>and the libraries for the two types of XML_Char have different names. > > Well, thats how it has to be. > But then where's the problem ? Don't understand the question. > > > >>For some situations it would be nice to have both sets of >>functions in the same library instance. This would mean reworking >>a lot of Expat. Since Expat is a volunteer drive effort, why could >>it not be you who contributes such a re-design? > > > No problem, if my work wouldn't fall into the trashcan of history > due missing interest of other people ... Can't really help you with build issues - as my Unix expertise in this area is not very good. I don't really want to commit a patch without having it reviewed by our Unix build experts. It is true, some of us have been too busy elsewhere, and haven't found the time to deal with outstanding build issues. > At the moment I'm not very hopeful at this point, since the folks > here didn't get a single trivial but important patch in for a > couple of month, neither gave any response to my posting. > > It really suxx if I have to fix trivial things again and again > and people even dont recognize my contributions. This makes > distro maintenance an ugly job. Let's hope you will get more response this time. Karl From weigelt at metux.de Tue Dec 14 04:58:52 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 04:58:54 2004 Subject: [Expat-discuss] RFC: unitool -> universal toolchain instead of autoshit Message-ID: <20041214035852.GG28636@nibiru.borg.metux.de> Hi folks, I've started to write down some concepts for an universal toolchain command as an platform-independent frontend for toolchain and installation commands. It is supports crosscompiling, sysroot-building, file installation + packaging, library management and system property database as fundamental concepts. No more "checking whether 1+1 ..." anylonger. Instead create a *strict* set of commands which can be really assumed to *work*, and of course *exactly the same on every platform. When using unitool, about 95% of the configure stuff is obsolete. We dont need to find out how to call several tools on a certain platform, now we definitivly *know* it. All platform or toolchain specific stuff is hidden behind the individual implementation of unitool. The wiki page can be found at: http://nibiru.borg.metux.de:7000/wiki/index.php/Univeral_Toolchain The maillist can be reached via unitool-dev@metux.de, subscriptions via majordomo@metux.de cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Tue Dec 14 05:10:43 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 05:10:40 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <41BE62C4.6060802@waclawek.net> References: <001801c4b48a$3fbffdd0$097ca8c0@karlglen188> <87brf1j9dp.fsf@dod.no> <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> <41BE5272.3030508@waclawek.net> <20041214031746.GF28636@nibiru.borg.metux.de> <41BE62C4.6060802@waclawek.net> Message-ID: <41BE67C3.3000907@waclawek.net> Karl Waclawek wrote: > > > Enrico Weigelt wrote: > >> * Karl Waclawek wrote: >> >> >> >>> Enrico Weigelt wrote: >>> >>>> Probably I missed something, but if *the same* API function exist >>>> with different behaviour (ie one time seeing char* as 8bit-string >>>> and another time as 16bit string) is a fundamental misdesign. >>> >>> >>> They are not the same API functions, as the signatures are >>> different due to the different type of XML_Char. >>> This overloading of names is quite common. >> >> >> >> Is the linker (ld.so) able to handle this overloading properly ? > > > I don't know about Linux - but there was a thread about it > a while ago. In Windows - where I usually work - this is no problem. I think I found the Linux discussion: Check the October 2004 archives and look for the thread "Conflict with two expat shared libraries". Karl From weigelt at metux.de Tue Dec 14 05:10:52 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 05:10:55 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <41BE62C4.6060802@waclawek.net> References: <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> <41BE5272.3030508@waclawek.net> <20041214031746.GF28636@nibiru.borg.metux.de> <41BE62C4.6060802@waclawek.net> Message-ID: <20041214041052.GH28636@nibiru.borg.metux.de> * Karl Waclawek wrote: Hi, > >Is the linker (ld.so) able to handle this overloading properly ? > > I don't know about Linux - but there was a thread about it > a while ago. In Windows - where I usually work - this is no problem. Well, on win32 its no problem, if you manually load the dll and find the functions one by one. This would also be possibly on gnu/linux with a prepared library loader (I remember in old days there was another .so loader library than libdl going around, which AFAIK did not publish all symbols as global). But thats really not the point! We also want to let the dynamic linker do evrything completely alone. And then we *need* different symbols for different things. AFAIK dll symbol tables do not support any kind of typing information and thus no overloading. Well, my last win32 works are several years ago, so it probably changed ... > >Well, thats how it has to be. > >But then where's the problem ? > > Don't understand the question. If both library variants work - as you said - by overloading (no idea how this should go w/o typeinfo), then I dont understand what's your problem. Obviously its the point that it does _not_ work. > >No problem, if my work wouldn't fall into the trashcan of history > >due missing interest of other people ... > > Can't really help you with build issues - as my Unix expertise > in this area is not very good. I don't really want to commit > a patch without having it reviewed by our Unix build experts. My dear, I'm now building distros for many years (created an automatic buildsystem which creates embedded firmware for almost any kind of linux-based device) and I've tested my patch a thousand times. Can't you just trust in me that I'm able to add an $(DESTDIR) variable reference in front of every installation location for 'cp', 'install' and similar commands without hurting anything else ? > It is true, some of us have been too busy elsewhere, and haven't > found the time to deal with outstanding build issues. So if your unix experts are too busy, why dont you just let me do this job ? I'm now working about 10 years as gnu/linux developer, so I feel competent enough :) cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From weigelt at metux.de Tue Dec 14 05:13:21 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 05:13:24 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <41BE67C3.3000907@waclawek.net> References: <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> <41BE5272.3030508@waclawek.net> <20041214031746.GF28636@nibiru.borg.metux.de> <41BE62C4.6060802@waclawek.net> <41BE67C3.3000907@waclawek.net> Message-ID: <20041214041321.GI28636@nibiru.borg.metux.de> * Karl Waclawek wrote: > I think I found the Linux discussion: Check the October 2004 archives > and look for the thread "Conflict with two expat shared libraries". yes, these are the postings i'm replying ... (sometimes it takes a while until I find the time to read all the thousands over thousands lists mails ;-)) cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Tue Dec 14 05:20:40 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 05:20:38 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <20041214041052.GH28636@nibiru.borg.metux.de> References: <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> <41BE5272.3030508@waclawek.net> <20041214031746.GF28636@nibiru.borg.metux.de> <41BE62C4.6060802@waclawek.net> <20041214041052.GH28636@nibiru.borg.metux.de> Message-ID: <41BE6A18.50503@waclawek.net> Enrico Weigelt wrote: > Well, on win32 its no problem, if you manually load the dll and > find the functions one by one. This would also be possibly on gnu/linux > with a prepared library loader (I remember in old days there was another > .so loader library than libdl going around, which AFAIK did not publish > all symbols as global). > > But thats really not the point! We also want to let the dynamic linker > do evrything completely alone. And then we *need* different symbols > for different things. AFAIK dll symbol tables do not support any kind > of typing information and thus no overloading. > > Well, my last win32 works are several years ago, so it probably changed ... > > >>>Well, thats how it has to be. >>>But then where's the problem ? >> >>Don't understand the question. > > If both library variants work - as you said - by overloading (no idea > how this should go w/o typeinfo), then I dont understand what's your > problem. I don't have a problem. > > Obviously its the point that it does _not_ work. > > > > >>>No problem, if my work wouldn't fall into the trashcan of history >>>due missing interest of other people ... >> >>Can't really help you with build issues - as my Unix expertise >>in this area is not very good. I don't really want to commit >>a patch without having it reviewed by our Unix build experts. > > My dear, I'm now building distros for many years (created an automatic > buildsystem which creates embedded firmware for almost any kind of > linux-based device) and I've tested my patch a thousand times. > > Can't you just trust in me that I'm able to add an $(DESTDIR) variable > reference in front of every installation location for 'cp', 'install' > and similar commands without hurting anything else ? I didn't even try to look at it closely, as I am not usually dealing with build issues. And no, I don't just trust you because I simply don't have the time to go out and check your credentials or evaluate your contributions which are of a type that I usually don't look at. Having said that, I can understand your impatience, but as volunteers we do not always have a consistent amount of time available. > > > >>It is true, some of us have been too busy elsewhere, and haven't >>found the time to deal with outstanding build issues. > > So if your unix experts are too busy, why dont you just let me do > this job ? > I'm now working about 10 years as gnu/linux developer, so I feel > competent enough :) Well, I am not the project administrator, but I am sure that Fred would welcome any competent help, busy as he is. He is cc'ed on this, so unless he really is not available (he could be on vacation or travelling), he will likely respond to you. Karl From weigelt at metux.de Tue Dec 14 06:05:55 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 06:05:58 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <41BE6A18.50503@waclawek.net> References: <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> <41BE5272.3030508@waclawek.net> <20041214031746.GF28636@nibiru.borg.metux.de> <41BE62C4.6060802@waclawek.net> <20041214041052.GH28636@nibiru.borg.metux.de> <41BE6A18.50503@waclawek.net> Message-ID: <20041214050555.GD22351@nibiru.borg.metux.de> * Karl Waclawek wrote: > I didn't even try to look at it closely, as I am not usually dealing > with build issues. And no, I don't just trust you because I simply Okay, then *who* is responsible for that ? > Having said that, I can understand your impatience, but as volunteers > we do not always have a consistent amount of time available. I'm also volunteer, and I'm willing to maintain the build stuff if the current maintainer(s) dont have enough time for that. > >So if your unix experts are too busy, why dont you just let me do > >this job ? > >I'm now working about 10 years as gnu/linux developer, so I feel > >competent enough :) > > Well, I am not the project administrator, but I am sure > that Fred would welcome any competent help, busy as he is. > He is cc'ed on this, so unless he really is not available > (he could be on vacation or travelling), he will likely > respond to you. Okay. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From gstein at lyra.org Tue Dec 14 09:35:54 2004 From: gstein at lyra.org (Greg Stein) Date: Tue Dec 14 09:32:18 2004 Subject: [Expat-discuss] RFC: unitool -> universal toolchain instead of autoshit In-Reply-To: <20041214035852.GG28636@nibiru.borg.metux.de> References: <20041214035852.GG28636@nibiru.borg.metux.de> Message-ID: <20041214083554.GC5913@lyra.org> The autoconf/libtool scripts work just fine for Expat. We dropped automake a long time (the worst of the bunch). I'm not sure how a switch would benefit Expat or its users. Cheers, -g On Tue, Dec 14, 2004 at 04:58:52AM +0100, Enrico Weigelt wrote: > > Hi folks, > > > I've started to write down some concepts for an universal toolchain > command as an platform-independent frontend for toolchain and > installation commands. It is supports crosscompiling, sysroot-building, > file installation + packaging, library management and system property > database as fundamental concepts. > > No more "checking whether 1+1 ..." anylonger. > Instead create a *strict* set of commands which can be really > assumed to *work*, and of course *exactly the same on every platform. > > When using unitool, about 95% of the configure stuff is obsolete. > We dont need to find out how to call several tools on a certain platform, > now we definitivly *know* it. All platform or toolchain specific stuff > is hidden behind the individual implementation of unitool. > > > The wiki page can be found at: > > http://nibiru.borg.metux.de:7000/wiki/index.php/Univeral_Toolchain > > The maillist can be reached via unitool-dev@metux.de, subscriptions > via majordomo@metux.de > > > cu > -- > --------------------------------------------------------------------- > Enrico Weigelt == metux IT service > > phone: +49 36207 519931 www: http://www.metux.de/ > fax: +49 36207 519932 email: contact@metux.de > cellphone: +49 174 7066481 > --------------------------------------------------------------------- > -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- > --------------------------------------------------------------------- > _______________________________________________ > Expat-discuss mailing list > Expat-discuss@libexpat.org > http://mail.libexpat.org/mailman/listinfo/expat-discuss -- -- Greg Stein, http://www.lyra.org/ From agandfr at yahoo.com Tue Dec 14 11:08:42 2004 From: agandfr at yahoo.com (Dean Agan) Date: Tue Dec 14 11:08:45 2004 Subject: [Expat-discuss] Maximum file size that can be parsed Message-ID: <20041214100842.11089.qmail@web52805.mail.yahoo.com> Hi! What's the maximum size that expat can parse? Can expat parse 1mb xml files? Also, I have lots of comments in my xml file(Just to make me remember what that inventory info is for). Can I set a size limit for expat? --------------------------------- Do you Yahoo!? Jazz up your holiday email with celebrity designs. Learn more. From rolf at pointsman.de Tue Dec 14 12:08:07 2004 From: rolf at pointsman.de (rolf@pointsman.de) Date: Tue Dec 14 12:03:33 2004 Subject: [Expat-discuss] Expat 1.95.8 fails on XML with newline In-Reply-To: <20041214015353.GA27907@nibiru.borg.metux.de> Message-ID: <20041214110810.241C580056@pointsman.pointsman.de> On 14 Dec, Enrico Weigelt wrote: > * Karl Waclawek wrote: >> AFAIK, starting an XML document with whitespace is an error, >> i.e. your document is not well-formed. > > Nevertheless its bad behaviour of expat and shall be fixed. Pardon? If an XML document starts with an XML declaration, then no space in front of the XML declartion is allowed. See the production 1 and 22 of the XML recommendation. That are the rules, no room for discussion; expat simply does the right thing, there's nothing to fix. If you absolutely don't like this, then don't use XML or go and pester the w3c XML group so that they change that in a new XML recommendation. Additionally, you're free to strip all white space prior to the XML declartion before you feed the data into the parser (but if you exchange your XML data with others don't expect that they also do this; you very probably will get complains). But the only good option you have is, to fix your documents. rolf From rolf at pointsman.de Tue Dec 14 12:26:55 2004 From: rolf at pointsman.de (rolf@pointsman.de) Date: Tue Dec 14 13:06:11 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] Message-ID: <20041214121047.09442801F7@pointsman.pointsman.de> On 14 Dec, Enrico Weigelt wrote: > * Karl Waclawek wrote: > > >> >Nevertheless its bad behaviour of expat and shall be fixed. >> >> For saying this, the wrath of the XML gods will come upon you. ;-) > I really dont care about such fundamentalists. > XML is a text format, and those normally have dozens over newlines. > I dont see where kindly accepting some additional newlines makes > any trouble. Expat is per definition an non-validating parser and > thus is not intendet for syntax checking. This is not about fundamentalism, but about playing under the rules. For the current matter, the rules are very clear. There are even a reason for that rule (to simplify a bit the process of autodetection of the document encoding). This is also no matter of validating versus non-validating parser. The rule in discussion is valid also for non-validating parsers. Sure, there is no absolute *must*, that the rules, especially this one, are exactly the way, they are. (And I'm not very happy about every decision of the xml wg.) I only say, that this mailing list is the wrong place to discuss this. Write to w3c XML wg (good luck!). rolf From rolf at pointsman.de Tue Dec 14 13:05:12 2004 From: rolf at pointsman.de (rolf@pointsman.de) Date: Tue Dec 14 13:06:31 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries Message-ID: <20041214121107.CD863801F7@pointsman.pointsman.de> On 13 Dec, Karl Waclawek wrote: > Enrico Weigelt wrote: >> My dear, I'm now building distros for many years (created an automatic >> buildsystem which creates embedded firmware for almost any kind of >> linux-based device) and I've tested my patch a thousand times. >> >> Can't you just trust in me that I'm able to add an $(DESTDIR) variable >> reference in front of every installation location for 'cp', 'install' >> and similar commands without hurting anything else ? > > I didn't even try to look at it closely, as I am not usually dealing > with build issues. And no, I don't just trust you because I simply > don't have the time to go out and check your credentials or > evaluate your contributions which are of a type that I usually don't > look at. > > Having said that, I can understand your impatience, but as volunteers > we do not always have a consistent amount of time available. >> >> >> >>>It is true, some of us have been too busy elsewhere, and haven't >>>found the time to deal with outstanding build issues. >> >> So if your unix experts are too busy, why dont you just let me do >> this job ? >> I'm now working about 10 years as gnu/linux developer, so I feel >> competent enough :) > > Well, I am not the project administrator, but I am sure > that Fred would welcome any competent help, busy as he is. > He is cc'ed on this, so unless he really is not available > (he could be on vacation or travelling), he will likely > respond to you. That's a question of politics: How open are the expat developing group. I'd say (but that's only my personal opinion): be liberal in to whom you give write access to the repository. If he does good, everone wins. If he mess up HEAD or change the API without consens with the other developers, revert to prior state (that's, what cvs is for) and write a mail with a few strong words to him. If he do it again, revert again and kick him from the developers list - nothing bad happend, beside some clean up trouble. (Well, it has come over me at some moments in the past, to volunteer as expat developer - you know, I bank on expat and it would simply feel better to have some 'official' influence. So far, I've withstood that temptation.) rolf From karl at waclawek.net Tue Dec 14 17:15:31 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 17:15:38 2004 Subject: [Expat-discuss] Maximum file size that can be parsed In-Reply-To: <20041214100842.11089.qmail@web52805.mail.yahoo.com> References: <20041214100842.11089.qmail@web52805.mail.yahoo.com> Message-ID: <41BF11A3.6010902@waclawek.net> Dean Agan wrote: > Hi! > What's the maximum size that expat can parse? Can expat parse 1mb xml files? Also, I have lots of comments in my xml file(Just to make me remember what that inventory info is for). Can I set a size limit for expat? Expat works by passing it buffers one after the other. It has no notion of a file or file size, only the buffer size is limited. So, just keep feeding it buffers, and it will happily parse them ad infinitum. Karl From karl at waclawek.net Tue Dec 14 17:20:24 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 17:20:27 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <20041214121107.CD863801F7@pointsman.pointsman.de> References: <20041214121107.CD863801F7@pointsman.pointsman.de> Message-ID: <41BF12C8.7040303@waclawek.net> rolf@pointsman.de wrote: > That's a question of politics: How open are the expat developing > group. I'd say (but that's only my personal opinion): be liberal in to > whom you give write access to the repository. If he does good, everone > wins. If he mess up HEAD or change the API without consens with the > other developers, revert to prior state (that's, what cvs is for) and > write a mail with a few strong words to him. If he do it again, revert > again and kick him from the developers list - nothing bad happend, > beside some clean up trouble. Well, considering your point, what I could do is check in Enrico's patch and leave it to the build experts to reject it if necessary. > > (Well, it has come over me at some moments in the past, to volunteer > as expat developer - you know, I bank on expat and it would simply > feel better to have some 'official' influence. So far, I've withstood > that temptation.) You are a strong man, Rolf! Karl From weigelt at metux.de Tue Dec 14 17:44:23 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 17:44:26 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <20041214112658.D65CE7FA5F@pointsman.pointsman.de> References: <20041214030817.GE28636@nibiru.borg.metux.de> <20041214112658.D65CE7FA5F@pointsman.pointsman.de> Message-ID: <20041214164422.GA13268@nibiru.borg.metux.de> * rolf@pointsman.de wrote: > > I really dont care about such fundamentalists. > > XML is a text format, and those normally have dozens over newlines. > > I dont see where kindly accepting some additional newlines makes > > any trouble. Expat is per definition an non-validating parser and > > thus is not intendet for syntax checking. > > This is not about fundamentalism, but about playing under the > rules. For the current matter, the rules are very clear. There are > even a reason for that rule (to simplify a bit the process of > autodetection of the document encoding). hmm, okay. overscanning some leading newlines (IMHO) wouldnt be problem, but its probably better to lay this in the hands of the application or add an special mode for that. at this point we also could implement other things like passing common entities as they are. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Tue Dec 14 17:53:54 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 17:54:00 2004 Subject: [Expat-discuss] DESTDIR patches Message-ID: <41BF1AA2.8050604@waclawek.net> It seems there are three patches available. 1) Patch # 779334: expat.Makefile.in.diff 2) Bug # 985235: expat-1.95.6-destdir.patch 3) Enrico's patch: Enrico.diff, Enrico, what is your bug #? They are all basically the same. One of them initializes DESTDIR to an empty string. Is that OK to do, or better left out? Karl From karl at waclawek.net Tue Dec 14 17:56:23 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 17:56:30 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <20041214164422.GA13268@nibiru.borg.metux.de> References: <20041214030817.GE28636@nibiru.borg.metux.de> <20041214112658.D65CE7FA5F@pointsman.pointsman.de> <20041214164422.GA13268@nibiru.borg.metux.de> Message-ID: <41BF1B37.3040708@waclawek.net> Enrico Weigelt wrote: > * rolf@pointsman.de wrote: > > > >>>I really dont care about such fundamentalists. >>>XML is a text format, and those normally have dozens over newlines. >>>I dont see where kindly accepting some additional newlines makes >>>any trouble. Expat is per definition an non-validating parser and >>>thus is not intendet for syntax checking. >> >>This is not about fundamentalism, but about playing under the >>rules. For the current matter, the rules are very clear. There are >>even a reason for that rule (to simplify a bit the process of >>autodetection of the document encoding). > > > hmm, okay. > > overscanning some leading newlines (IMHO) wouldnt be problem, but its > probably better to lay this in the hands of the application or add > an special mode for that. This is easy to do, as you are passing a buffer to Expat. Just adjust the buffer pointer and decrement the count accordingly. at this point we also could implement > other things like passing common entities as they are. What do you mean? Karl > > > cu From weigelt at metux.de Tue Dec 14 21:23:09 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 21:23:11 2004 Subject: [Expat-discuss] RFC: unitool -> universal toolchain instead of autoshit In-Reply-To: <20041214083554.GC5913@lyra.org> References: <20041214035852.GG28636@nibiru.borg.metux.de> <20041214083554.GC5913@lyra.org> Message-ID: <20041214202309.GG29394@nibiru.borg.metux.de> * Greg Stein wrote: Hi, > The autoconf/libtool scripts work just fine for Expat. We dropped > automake a long time (the worst of the bunch). > > I'm not sure how a switch would benefit Expat or its users. Well, autoconf often enough does *not* work. For example it doesn't pass -E to grep when using extendet regex, which breaks with slightly older grep (2.4.x). This ends up in configure thinking you dont have gnu ld and so refuses to build an shared library. Why the hell does an plain package, which ig. is platform independent, have to cope with platform dependend toolchain commands ?! Why can't there simply be *one* reliable command set for *every platform* ? And that's exactly where I want to do with unitool: We have *one* package which encapsulates all platform dependent commands behind some, well engineered and reliable fontend interface and is ported *once* per platform. All those common things like how to run several tools, what binfmt to use, how to find certain imported packages, cope with crosscompile and sysroot, etc, etc, are all hidden behind unitool. Evrything unitool-based packages have to do is, to simply throw the input files in front of unitool's feet and call it "build this stuff". This also makes locating and fixing build problems much easier, ie. the DESTDIR issue (remember: i posted a fix half a year ago ...) is then simply irrelevant to expat developers, since unitool handles it, and everyone knows how to instruct unitool. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From weigelt at metux.de Tue Dec 14 21:30:50 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 21:30:55 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <41BF1B37.3040708@waclawek.net> References: <20041214030817.GE28636@nibiru.borg.metux.de> <20041214112658.D65CE7FA5F@pointsman.pointsman.de> <20041214164422.GA13268@nibiru.borg.metux.de> <41BF1B37.3040708@waclawek.net> Message-ID: <20041214203050.GH29394@nibiru.borg.metux.de> * Karl Waclawek wrote: > This is easy to do, as you are passing a buffer to Expat. > Just adjust the buffer pointer and decrement the count accordingly. Right. Its trivial. One line: while ((*buf==10)||(*buf==13)||(*buf==32)||(*buf==7)) buf++; > at this point we also could implement > >other things like passing common entities as they are. > > What do you mean? In some php application I need to get some html entities for umlauts as they are, and I was really too lazy writing an extra doctype for that. (in fact the documents dont have any doctype declaration). I solved this by simply replacing these entities to some other sequences before passing the xml to expat and later converting back to htmlentities. Yes, its a dirty hack but it works. Probably there's a more elegant way ... cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Tue Dec 14 21:34:20 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 14 21:34:29 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <20041214203050.GH29394@nibiru.borg.metux.de> References: <20041214030817.GE28636@nibiru.borg.metux.de> <20041214112658.D65CE7FA5F@pointsman.pointsman.de> <20041214164422.GA13268@nibiru.borg.metux.de> <41BF1B37.3040708@waclawek.net> <20041214203050.GH29394@nibiru.borg.metux.de> Message-ID: <41BF4E4C.8030507@waclawek.net> Enrico Weigelt wrote: >>>other things like passing common entities as they are. >> >>What do you mean? > > In some php application I need to get some html entities for > umlauts as they are, and I was really too lazy writing an extra > doctype for that. > (in fact the documents dont have any doctype declaration). > > I solved this by simply replacing these entities to some other > sequences before passing the xml to expat and later converting > back to htmlentities. > > Yes, its a dirty hack but it works. Probably there's a more > elegant way ... Have a look at the XML_UseForeignDTD() API. It allows you to use a DTD with documents that don't have one. Karl From weigelt at metux.de Tue Dec 14 21:47:16 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 21:47:20 2004 Subject: [Expat-discuss] Re: DESTDIR patches In-Reply-To: <41BF1AA2.8050604@waclawek.net> References: <41BF1AA2.8050604@waclawek.net> Message-ID: <20041214204716.GI29394@nibiru.borg.metux.de> * Karl Waclawek wrote: > 2) Bug # 985235: expat-1.95.6-destdir.patch sounds like mine ? > 3) Enrico's patch: Enrico.diff, Enrico, what is your bug #? uff ? I've really got no idea if I ever had one ... I simply posted my patch on some hot day in July ... > They are all basically the same. One of them initializes > DESTDIR to an empty string. Is that OK to do, or better left out? hmm, better leave it out. it seems that this initialization overwrites variables coming from env, while those coming from commandline survive. thats at least my personal experience w/ gmake-3.79.1 I'm making heavy use of env for my buildfarm, so I'd really have it not to be initialized. *BUT* we probably should have a short talk about DESTDIR vs. INSTALL_ROOT. Many autoconf-based packages use DESTDIR, so I also tool this. But INSTALL_ROOT is also very common and seems to be clearer for me - it really states that evryting's installed as normal, but within some other root directory. (ie. PREFIX does something really different) Its probably best to support both. I've attached a patch which sets $(DESTDIR) to $(INSTALL_ROOT) if $(INSTALL_ROOT) is set. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- -------------- next part -------------- diff -ruN expat-1.95.8.orig/Makefile.in expat-1.95.8/Makefile.in --- expat-1.95.8.orig/Makefile.in Tue Dec 14 21:40:57 2004 +++ expat-1.95.8/Makefile.in Tue Dec 14 21:43:34 2004 @@ -44,6 +44,9 @@ APIHEADER = $(srcdir)/lib/expat.h $(srcdir)/lib/expat_external.h LIBRARY = libexpat.la +ifdef INSTALL_ROOT +DESTDIR=$(INSTALL_ROOT) +endif default: buildlib xmlwf/xmlwf -------------- next part -------------- diff -ruN expat-1.95.8.orig/Makefile.in expat-1.95.8/Makefile.in --- expat-1.95.8.orig/Makefile.in 2004-12-03 16:48:45.000000000 +0100 +++ expat-1.95.8/Makefile.in 2004-12-03 16:50:03.000000000 +0100 @@ -73,22 +73,22 @@ tests/runtests install: xmlwf/xmlwf installlib - $(mkinstalldirs) $(bindir) $(man1dir) - $(LIBTOOL) --mode=install $(INSTALL_PROGRAM) xmlwf/xmlwf $(bindir)/xmlwf - $(INSTALL_DATA) $(MANFILE) $(man1dir) + $(mkinstalldirs) $(DESTDIR)$(bindir) $(DESTDIR)$(man1dir) + $(LIBTOOL) --mode=install $(INSTALL_PROGRAM) xmlwf/xmlwf $(DESTDIR)$(bindir)/xmlwf + $(INSTALL_DATA) $(MANFILE) $(DESTDIR)$(man1dir) installlib: $(LIBRARY) $(APIHEADER) - $(mkinstalldirs) $(libdir) $(includedir) - $(LIBTOOL) --mode=install $(INSTALL) $(LIBRARY) $(libdir)/$(LIBRARY) - $(INSTALL_DATA) $(APIHEADER) $(includedir) + $(mkinstalldirs) $(DESTDIR)$(libdir) $(DESTDIR)$(includedir) + $(LIBTOOL) --mode=install $(INSTALL) $(LIBRARY) $(DESTDIR)$(libdir)/$(LIBRARY) + $(INSTALL_DATA) $(APIHEADER) $(DESTDIR)$(includedir) uninstall: uninstalllib - $(LIBTOOL) --mode=uninstall rm -f $(bindir)/xmlwf - rm -f $(man1dir)/xmlwf.1 + $(LIBTOOL) --mode=uninstall rm -f $(DESTDIR)$(bindir)/xmlwf + rm -f $(DESTDIR)$(man1dir)/xmlwf.1 uninstalllib: - $(LIBTOOL) --mode=uninstall rm -f $(libdir)/$(LIBRARY) - rm -f $(includedir)/$(APIHEADER) + $(LIBTOOL) --mode=uninstall rm -f $(DESTDIR)$(libdir)/$(LIBRARY) + rm -f $(DESTDIR)$(includedir)/$(APIHEADER) # for VPATH builds (invoked by configure) mkdir-init: From sb at dod.no Tue Dec 14 21:57:40 2004 From: sb at dod.no (Steinar Bang) Date: Tue Dec 14 21:57:47 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries References: <001301c4b451$665825a0$0200a8c0@karlglen188> <87k6tpjhhu.fsf@dod.no> <001801c4b48a$3fbffdd0$097ca8c0@karlglen188> <87brf1j9dp.fsf@dod.no> <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <874qkqbnlw.fsf@dod.no> <001b01c4b605$cff09a00$057ca8c0@karlglen188> <87zn2ia24t.fsf@dod.no> <001d01c4b61f$a39aa940$047ca8c0@karlglen188> <87brdy4uhj.fsf@dod.no> <20041214022643.GC28636@nibiru.borg.metux.de> Message-ID: <874qiooawr.fsf@dod.no> >>>>> Enrico Weigelt : > What stands against providing either *one* API (8bit *XOR* 16bit) > and putting write frontend of the other variant or provide two > different sets of functions (ie xml8_xxx, xml16_xxx) ? My suggestion http://sourceforge.net/tracker/index.php?func=detail&aid=1048448&group_id=10127&atid=110127 was use preprosessor magic to accomplish something like the above function names. The reason for using preprocessor magic, was to allow current expat-using code to be compile time compatible. Have your patch been attached to a message posted to the list? If so, I've missed it. What did it do? From sb at dod.no Tue Dec 14 22:09:34 2004 From: sb at dod.no (Steinar Bang) Date: Tue Dec 14 22:09:41 2004 Subject: [Expat-discuss] Re: Bugs in expat 1.95.8 References: <41A33F33.9040103@hypershell.com> <001b01c4d163$583c70a0$9e539696@torstar.net> <20041214015353.GA27907@nibiru.borg.metux.de> <41BE5093.9040203@waclawek.net> <20041214030817.GE28636@nibiru.borg.metux.de> Message-ID: <87zn0gmvsh.fsf@dod.no> >>>>> Enrico Weigelt : > * Karl Waclawek wrote: > >> >Nevertheless its bad behaviour of expat and shall be fixed. >> For saying this, the wrath of the XML gods will come upon you. ;-) > I really dont care about such fundamentalists. XML is a text format, > and those normally have dozens over newlines. I dont see where > kindly accepting some additional newlines makes any trouble. Expat > is per definition an non-validating parser and thus is not intendet > for syntax checking. No, but it is a well-formedness checking parser. And well-formed XML MUST start with " References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> Message-ID: <20041214211405.GA20532@nibiru.borg.metux.de> * Karl Waclawek wrote: > This patch doesn't work for current CVS. > Why don't you do a quick merge from CVS and send me the patch again? current cvs doesnt work for me at all. it argues: # configure.in:56: error: possibly undefined macro: AC_LIBTOOL_WIN32_DLL # If this token and others are legitimate, please use m4_pattern_allow. # See the Autoconf documentation. # configure.in:57: error: possibly undefined macro: AC_PROG_LIBTOOL an in fact these macros arent replaced and appear in configure as they are in configure.in also configure breaks: ./configure: line 1344: AC_LIBTOOL_WIN32_DLL: command not found ./configure: line 1345: AC_PROG_LIBTOOL: command not found checking for gcc... gcc checking for C compiler default output file name... a.out checking whether the C compiler works... yes checking whether we are cross compiling... no checking for suffix of executables... checking for suffix of object files... o checking whether we are using the GNU C compiler... yes checking whether gcc accepts -g... yes checking for gcc option to accept ANSI C... none needed checking for a BSD-compatible install... /usr/bin/ginstall -c checking whether gcc accepts -fexceptions... yes checking how to run the C preprocessor... gcc -E checking for egrep... grep -E checking for ANSI C header files... yes checking whether byte ordering is bigendian... no checking for an ANSI C-conforming const... yes checking for sys/types.h... yes checking for sys/stat.h... yes checking for stdlib.h... yes checking for string.h... yes checking for memory.h... yes checking for strings.h... yes checking for inttypes.h... yes checking for stdint.h... yes checking for unistd.h... yes checking for size_t... yes checking for memmove... yes checking for bcopy... yes checking fcntl.h usability... yes checking fcntl.h presence... yes checking for fcntl.h... yes checking for unistd.h... (cached) yes checking for off_t... yes checking for stdlib.h... (cached) yes checking for unistd.h... (cached) yes checking for getpagesize... yes checking for working mmap... yes checking check.h usability... no checking check.h presence... no checking for check.h... no checking for check.h... (cached) no configure: creating ./config.status config.status: creating Makefile config.status: creating expat_config.h config.status: error: cannot find input file: expat_config.h.in so, current cvs version is unusable. cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From weigelt at metux.de Tue Dec 14 22:33:42 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 22:33:45 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <41BF4E4C.8030507@waclawek.net> References: <20041214030817.GE28636@nibiru.borg.metux.de> <20041214112658.D65CE7FA5F@pointsman.pointsman.de> <20041214164422.GA13268@nibiru.borg.metux.de> <41BF1B37.3040708@waclawek.net> <20041214203050.GH29394@nibiru.borg.metux.de> <41BF4E4C.8030507@waclawek.net> Message-ID: <20041214213342.GJ29394@nibiru.borg.metux.de> * Karl Waclawek wrote: > Have a look at the XML_UseForeignDTD() API. > It allows you to use a DTD with documents that don't have one. ah, if I enable this, the externelentitfyref handler is called when an (undefined) entity comes in and it just have to return the text which the given entity has to be replaced with ? cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From weigelt at metux.de Tue Dec 14 22:35:38 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Tue Dec 14 22:46:04 2004 Subject: [Expat-discuss] Re: Bugs in expat 1.95.8 In-Reply-To: <87zn0gmvsh.fsf@dod.no> References: <41A33F33.9040103@hypershell.com> <001b01c4d163$583c70a0$9e539696@torstar.net> <20041214015353.GA27907@nibiru.borg.metux.de> <41BE5093.9040203@waclawek.net> <20041214030817.GE28636@nibiru.borg.metux.de> <87zn0gmvsh.fsf@dod.no> Message-ID: <20041214213538.GK29394@nibiru.borg.metux.de> * Steinar Bang wrote: > No, but it is a well-formedness checking parser. And well-formed XML > MUST start with " http://w3.org/TR/2004/REC-xml-20040204/#sec-well-formed > http://w3.org/TR/2004/REC-xml-20040204/#charsets > (for the first link follow the "prolog" link in the BNF production). hmm, can this be disabled or do I have to put an kind of preprocessor in front of it when my files dont have References: <20041214030817.GE28636@nibiru.borg.metux.de> <20041214112658.D65CE7FA5F@pointsman.pointsman.de> <20041214164422.GA13268@nibiru.borg.metux.de> <41BF1B37.3040708@waclawek.net> <20041214203050.GH29394@nibiru.borg.metux.de> <41BF4E4C.8030507@waclawek.net> <20041214213342.GJ29394@nibiru.borg.metux.de> Message-ID: <41BF7268.7010701@waclawek.net> Enrico Weigelt wrote: > * Karl Waclawek wrote: > > > >>Have a look at the XML_UseForeignDTD() API. >>It allows you to use a DTD with documents that don't have one. > > ah, if I enable this, the externelentitfyref handler is called > when an (undefined) entity comes in and it just have to return > the text which the given entity has to be replaced with ? No, what you have to do is parse your own custom DTD in the ExternalEntityRefHandler, just like a DTD supplied with the document. This custom DTD should contain all the entity declarations you need. So, what the option does: It triggers a call-back to the ExternalEntityRefHandler at the point where a DTD would be normally processed, even if there is no DocType declaration. Karl From karl at waclawek.net Wed Dec 15 00:14:51 2004 From: karl at waclawek.net (Karl Waclawek) Date: Wed Dec 15 00:14:47 2004 Subject: [Expat-discuss] Re: DESTDIR patches In-Reply-To: <20041214211405.GA20532@nibiru.borg.metux.de> References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> Message-ID: <41BF73EB.6080603@waclawek.net> Enrico Weigelt wrote: > * Karl Waclawek wrote: > > >>This patch doesn't work for current CVS. >>Why don't you do a quick merge from CVS and send me the patch again? > > > current cvs doesnt work for me at all. it argues: I was talking about your patch to Makefile.in only. The rest is another issue. Can you figure out what is wrong with it? Karl From karl at waclawek.net Wed Dec 15 00:26:54 2004 From: karl at waclawek.net (Karl Waclawek) Date: Wed Dec 15 00:26:51 2004 Subject: [Expat-discuss] Re: Bugs in expat 1.95.8 In-Reply-To: <20041214213538.GK29394@nibiru.borg.metux.de> References: <41A33F33.9040103@hypershell.com> <001b01c4d163$583c70a0$9e539696@torstar.net> <20041214015353.GA27907@nibiru.borg.metux.de> <41BE5093.9040203@waclawek.net> <20041214030817.GE28636@nibiru.borg.metux.de> <87zn0gmvsh.fsf@dod.no> <20041214213538.GK29394@nibiru.borg.metux.de> Message-ID: <41BF76BE.80006@waclawek.net> Enrico Weigelt wrote: > * Steinar Bang wrote: > > > >>No, but it is a well-formedness checking parser. And well-formed XML >>MUST start with "> http://w3.org/TR/2004/REC-xml-20040204/#sec-well-formed >> http://w3.org/TR/2004/REC-xml-20040204/#charsets >>(for the first link follow the "prolog" link in the BNF production). > > hmm, can this be disabled or do I have to put an kind of preprocessor > in front of it when my files dont have <001b01c4d163$583c70a0$9e539696@torstar.net> <20041214015353.GA27907@nibiru.borg.metux.de> <41BE5093.9040203@waclawek.net> <20041214030817.GE28636@nibiru.borg.metux.de> <87zn0gmvsh.fsf@dod.no> <20041214213538.GK29394@nibiru.borg.metux.de> <41BF76BE.80006@waclawek.net> Message-ID: <87pt1bzt93.fsf@dod.no> >>>>> Karl Waclawek : > Enrico Weigelt wrote: >> * Steinar Bang wrote: >> >>> http://w3.org/TR/2004/REC-xml-20040204/#sec-well-formed > The XML declaration is optional, so you don't need to add one. Hm... so it is. I've never noticed that question mark before. Has it been optional all along? I thought I once got a well-formedness error because it was missing? From rolf at pointsman.de Wed Dec 15 13:01:36 2004 From: rolf at pointsman.de (rolf@pointsman.de) Date: Wed Dec 15 12:57:09 2004 Subject: [Expat-discuss] Re: Bugs in expat 1.95.8 In-Reply-To: <87pt1bzt93.fsf@dod.no> Message-ID: <20041215120139.D40C18083D@pointsman.pointsman.de> On 15 Dec, Steinar Bang wrote: >>>>>> Karl Waclawek : >> The XML declaration is optional, so you don't need to add one. > > Hm... so it is. I've never noticed that question mark before. Has it > been optional all along? I thought I once got a well-formedness error > because it was missing? The XML declaration was alwas optional, since the first edition of the XML recommendation, more than six years (or better almost seven years now) ago. See http://www.w3.org/TR/1998/REC-xml-19980210#sec-prolog-dtd rolf From karl at waclawek.net Wed Dec 15 14:54:05 2004 From: karl at waclawek.net (Karl Waclawek) Date: Wed Dec 15 14:54:10 2004 Subject: [Expat-discuss] Re: Bugs in expat 1.95.8 In-Reply-To: <87pt1bzt93.fsf@dod.no> References: <41A33F33.9040103@hypershell.com> <001b01c4d163$583c70a0$9e539696@torstar.net> <20041214015353.GA27907@nibiru.borg.metux.de> <41BE5093.9040203@waclawek.net> <20041214030817.GE28636@nibiru.borg.metux.de> <87zn0gmvsh.fsf@dod.no> <20041214213538.GK29394@nibiru.borg.metux.de> <41BF76BE.80006@waclawek.net> <87pt1bzt93.fsf@dod.no> Message-ID: <41C041FD.5090602@waclawek.net> Steinar Bang wrote: >>The XML declaration is optional, so you don't need to add one. > > > Hm... so it is. I've never noticed that question mark before. Has it > been optional all along? I thought I once got a well-formedness error > because it was missing? Currently it seems to work correctly. I don't remember that I fixed such an issue, but it is possible that it didn't work at some point in the past. Karl From kward at orange.us Tue Dec 14 23:58:38 2004 From: kward at orange.us (Ken Ward) Date: Wed Dec 15 16:19:06 2004 Subject: [Expat-discuss] 'ASCII encoding error' with encoding 'iso-8859-1' Message-ID: Hello, I'm using expat from Python 2.1.1 on RedHat Linux. I think it's expat 1.95.1, based on running 'strings' on the library. I'm trying to parse a file with Latin-1 characters above 128. I set the encoding to "iso-8859-1" in ParserCreate, but I get this error: UnicodeError: ASCII encoding error: ordinal not in range(128) Is this expected to work in that old version? Thank you, Ken Ward From karl at waclawek.net Wed Dec 15 16:56:59 2004 From: karl at waclawek.net (Karl Waclawek) Date: Wed Dec 15 16:57:11 2004 Subject: [Expat-discuss] 'ASCII encoding error' with encoding 'iso-8859-1' In-Reply-To: References: Message-ID: <41C05ECB.3050504@waclawek.net> Ken Ward wrote: > Hello, > > I'm using expat from Python 2.1.1 on RedHat Linux. I think it's expat > 1.95.1, based on running 'strings' on the library. > > I'm trying to parse a file with Latin-1 characters above 128. I set the > encoding to "iso-8859-1" in ParserCreate, but I get this error: > UnicodeError: ASCII encoding error: ordinal not in range(128) This does not seem to be an error message generated by Expat. Karl From kward at orange.us Wed Dec 15 18:04:44 2004 From: kward at orange.us (Ken Ward) Date: Wed Dec 15 18:04:46 2004 Subject: [Expat-discuss] 'ASCII encoding error' with encoding 'iso-8859-1' Message-ID: No, it's from Python, but I suppose that it results from expat rejecting the non-ASCII characters. Maybe not. Python reads the characters OK - my insufficient workaround is to translate all characters over 127 to '?' before feeding the XML to the parser object which expat underlies. > -----Original Message----- > From: Karl Waclawek [mailto:karl@waclawek.net] > Sent: Wednesday, December 15, 2004 10:57 AM > To: Ken Ward > Cc: expat-discuss@libexpat.org > Subject: Re: [Expat-discuss] 'ASCII encoding error' with > encoding 'iso-8859-1' > > Ken Ward wrote: > > Hello, > > > > I'm using expat from Python 2.1.1 on RedHat Linux. I think > it's expat > > 1.95.1, based on running 'strings' on the library. > > > > I'm trying to parse a file with Latin-1 characters above > 128. I set the > > encoding to "iso-8859-1" in ParserCreate, but I get this error: > > UnicodeError: ASCII encoding error: ordinal not in range(128) > > This does not seem to be an error message generated by Expat. > > Karl > > From karl at waclawek.net Wed Dec 15 18:12:38 2004 From: karl at waclawek.net (Karl Waclawek) Date: Wed Dec 15 18:12:47 2004 Subject: [Expat-discuss] 'ASCII encoding error' with encoding 'iso-8859-1' In-Reply-To: References: Message-ID: <41C07086.4080405@waclawek.net> Ken Ward wrote: > No, it's from Python, but I suppose that it results from expat rejecting > the non-ASCII characters. Maybe not. > > Python reads the characters OK - my insufficient workaround is to > translate all characters over 127 to '?' before feeding the XML to the > parser object which expat underlies. We haven't supported Expat 1.95.1 for a very long time. If at all possible, you should upgrade. If the problem still exists, then it is likely not Expat. Karl From peterv at propagandism.org Thu Dec 16 23:15:51 2004 From: peterv at propagandism.org (Peter Van der Beken) Date: Thu Dec 16 23:15:53 2004 Subject: [Expat-discuss] XML_ParserCreateNS In-Reply-To: <41BE3F4E.3060209@waclawek.net> References: <41BE04A1.2030104@propagandism.org> <41BE3F4E.3060209@waclawek.net> Message-ID: <41C20917.2010107@propagandism.org> Karl Waclawek wrote: > I am inclined to change the code to simply use whatever > namespace separator is supplied, even if it is '\0'. > > Can you think of any use case for having the separator omitted? > Fred, if you read this - do you remember why this feature > was provided? I don't see how '\0' can ever be useful as a separator, since Expat can send 1, 2 or 3 pieces (namespace URI, name, prefix) and the string is null-terminated. There's no way to find out if one should read beyond the '\0', since it can be the terminator of the string or the separator. > So, I suggest two steps: > - modify namespace separator code as above and adjust comments/docs Not sure about this one, see above. > - remove the sentence above from the comments/docs Sounds good. Thanks, Peter From karl at waclawek.net Fri Dec 17 01:01:52 2004 From: karl at waclawek.net (Karl Waclawek) Date: Fri Dec 17 01:01:54 2004 Subject: [Expat-discuss] XML_ParserCreateNS In-Reply-To: <41C20917.2010107@propagandism.org> References: <41BE04A1.2030104@propagandism.org> <41BE3F4E.3060209@waclawek.net> <41C20917.2010107@propagandism.org> Message-ID: <41C221F0.80200@waclawek.net> Peter Van der Beken wrote: > Karl Waclawek wrote: > >> I am inclined to change the code to simply use whatever >> namespace separator is supplied, even if it is '\0'. >> >> Can you think of any use case for having the separator omitted? >> Fred, if you read this - do you remember why this feature >> was provided? > > > I don't see how '\0' can ever be useful as a separator, since Expat can > send 1, 2 or 3 pieces (namespace URI, name, prefix) and the string is > null-terminated. There's no way to find out if one should read beyond > the '\0', since it can be the terminator of the string or the separator. Of course, useful it would not be. Just like many other characters that would be bad choices ('/', ':', A..Z, a..z, 0..9, ...), and we don't check for them either. The question is, since the programmer can already shoot him/herself in the foot, what damage could one more bull's eye on the foot do? Karl From weigelt at metux.de Fri Dec 17 18:28:37 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Fri Dec 17 18:28:40 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <41BF7268.7010701@waclawek.net> References: <20041214030817.GE28636@nibiru.borg.metux.de> <20041214112658.D65CE7FA5F@pointsman.pointsman.de> <20041214164422.GA13268@nibiru.borg.metux.de> <41BF1B37.3040708@waclawek.net> <20041214203050.GH29394@nibiru.borg.metux.de> <41BF4E4C.8030507@waclawek.net> <20041214213342.GJ29394@nibiru.borg.metux.de> <41BF7268.7010701@waclawek.net> Message-ID: <20041217172837.GD26310@nibiru.borg.metux.de> * Karl Waclawek wrote: > No, what you have to do is parse your own custom DTD in > the ExternalEntityRefHandler, just like a DTD supplied > with the document. This custom DTD should contain all the > entity declarations you need. hmm. could you please give a short example ? I've really no experiences with DTD at all :(( cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From weigelt at metux.de Fri Dec 17 18:40:34 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Fri Dec 17 18:40:38 2004 Subject: [Expat-discuss] Re: DESTDIR patches In-Reply-To: <41BF73EB.6080603@waclawek.net> References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> Message-ID: <20041217174034.GE26310@nibiru.borg.metux.de> * Karl Waclawek wrote: > I was talking about your patch to Makefile.in only. > > The rest is another issue. Can you figure out what is wrong with it? Here's a new patch against cvs Makefile.in cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- -------------- next part -------------- diff -ruN expat.orig/Makefile.in expat/Makefile.in --- expat.orig/Makefile.in 2004-12-17 18:30:25.000000000 +0100 +++ expat/Makefile.in 2004-12-17 18:34:40.000000000 +0100 @@ -44,6 +44,9 @@ APIHEADER = $(srcdir)/lib/expat.h $(srcdir)/lib/expat_external.h LIBRARY = libexpat.la +ifndef INSTALL_ROOT +INSTALL_ROOT=$(DESTDIR) +endif default: buildlib xmlwf/xmlwf @@ -73,22 +76,22 @@ tests/runtests install: xmlwf/xmlwf installlib - $(mkinstalldirs) $(bindir) $(man1dir) - $(LIBTOOL) --mode=install $(INSTALL_PROGRAM) xmlwf/xmlwf $(bindir)/xmlwf - $(INSTALL_DATA) $(MANFILE) $(man1dir) + $(mkinstalldirs) $(INSTALL_ROOT)$(bindir) $(INSTALL_ROOT)$(man1dir) + $(LIBTOOL) --mode=install $(INSTALL_PROGRAM) xmlwf/xmlwf $(INSTALL_ROOT)$(bindir)/xmlwf + $(INSTALL_DATA) $(MANFILE) $(INSTALL_ROOT)$(man1dir) installlib: $(LIBRARY) $(APIHEADER) - $(mkinstalldirs) $(libdir) $(includedir) - $(LIBTOOL) --mode=install $(INSTALL) $(LIBRARY) $(libdir)/$(LIBRARY) - for FN in $(APIHEADER) ; do $(INSTALL_DATA) $$FN $(includedir) ; done + $(mkinstalldirs) $(INSTALL_ROOT)$(libdir) $(INSTALL_ROOT)$(includedir) + $(LIBTOOL) --mode=install $(INSTALL) $(LIBRARY) $(INSTALL_ROOT)$(libdir)/$(LIBRARY) + for FN in $(APIHEADER) ; do $(INSTALL_DATA) $$FN $(INSTALL_ROOT)$(includedir) ; done uninstall: uninstalllib - $(LIBTOOL) --mode=uninstall rm -f $(bindir)/xmlwf - rm -f $(man1dir)/xmlwf.1 + $(LIBTOOL) --mode=uninstall rm -f $(INSTALL_ROOT)$(bindir)/xmlwf + rm -f $(INSTALL_ROOT)$(man1dir)/xmlwf.1 uninstalllib: - $(LIBTOOL) --mode=uninstall rm -f $(libdir)/$(LIBRARY) - rm -f $(includedir)/$(APIHEADER) + $(LIBTOOL) --mode=uninstall rm -f $(INSTALL_ROOT)$(libdir)/$(LIBRARY) + rm -f $(INSTALL_ROOT)$(includedir)/$(APIHEADER) # for VPATH builds (invoked by configure) mkdir-init: From karl at waclawek.net Fri Dec 17 18:59:53 2004 From: karl at waclawek.net (Karl Waclawek) Date: Fri Dec 17 18:59:59 2004 Subject: [Expat-discuss] Bugs in expat 1.95.8 [WAS: Expat 1.95.8 fails on XML with newline] In-Reply-To: <20041217172837.GD26310@nibiru.borg.metux.de> References: <20041214030817.GE28636@nibiru.borg.metux.de> <20041214112658.D65CE7FA5F@pointsman.pointsman.de> <20041214164422.GA13268@nibiru.borg.metux.de> <41BF1B37.3040708@waclawek.net> <20041214203050.GH29394@nibiru.borg.metux.de> <41BF4E4C.8030507@waclawek.net> <20041214213342.GJ29394@nibiru.borg.metux.de> <41BF7268.7010701@waclawek.net> <20041217172837.GD26310@nibiru.borg.metux.de> Message-ID: <41C31E99.7030202@waclawek.net> Enrico Weigelt wrote: > * Karl Waclawek wrote: > > > >>No, what you have to do is parse your own custom DTD in >>the ExternalEntityRefHandler, just like a DTD supplied >>with the document. This custom DTD should contain all the >>entity declarations you need. > > > hmm. could you please give a short example ? > I've really no experiences with DTD at all :(( The basic idea with Expat is that external entities are parsed with a separate Expat parser instance. This is performed within the ExternalEntityRefHandler. The "child" parser is created using XML_ExternalEntityParserCreate(), so that it inherits all the handlers. The DTD is just another entity to be parsed, and the parser will call the ExternalEntityRefHandler() at the right point. For an external DTD that is not referenced in the XML document, the ExternalEntityRefHandler() is called with both publicId and systemId set to NULL, and that is how you identify it. If you have a look at the xmlwf demo app included with Expat, there is an external entity reference handler: static int externalEntityRefFilemap(XML_Parser parser, const XML_Char *context, const XML_Char *base, const XML_Char *systemId, const XML_Char *publicId) { int result; XML_Char *s; const XML_Char *filename; XML_Parser entParser = XML_ExternalEntityParserCreate(parser, context, 0); PROCESS_ARGS args; args.retPtr = &result; args.parser = entParser; filename = resolveSystemId(base, systemId, &s); XML_SetBase(entParser, filename); if (!filemap(filename, processFile, &args)) result = 0; free(s); XML_ParserFree(entParser); return result; } which uses this: static void processFile(const void *data, size_t size, const XML_Char *filename, void *args) { XML_Parser parser = ((PROCESS_ARGS *)args)->parser; int *retPtr = ((PROCESS_ARGS *)args)->retPtr; if (XML_Parse(parser, (const char *)data, size, 1) == XML_STATUS_ERROR) { reportError(parser, filename); *retPtr = 0; } else *retPtr = 1; } Karl From karl at waclawek.net Fri Dec 17 19:08:35 2004 From: karl at waclawek.net (Karl Waclawek) Date: Fri Dec 17 19:08:43 2004 Subject: [Expat-discuss] Re: DESTDIR patches In-Reply-To: <20041217174034.GE26310@nibiru.borg.metux.de> References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> <20041217174034.GE26310@nibiru.borg.metux.de> Message-ID: <41C320A3.4000607@waclawek.net> Enrico Weigelt wrote: > * Karl Waclawek wrote: > > >>I was talking about your patch to Makefile.in only. >> >>The rest is another issue. Can you figure out what is wrong with it? > > > Here's a new patch against cvs Makefile.in Applied to CVS. Karl From fdrake at acm.org Wed Dec 15 17:18:16 2004 From: fdrake at acm.org (Fred L. Drake, Jr.) Date: Sun Dec 19 21:59:43 2004 Subject: [Expat-discuss] XML_ParserCreateNS In-Reply-To: <41BE3F4E.3060209@waclawek.net> References: <41BE04A1.2030104@propagandism.org> <41BE3F4E.3060209@waclawek.net> Message-ID: <200412150618.16347.fdrake@acm.org> On Monday 13 December 2004 15:18, Karl Waclawek wrote: > Can you think of any use case for having the separator omitted? > Fred, if you read this - do you remember why this feature > was provided? This is for RDF users. RDF/XML has a very weird intepretation of namespace, likely stemming from the RDF crowd's convention of ending namespace URIs with the "#" character. The namespace URI and the local name are simply jammed together. So the following two elements should both have a "fuill name" of 'http://xml.example.org/foo': While I'm of the mind that this is insane, that is what the RDF/XML specs say must be done. We should keep the special case for a namespace separator of '\0' and fix anyplace this is handled according to the documentation. We should also document that this should not be combined with namespace triplets. That's a programmer error; we don't actually need to catch that case and return an error code; we need not to segfault. > This makes no sense to me either - when namespace processing is > turned on - as the parser would then not conform to the namespaces > specification, if I remember correctly. Agreed. Expat used to do that, but we determined that was wrong after we'd read the spec enough times. > So, I suggest two steps: > - modify namespace separator code as above and adjust comments/docs -1, as discussed above. > - remove the sentence above from the comments/docs +1. -Fred -- Fred L. Drake, Jr. From fdrake at acm.org Wed Dec 15 17:25:41 2004 From: fdrake at acm.org (Fred L. Drake, Jr.) Date: Sun Dec 19 21:59:45 2004 Subject: [Expat-discuss] Re: Conflict with two expat shared libraries In-Reply-To: <41BE6A18.50503@waclawek.net> References: <002701c4b49d$1a5b6940$097ca8c0@karlglen188> <20041214041052.GH28636@nibiru.borg.metux.de> <41BE6A18.50503@waclawek.net> Message-ID: <200412150625.41448.fdrake@acm.org> On Monday 13 December 2004 18:20, Karl Waclawek wrote: > Well, I am not the project administrator, but I am sure > that Fred would welcome any competent help, busy as he is. > He is cc'ed on this, so unless he really is not available > (he could be on vacation or travelling), he will likely > respond to you. I am indeed travelling (Waikiki on business). Hopefully I'll be able to catch up on all the emails before the end of the year, but I'll be spending a few more days travelling with my family before the new year. -Fred -- Fred L. Drake, Jr. From karl at waclawek.net Mon Dec 20 03:36:43 2004 From: karl at waclawek.net (Karl Waclawek) Date: Mon Dec 20 03:36:41 2004 Subject: [Expat-discuss] XML_ParserCreateNS In-Reply-To: <200412150618.16347.fdrake@acm.org> References: <41BE04A1.2030104@propagandism.org> <41BE3F4E.3060209@waclawek.net> <200412150618.16347.fdrake@acm.org> Message-ID: <41C63ABB.4050302@waclawek.net> Fred L. Drake, Jr. wrote: > On Monday 13 December 2004 15:18, Karl Waclawek wrote: > > Can you think of any use case for having the separator omitted? > > Fred, if you read this - do you remember why this feature > > was provided? > > This is for RDF users. RDF/XML has a very weird intepretation of namespace, > likely stemming from the RDF crowd's convention of ending namespace URIs with > the "#" character. The namespace URI and the local name are simply jammed > together. So the following two elements should both have a "fuill name" of > 'http://xml.example.org/foo': > > > > > > While I'm of the mind that this is insane, that is what the RDF/XML specs say > must be done. We should keep the special case for a namespace separator of > '\0' and fix anyplace this is handled according to the documentation. OK, so we keep it to allow RDF processors to skip an extra step of uri+localName concatenation, as we are all about performance, right? There is probably a bug in the attribute name generation, where it does not deal with this special case. I may have introduced this bug when I fixed the bug of Expat not detecting duplicate attributes with different prefixes. However, I really would like to change the name reporting for post 2.0 releases to pass uri, localName and prefix as three separate arguments, especially since we already have them available. Building the current name argument using separators is just extra work for Expat. For RDF processor there would then be the extra string concatenation mentioned above, but this would be offset by what we gain in Expat. > We should also document that this should not be combined with namespace > triplets. That's a programmer error; we don't actually need to catch that > case and return an error code; we need not to segfault. I don't think we segfault currently, do we? > > This makes no sense to me either - when namespace processing is > > turned on - as the parser would then not conform to the namespaces > > specification, if I remember correctly. > > Agreed. Expat used to do that, but we determined that was wrong after we'd > read the spec enough times. > > > So, I suggest two steps: > > - modify namespace separator code as above and adjust comments/docs > > -1, as discussed above. > > > - remove the sentence above from the comments/docs > > +1. OK, for now. ;-) Karl From karl at waclawek.net Mon Dec 20 05:19:52 2004 From: karl at waclawek.net (Karl Waclawek) Date: Mon Dec 20 05:19:49 2004 Subject: [Expat-discuss] XML_ParserCreateNS In-Reply-To: <41C63ABB.4050302@waclawek.net> References: <41BE04A1.2030104@propagandism.org> <41BE3F4E.3060209@waclawek.net> <200412150618.16347.fdrake@acm.org> <41C63ABB.4050302@waclawek.net> Message-ID: <41C652E8.2050503@waclawek.net> Karl Waclawek wrote: > >> While I'm of the mind that this is insane, that is what the RDF/XML >> specs say must be done. We should keep the special case for a >> namespace separator of '\0' and fix anyplace this is handled according >> to the documentation. > > > OK, so we keep it to allow RDF processors to skip an extra step of > uri+localName concatenation, as we are all about performance, right? > There is probably a bug in the attribute name generation, where it does > not deal with this special case. I may have introduced this bug when I > fixed > the bug of Expat not detecting duplicate attributes with different > prefixes. Follow up after looking at the code: I think in the case of namespace triplets the simplest fix for the bug above is to *always* add the namespace separator between local name and prefix even if it is '\0'. As I have already changed the docs to state that it is an error to combine '\0' with triplets, the implementation is free to choose its behaviour for that case, as long as it is safe. Karl From peterv at propagandism.org Mon Dec 20 12:09:45 2004 From: peterv at propagandism.org (Peter Van der Beken) Date: Mon Dec 20 12:09:49 2004 Subject: [Expat-discuss] XML_ParserCreateNS In-Reply-To: <41C63ABB.4050302@waclawek.net> References: <41BE04A1.2030104@propagandism.org> <41BE3F4E.3060209@waclawek.net> <200412150618.16347.fdrake@acm.org> <41C63ABB.4050302@waclawek.net> Message-ID: <41C6B2F9.8030906@propagandism.org> Karl Waclawek wrote: > However, I really would like to change the name reporting for post 2.0 > releases to pass uri, localName and prefix as three separate arguments, > especially since we already have them available. Building the current > name argument using separators is just extra work for Expat. I totally want this :-). Right now I have to parse the resulting string that I get from Expat to split out the three arguments again. Moreover, as the length of the string is not passed along I can't start scanning for the separator from the end of the string and so often we have to go over the whole namespace URI, which is the longest of the three arguments in most cases. A lot of overhead that we could avoid. Peter From xcross at us.ibm.com Tue Dec 21 13:29:45 2004 From: xcross at us.ibm.com (Chris Cross) Date: Tue Dec 21 13:42:24 2004 Subject: [Expat-discuss] Chris Cross/West Palm Beach/IBM is out of the office. Message-ID: I will be out of the office starting 12/17/2004 and will not return until 01/03/2005. I will be checking my email occasionally in "vacation mode." From fdrake at acm.org Tue Dec 21 17:57:21 2004 From: fdrake at acm.org (Fred L. Drake, Jr.) Date: Tue Dec 21 17:58:05 2004 Subject: [Expat-discuss] XML_ParserCreateNS In-Reply-To: <41C652E8.2050503@waclawek.net> References: <41BE04A1.2030104@propagandism.org> <41C63ABB.4050302@waclawek.net> <41C652E8.2050503@waclawek.net> Message-ID: <200412211155.42411.fdrake@acm.org> On Sunday 19 December 2004 23:19, Karl Waclawek wrote: > I think in the case of namespace triplets the simplest fix > for the bug above is to *always* add the namespace separator > between local name and prefix even if it is '\0'. > As I have already changed the docs to state that it is an error > to combine '\0' with triplets, the implementation is free to choose > its behaviour for that case, as long as it is safe. Yes, that's perfectly fine. It's only the namespace URI / local name joining that's important when the separator is '\0'. -Fred -- Fred L. Drake, Jr. From weigelt at metux.de Sun Dec 26 19:29:37 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Sun Dec 26 19:37:04 2004 Subject: [Expat-discuss] Re: DESTDIR patches In-Reply-To: <41C320A3.4000607@waclawek.net> References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> <20041217174034.GE26310@nibiru.borg.metux.de> <41C320A3.4000607@waclawek.net> Message-ID: <20041226182937.GG25408@nibiru.borg.metux.de> * Karl Waclawek wrote: > >Here's a new patch against cvs Makefile.in > > Applied to CVS. Okay. Great. Is the current release also fixed ? cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Mon Dec 27 00:08:43 2004 From: karl at waclawek.net (Karl Waclawek) Date: Mon Dec 27 00:08:42 2004 Subject: [Expat-discuss] Re: DESTDIR patches In-Reply-To: <20041226182937.GG25408@nibiru.borg.metux.de> References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> <20041217174034.GE26310@nibiru.borg.metux.de> <41C320A3.4000607@waclawek.net> <20041226182937.GG25408@nibiru.borg.metux.de> Message-ID: <41CF447B.1040209@waclawek.net> We never change releases, especially not ones that have been out for a while, we rather make new ones. All the changes in CVS will be part of Expat 2.0. Since you will ask when it will be out: This depends mostly on how much time we have writing regression tests for the fixes and packaging it all up. Karl Enrico Weigelt wrote: > * Karl Waclawek wrote: > > > >>>Here's a new patch against cvs Makefile.in >> >>Applied to CVS. > > Okay. Great. > > Is the current release also fixed ? > > > cu From weigelt at metux.de Mon Dec 27 17:31:00 2004 From: weigelt at metux.de (Enrico Weigelt) Date: Mon Dec 27 17:31:04 2004 Subject: [Expat-discuss] Re: DESTDIR patches In-Reply-To: <41CF447B.1040209@waclawek.net> References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> <20041217174034.GE26310@nibiru.borg.metux.de> <41C320A3.4000607@waclawek.net> <20041226182937.GG25408@nibiru.borg.metux.de> <41CF447B.1040209@waclawek.net> Message-ID: <20041227163059.GA28890@nibiru.borg.metux.de> * Karl Waclawek wrote: > We never change releases, especially not ones > that have been out for a while, we rather > make new ones. > > All the changes in CVS will be part of Expat 2.0. > Since you will ask when it will be out: > This depends mostly on how much time we have > writing regression tests for the fixes > and packaging it all up. Okay, why not releasing 1.95.8.1 with my fix against 1.95.8 applied ? cu -- --------------------------------------------------------------------- Enrico Weigelt == metux IT service phone: +49 36207 519931 www: http://www.metux.de/ fax: +49 36207 519932 email: contact@metux.de cellphone: +49 174 7066481 --------------------------------------------------------------------- -- DSL ab 0 Euro. -- statische IP -- UUCP -- Hosting -- Webshops -- --------------------------------------------------------------------- From karl at waclawek.net Mon Dec 27 17:45:43 2004 From: karl at waclawek.net (Karl Waclawek) Date: Mon Dec 27 17:45:41 2004 Subject: [Expat-discuss] Re: DESTDIR patches In-Reply-To: <20041227163059.GA28890@nibiru.borg.metux.de> References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> <20041217174034.GE26310@nibiru.borg.metux.de> <41C320A3.4000607@waclawek.net> <20041226182937.GG25408@nibiru.borg.metux.de> <41CF447B.1040209@waclawek.net> <20041227163059.GA28890@nibiru.borg.metux.de> Message-ID: <41D03C37.70001@waclawek.net> Enrico Weigelt wrote: > * Karl Waclawek wrote: > > >>We never change releases, especially not ones >>that have been out for a while, we rather >>make new ones. >> >>All the changes in CVS will be part of Expat 2.0. >>Since you will ask when it will be out: >>This depends mostly on how much time we have >>writing regression tests for the fixes >>and packaging it all up. > > > Okay, why not releasing 1.95.8.1 with my fix against 1.95.8 applied ? Because a release is an effort, and there are other fixes waiting to get packaged too. We do plan to get a new release out as soon as we can - once Fred is back we will know more. Karl From ygblue at hotmail.com Tue Dec 28 19:40:49 2004 From: ygblue at hotmail.com (huang jianfeng) Date: Tue Dec 28 03:39:03 2004 Subject: [Expat-discuss] how to simply expat Message-ID: Hi, I'm using expat well, except it is a bit large. The xml file we used is = very simple and it doesn't contain any DTD related information, so I = want to cut DTD related code to get a simple version. I use expat for = our embedded system, and you know the memory is limited.=20 Any suggestions? Best Regards, CarlFrom karl at waclawek.net Tue Dec 28 04:25:00 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 28 04:24:58 2004 Subject: [Expat-discuss] how to simply expat In-Reply-To: References: Message-ID: <41D0D20C.1010801@waclawek.net> Try to undefine XML_DTD and XML_CONTEXT_BYTES. It should be documented. in reference.html. huang jianfeng wrote: > Hi, > I'm using expat well, except it is a bit large. The xml file we used is very simple and it doesn't contain any DTD related information, so I want to cut DTD related code to get a simple version. I use expat for our embedded system, and you know the memory is limited. > Any suggestions? > > Best Regards, > Carl > _______________________________________________ > Expat-discuss mailing list > Expat-discuss@libexpat.org > http://mail.libexpat.org/mailman/listinfo/expat-discuss > From ygblue at hotmail.com Tue Dec 28 22:28:29 2004 From: ygblue at hotmail.com (huang jianfeng) Date: Tue Dec 28 06:27:08 2004 Subject: [Expat-discuss] how to simply expat References: <41D0D20C.1010801@waclawek.net> Message-ID: Yes,I have found that XML_DTD and XML_CONTEXT_BYTES can meet the requirement. But I've also found it seems that XML_DTD doesn't include all the DTD related code. I suppose xmlrole.c is just for DTD, isn't it? I try to read the source code, but it's really hard to understand. Are there any docs about expat's implementation? Thanks. Carl ----- Karl Waclawek wrote: > Try to undefine XML_DTD and XML_CONTEXT_BYTES. > It should be documented. in reference.html. > > huang jianfeng wrote: > > Hi, > > I'm using expat well, except it is a bit large. The xml file we used is very simple and it doesn't contain any DTD related information, so I want to cut DTD related code to get a simple version. I use expat for our embedded system, and you know the memory is limited. > > Any suggestions? > > > > Best Regards, > > Carl > > _______________________________________________ > > Expat-discuss mailing list > > Expat-discuss@libexpat.org > > http://mail.libexpat.org/mailman/listinfo/expat-discuss > > > From karl at waclawek.net Tue Dec 28 08:54:37 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 28 08:54:34 2004 Subject: [Expat-discuss] how to simply expat In-Reply-To: References: <41D0D20C.1010801@waclawek.net> Message-ID: <41D1113D.9030701@waclawek.net> XML_DTD only affects some of the DTD related code. There is no implementation docs, so you have to read the code, which is not easy, I admit. Karl huang jianfeng wrote: > Yes,I have found that XML_DTD and XML_CONTEXT_BYTES can meet the > requirement. But I've also found it seems that XML_DTD doesn't include all > the DTD related code. I suppose xmlrole.c is just for DTD, isn't it? > > I try to read the source code, but it's really hard to understand. Are there > any docs about expat's implementation? > > Thanks. > > Carl > > ----- > Karl Waclawek wrote: > > > >>Try to undefine XML_DTD and XML_CONTEXT_BYTES. >>It should be documented. in reference.html. >> >>huang jianfeng wrote: >> >>>Hi, >>>I'm using expat well, except it is a bit large. The xml file we used is > > very simple and it doesn't contain any DTD related information, so I want to > cut DTD related code to get a simple version. I use expat for our embedded > system, and you know the memory is limited. > >>>Any suggestions? >>> >>>Best Regards, >>>Carl >>>_______________________________________________ >>>Expat-discuss mailing list >>>Expat-discuss@libexpat.org >>>http://mail.libexpat.org/mailman/listinfo/expat-discuss >>> >> > From magi at funambol.com Tue Dec 28 10:36:37 2004 From: magi at funambol.com (Marco Magistrali) Date: Tue Dec 28 10:39:04 2004 Subject: [Expat-discuss] Using expat for PDA device References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> <20041217174034.GE26310@nibiru.borg.metux.de> <41C320A3.4000607@waclawek.net><20041226182937.GG25408@nibiru.borg.metux.de> <41CF447B.1040209@waclawek.net> Message-ID: <012301c4ecc0$d14fa430$0400a8c0@Delicious> Hi people, I am evaluating to use expat to parse an xml message, that can be a big xml, to parse xml elements and create an object structure representing the xml structure. It should work on a PDA device like a PocketPC. I try the expat on a win32 and it seems it works fine with big xml file but it seems it use much memory that is not so good for an embedded system. My question are: - There is a way to use less memory? - there is an automatic way to create an object structure starting from an xml structure or some hints to make that? I think something like xmlToObject.... Thank you in advance -- Marco Magistrali magi@funambol.com Sync4j- the open source SyncML mobile application platform www.sync4j.org From msridh at yahoo.com Tue Dec 28 10:43:58 2004 From: msridh at yahoo.com (Sridhar M) Date: Tue Dec 28 10:44:01 2004 Subject: [Expat-discuss] Regarding Expat (Beginners Question) Message-ID: <20041228094358.57100.qmail@web60003.mail.yahoo.com> HI, This is Sridhar from Sweden. I am new to Expat and I need following help 1. Is Expat is freeware? 2. Where can I found Unix Version of Expat 3. I need a simple "C" program using Expat Currently I working on Sun Solaris Environment. Please help me in this regard. Merry Chirstmas and Happy New Year Thanks and Regards, Sridhar __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From msridh at yahoo.com Tue Dec 28 10:44:34 2004 From: msridh at yahoo.com (Sridhar M) Date: Tue Dec 28 10:44:37 2004 Subject: [Expat-discuss] Regarding Expat (Beginners Question) Message-ID: <20041228094434.73632.qmail@web60004.mail.yahoo.com> HI, This is Sridhar from Sweden. I am new to Expat and I need following help 1. Is Expat is freeware? 2. Where can I found Unix Version of Expat 3. I need a simple "C" program using Expat Currently I working on Sun Solaris Environment. Please help me in this regard. Merry Chirstmas and Happy New Year Thanks and Regards, Sridhar __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From magi at funambol.com Tue Dec 28 12:42:23 2004 From: magi at funambol.com (Marco Magistrali) Date: Tue Dec 28 12:44:15 2004 Subject: [Expat-discuss] Using expat for PDA device References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> <20041217174034.GE26310@nibiru.borg.metux.de> <41C320A3.4000607@waclawek.net><20041226182937.GG25408@nibiru.borg.metux.de><41CF447B.1040209@waclawek.net> <012301c4ecc0$d14fa430$0400a8c0@Delicious> Message-ID: <01a801c4ecd2$4ff414e0$0400a8c0@Delicious> Hi I'm trying to build expat with eVC++ 4.0 to create not a lib or a dll but an exe. I include the files xmlparse.c, xmlrole.c and xmltok.c in my project where there is a file example.c with the main function. I get many warning and errors like xmltok.c C:\Expat-1.95.8\Source\lib\xmltok.c(25) : warning C4005: 'IGNORE_SECTION_TOK_VTABLE' : macro redefinition C:\Expat-1.95.8\Source\lib\xmltok.c(23) : see previous definition of 'IGNORE_SECTION_TOK_VTABLE' C:\Expat-1.95.8\Source\lib\xmltok.c(206) : warning C4005: 'STANDARD_VTABLE' : macro redefinition C:\Expat-1.95.8\Source\lib\xmltok.c(202) : see previous definition of 'STANDARD_VTABLE' C:\Expat-1.95.8\Source\lib\xmltok.c(235) : warning C4005: 'MINBPC' : macro redefinition C:\Expat-1.95.8\Source\lib\xmltok.c(232) : see previous definition of 'MINBPC' C:\Expat-1.95.8\Source\lib\xmltok.c(250) : warning C4005: 'BYTE_TYPE' : macro redefinition ......... xmlparse.c C:\Expat-1.95.8\Source\lib\xmlparse.c(30) : warning C4005: 'XML_ENCODE_MAX' : macro redefinition C:\Expat-1.95.8\Source\lib\xmlparse.c(22) : see previous definition of 'XML_ENCODE_MAX' C:\Expat-1.95.8\Source\lib\xmlparse.c(31) : warning C4005: 'XmlConvert' : macro redefinition C:\Expat-1.95.8\Source\lib\xmlparse.c(23) : see previous definition of 'XmlConvert' C:\Expat-1.95.8\Source\lib\xmlparse.c(32) : warning C4005: 'XmlGetInternalEncoding' : macro redefinition C:\Expat-1.95.8\Source\lib\xmlparse.c(24) : see previous definition of 'XmlGetInternalEncoding' C:\Expat-1.95.8\Source\lib\xmlparse.c(33) : warning C4005: 'XmlGetInternalEncodingNS' : macro redefinition C:\Expat-1.95.8\Source\lib\xmlparse.c(25) : see previous definition of 'XmlGetInternalEncodingNS' C:\Expat-1.95.8\Source\lib\xmlparse.c(34) : warning C4005: 'XmlEncode' : macro redefinition C:\Expat-1.95.8\Source\lib\xmlparse.c(106) : warning C4005: 'CHAR_HASH' : macro redefinition C:\Expat-1.95.8\Source\lib\xmlparse.c(103) : see previous definition of 'CHAR_HASH' C:\Expat-1.95.8\Source\lib\xmlparse.c(6184) : fatal error C1010: unexpected end of file while looking for precompiled header directive I try to use some directive as XML_DTD, HAVE_EXPAT_CONFIG_H, XML_UNICODE_WCHAR_T, XML_UNICODE but with no success.. Has someone got any hints? Many thanks Regards -- Marco Magistrali magi@funambol.com Sync4j- the open source SyncML mobile application platform www.sync4j.org From karl at waclawek.net Tue Dec 28 16:52:22 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 28 16:52:19 2004 Subject: [Expat-discuss] Regarding Expat (Beginners Question) In-Reply-To: <20041228094434.73632.qmail@web60004.mail.yahoo.com> References: <20041228094434.73632.qmail@web60004.mail.yahoo.com> Message-ID: <41D18136.3090607@waclawek.net> Download Expat 1.95.8 from http://sf.net/projects/expat . It contains the license (BSD Open Source license), the source and binaries for Win32, Linux and other Unixes. Demo programs are included. Karl Sridhar M wrote: > HI, > > This is Sridhar from Sweden. I am new to Expat and I need following help > > 1. Is Expat is freeware? > 2. Where can I found Unix Version of Expat > 3. I need a simple "C" program using Expat > > Currently I working on Sun Solaris Environment. Please help me in this regard. > > Merry Chirstmas and Happy New Year > Thanks and Regards, > Sridhar > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > _______________________________________________ > Expat-discuss mailing list > Expat-discuss@libexpat.org > http://mail.libexpat.org/mailman/listinfo/expat-discuss > From karl at waclawek.net Tue Dec 28 16:53:46 2004 From: karl at waclawek.net (Karl Waclawek) Date: Tue Dec 28 16:53:44 2004 Subject: [Expat-discuss] Using expat for PDA device In-Reply-To: <012301c4ecc0$d14fa430$0400a8c0@Delicious> References: <41BF1AA2.8050604@waclawek.net> <20041214204716.GI29394@nibiru.borg.metux.de> <41BF530C.6040606@waclawek.net> <20041214211405.GA20532@nibiru.borg.metux.de> <41BF73EB.6080603@waclawek.net> <20041217174034.GE26310@nibiru.borg.metux.de> <41C320A3.4000607@waclawek.net><20041226182937.GG25408@nibiru.borg.metux.de> <41CF447B.1040209@waclawek.net> <012301c4ecc0$d14fa430$0400a8c0@Delicious> Message-ID: <41D1818A.6030505@waclawek.net> Marco Magistrali wrote: > Hi people, > I am evaluating to use expat to parse an xml message, that can be a big > xml, to parse xml elements and create an object structure representing the > xml structure. It should work on a PDA device like a PocketPC. > I try the expat on a win32 and it seems it works fine with big xml file but > it seems it use much memory that is not so good for an embedded system. > My question are: > > - There is a way to use less memory? Already answered in this thread.. > - there is an automatic way to create an object structure starting from an > xml structure or some hints to make that? I think something like > xmlToObject.... No, you have to do that yourself. Karl From odisseo at despammed.com Wed Dec 29 02:31:20 2004 From: odisseo at despammed.com (odisseo@despammed.com) Date: Wed Dec 29 02:47:56 2004 Subject: [Expat-discuss] How to handle text-data associated to a tag Message-ID: <200412290131.iBT1VKk05623@mailout.despammed.com> Ho can I handle the text-data in the example: 12 -9 232 I want to create a map < , , > so I need to know the value between start and end tag. The code for the start tag function could be static void startElement(void *userData, const char *name, const char **atts) { if (string(name) == "value") std::cout << "Var name" << atts[1] << std::endl; } now how can associate to the tag, the text value? From marco.forberg at gmx.net Wed Dec 29 10:20:48 2004 From: marco.forberg at gmx.net (Marco Forberg) Date: Wed Dec 29 10:10:20 2004 Subject: [Expat-discuss] How to handle text-data associated to a tag In-Reply-To: <200412290131.iBT1VKk05623@mailout.despammed.com> References: <200412290131.iBT1VKk05623@mailout.despammed.com> Message-ID: You will have to set a CharacterDataHandler in order to deal with text between the tags. The tag around the text is the last one that triggered the call to your StartElementHandlers. So you might have to store the variable-name to know where the value belongs. Using a flag could also be usefull to make sure you're parsing text inside the value-element. Am Tue, 28 Dec 2004 19:31:20 -0600 (CST) hat geschrieben: > Ho can I handle the text-data in the example: > 12 > -9 > 232 > I want to create a map > < , , > > so I need to know the value between start and end tag. > > The code for the start tag function could be > > static void startElement(void *userData, const char *name, const char > **atts) { > if (string(name) == "value") > std::cout << "Var name" << atts[1] << std::endl; > } > > now how can associate to the tag, the text value? > > > _______________________________________________ > Expat-discuss mailing list > Expat-discuss@libexpat.org > http://mail.libexpat.org/mailman/listinfo/expat-discuss > From magi at funambol.com Wed Dec 29 12:10:36 2004 From: magi at funambol.com (Marco Magistrali) Date: Wed Dec 29 12:12:29 2004 Subject: [Expat-discuss] Again on expat and pocket pc Message-ID: <03e801c4ed97$09b68e20$0400a8c0@Delicious> Hi list, I have again the problem to build expat for pocketpc 2003 with arm processor. I try to create a expat static lib with the for the arm processor and it was created successfully, it seems... I use the COMPILED_FROM_DSP directive only. The in my second project, I build my elements.c but I get some link error like Compiling... elements.c Linking... elements.obj : error LNK2019: unresolved external symbol __imp__XML_DefaultCurrent referenced in function _printcurrent elements.obj : error LNK2019: unresolved external symbol __imp__XML_SetDefaultHandler referenced in function _printcurrent elements.obj : error LNK2019: unresolved external symbol __imp__XML_ParserFree referenced in function _main elements.obj : error LNK2019: unresolved external symbol __imp__XML_GetCurrentLineNumber referenced in function _main elements.obj : error LNK2019: unresolved external symbol __imp__XML_ErrorString referenced in function _main elements.obj : error LNK2019: unresolved external symbol __imp__XML_GetErrorCode referenced in function _main elements.obj : error LNK2019: unresolved external symbol __imp__XML_Parse referenced in function _main elements.obj : error LNK2019: unresolved external symbol __imp__XML_SetCharacterDataHandler referenced in function _main elements.obj : error LNK2019: unresolved external symbol __imp__XML_SetElementHandler referenced in function _main elements.obj : error LNK2019: unresolved external symbol __imp__XML_UseParserAsHandlerArg referenced in function _main elements.obj : error LNK2019: unresolved external symbol __imp__XML_ParserCreate referenced in function _main corelibc.lib(pegwmain.obj) : error LNK2019: unresolved external symbol _WinMain referenced in function _WinMainCRTStartup emulatorDbg/program.exe : fatal error LNK1120: 12 unresolved externals Could someone help me please? Thanks a lot Marco From karl at waclawek.net Wed Dec 29 14:55:27 2004 From: karl at waclawek.net (Karl Waclawek) Date: Wed Dec 29 14:55:33 2004 Subject: [Expat-discuss] Again on expat and pocket pc In-Reply-To: <03e801c4ed97$09b68e20$0400a8c0@Delicious> References: <03e801c4ed97$09b68e20$0400a8c0@Delicious> Message-ID: <41D2B74F.4080305@waclawek.net> Marco Magistrali wrote: > Hi list, > I have again the problem to build expat for pocketpc 2003 with arm > processor. > I try to create a expat static lib with the for the arm processor and it was > created successfully, it seems... > I use the COMPILED_FROM_DSP directive only. There is a directive for building static libs. Don't remember the exact name, but it is documented in one of the text files included. Karl From magi at funambol.com Wed Dec 29 18:12:07 2004 From: magi at funambol.com (Marco Magistrali) Date: Wed Dec 29 18:15:21 2004 Subject: [Expat-discuss] Again on expat and pocket pc References: <03e801c4ed97$09b68e20$0400a8c0@Delicious> <41D2B74F.4080305@waclawek.net> Message-ID: <046001c4edc9$8a86dc80$0400a8c0@Delicious> Hi Karl, thank you for answer. I found the XML_STATIC directive to create a static expat lib. I use this togheter with COMPILED_FROM_DSP . But I get the same error... -- Marco Magistrali magi@funambol.com Sync4j- the open source SyncML mobile application platform www.sync4j.org ----- Original Message ----- From: "Karl Waclawek" To: "Marco Magistrali" Cc: Sent: Wednesday, December 29, 2004 2:55 PM Subject: Re: [Expat-discuss] Again on expat and pocket pc > Marco Magistrali wrote: > > Hi list, > > I have again the problem to build expat for pocketpc 2003 with arm > > processor. > > I try to create a expat static lib with the for the arm processor and it was > > created successfully, it seems... > > I use the COMPILED_FROM_DSP directive only. > > There is a directive for building static libs. > Don't remember the exact name, but it is documented in > one of the text files included. > > Karl > > From karl at waclawek.net Wed Dec 29 19:44:25 2004 From: karl at waclawek.net (Karl Waclawek) Date: Wed Dec 29 19:44:32 2004 Subject: [Expat-discuss] Again on expat and pocket pc In-Reply-To: <046001c4edc9$8a86dc80$0400a8c0@Delicious> References: <03e801c4ed97$09b68e20$0400a8c0@Delicious> <41D2B74F.4080305@waclawek.net> <046001c4edc9$8a86dc80$0400a8c0@Delicious> Message-ID: <41D2FB09.5050304@waclawek.net> Marco Magistrali wrote: > Hi Karl, > thank you for answer. > I found the XML_STATIC directive to create a static expat lib. I use this > togheter with COMPILED_FROM_DSP . > But I get the same error... What compiler are you using? Unfortunately I know nothing about building for PocketPC. COMPILED_FROM_DSP is meant for Windows/VC++. Karl From magi at funambol.com Wed Dec 29 21:03:21 2004 From: magi at funambol.com (Marco Magistrali) Date: Wed Dec 29 21:05:16 2004 Subject: [Expat-discuss] Again on expat and pocket pc References: <20041229195238.AA01430850@ns.arachne.cz> Message-ID: <04be01c4ede1$76a69670$0400a8c0@Delicious> Hi Mike, Thanks for your comments. The thing is more weird for me is that I can build succesfully the lib with eVC4 for the target ppc 2003 emulator and device. But I cannot link the library in an external project for every platform... I will try to look for the net again... Of course if you could retrieve any stuff about directives and so on.... ;) Thank you again Marco ----- Original Message ----- From: "Mike" To: "'Marco Magistrali'" Cc: Sent: Wednesday, December 29, 2004 8:52 PM Subject: RE: [Expat-discuss] Again on expat and pocket pc > Hmm weird. I compiled expat (1.95.7) year ago sucessfully > with eVC4. Unfortunatelly I haven't eVC nor sources anymore, > but AFAIK I used both directives, too. > Try to look for port to WinCE for an inspiration, I saw > it somewhere on the net (don't know which version, but it's > not important). Take a look into their .vcp. > > Mike > > P.S. Don't be afraid of memory consuming on a PDA with expat, > I never had any problem. It's efficient and quick as a wind ;-) > > > -----Original Message----- > > From: expat-discuss-bounces@libexpat.org > > [mailto:expat-discuss-bounces@libexpat.org] On Behalf Of > > Marco Magistrali > > Sent: Wednesday, December 29, 2004 6:12 PM > > To: Karl Waclawek > > Cc: expat-discuss@libexpat.org > > Subject: Re: [Expat-discuss] Again on expat and pocket pc > > > > Hi Karl, > > thank you for answer. > > I found the XML_STATIC directive to create a static expat > > lib. I use this > > togheter with COMPILED_FROM_DSP . > > But I get the same error... > > > > -- > > Marco Magistrali magi@funambol.com > > Sync4j- the open source SyncML mobile application platform > > www.sync4j.org > > > > ----- Original Message ----- > > From: "Karl Waclawek" > > To: "Marco Magistrali" > > Cc: > > Sent: Wednesday, December 29, 2004 2:55 PM > > Subject: Re: [Expat-discuss] Again on expat and pocket pc > > > > > > > Marco Magistrali wrote: > > > > Hi list, > > > > I have again the problem to build expat for pocketpc > > 2003 with arm > > > > processor. > > > > I try to create a expat static lib with the for the arm > > processor and it > > was > > > > created successfully, it seems... > > > > I use the COMPILED_FROM_DSP directive only. > > > > > > There is a directive for building static libs. > > > Don't remember the exact name, but it is documented in > > > one of the text files included. > > > > > > Karl > > > > > > > > > > > > _______________________________________________ > > Expat-discuss mailing list > > Expat-discuss@libexpat.org > > http://mail.libexpat.org/mailman/listinfo/expat-discuss > > > > > > > > From mike at mikeware.cz Wed Dec 29 20:52:34 2004 From: mike at mikeware.cz (Mike) Date: Thu Dec 30 23:02:28 2004 Subject: [Expat-discuss] Again on expat and pocket pc In-Reply-To: <046001c4edc9$8a86dc80$0400a8c0@Delicious> Message-ID: <20041229195238.AA01430850@ns.arachne.cz> Hmm weird. I compiled expat (1.95.7) year ago sucessfully with eVC4. Unfortunatelly I haven't eVC nor sources anymore, but AFAIK I used both directives, too. Try to look for port to WinCE for an inspiration, I saw it somewhere on the net (don't know which version, but it's not important). Take a look into their .vcp. Mike P.S. Don't be afraid of memory consuming on a PDA with expat, I never had any problem. It's efficient and quick as a wind ;-) > -----Original Message----- > From: expat-discuss-bounces@libexpat.org > [mailto:expat-discuss-bounces@libexpat.org] On Behalf Of > Marco Magistrali > Sent: Wednesday, December 29, 2004 6:12 PM > To: Karl Waclawek > Cc: expat-discuss@libexpat.org > Subject: Re: [Expat-discuss] Again on expat and pocket pc > > Hi Karl, > thank you for answer. > I found the XML_STATIC directive to create a static expat > lib. I use this > togheter with COMPILED_FROM_DSP . > But I get the same error... > > -- > Marco Magistrali magi@funambol.com > Sync4j- the open source SyncML mobile application platform > www.sync4j.org > > ----- Original Message ----- > From: "Karl Waclawek" > To: "Marco Magistrali" > Cc: > Sent: Wednesday, December 29, 2004 2:55 PM > Subject: Re: [Expat-discuss] Again on expat and pocket pc > > > > Marco Magistrali wrote: > > > Hi list, > > > I have again the problem to build expat for pocketpc > 2003 with arm > > > processor. > > > I try to create a expat static lib with the for the arm > processor and it > was > > > created successfully, it seems... > > > I use the COMPILED_FROM_DSP directive only. > > > > There is a directive for building static libs. > > Don't remember the exact name, but it is documented in > > one of the text files included. > > > > Karl > > > > > > > _______________________________________________ > Expat-discuss mailing list > Expat-discuss@libexpat.org > http://mail.libexpat.org/mailman/listinfo/expat-discuss > > > From gstein at lyra.org Fri Dec 31 14:27:59 2004 From: gstein at lyra.org (Greg Stein) Date: Fri Dec 31 14:23:46 2004 Subject: [Expat-discuss] RFC: unitool -> universal toolchain instead of autoshit In-Reply-To: <20041214202309.GG29394@nibiru.borg.metux.de> References: <20041214035852.GG28636@nibiru.borg.metux.de> <20041214083554.GC5913@lyra.org> <20041214202309.GG29394@nibiru.borg.metux.de> Message-ID: <20041231132758.GC12879@lyra.org> On Tue, Dec 14, 2004 at 09:23:09PM +0100, Enrico Weigelt wrote: > * Greg Stein wrote: > > Hi, > > > The autoconf/libtool scripts work just fine for Expat. We dropped > > automake a long time (the worst of the bunch). > > > > I'm not sure how a switch would benefit Expat or its users. > > Well, autoconf often enough does *not* work. > > For example it doesn't pass -E to grep when using extendet regex, > which breaks with slightly older grep (2.4.x). This ends up in > configure thinking you dont have gnu ld and so refuses to build > an shared library. I haven't seen this reported before. Neither here, nor in the firestorm that you raised over in Apache httpd. >... > And that's exactly where I want to do with unitool: This is fine. *Complete* your tool, and then it might be nice to contemplate. But from all that I've seen, it is incomplete -- you're still developing an initial release. Until it is done, then I don't see how we could even consider it for Expat. Cheers, -g -- Greg Stein, http://www.lyra.org/