From: Julio S. <jul...@gm...> - 2005-06-26 16:42:38
|
Hello, How are these methods and the associated data meant to be used? I have problems because of the result of get_document() being bad. It happens with several reports (IndivComplete, Timeline, ...), but I am not sure what is the triggering condition, seems to be related to the particular filter chosen: Everyone works, one single person works, ascendants fail and form my custom filters some work and some don't and I do not understand why. One sample message is: GRAMPS : 2.0.4-0.CVS20050619 LANG : es_ES.UTF-8 Python : 2.4.1 final GTK : 2.6.7 PyGTK : 2.6.0 OS : Fedora Core release 4 (Stentz) Traceback (most recent call last): File "/home/jsanchez/devel/gramps/cvs/gramps2-gramps20-ro/src/Report.py", line 1766, in report MyReport =3D report_class(dialog.db,dialog.person,dialog.options) File "./plugins/IndivComplete.py", line 76, in __init__ Report.Report.__init__(self,database,person,options_class) File "/home/jsanchez/devel/gramps/cvs/gramps2-gramps20-ro/src/Report.py", line 215, in __init__ self.doc.creator(creator) AttributeError: 'NoneType' object has no attribute 'creator' Here, self.doc was received from a call to get_document so it seems a value is not provided, but it is unclear to me what part of GRAMPS is responsible for this. Regards, Julio |
From: Alex R. <sh...@gr...> - 2005-07-02 03:01:18
|
Julio, Is this still the issue with 2.0.4? I think the failure with the filters leads to self.doc being None instead of a BaseDoc instance. Seems that you fixed some filter issues, so please let me know if this still needs attention. Thanks, Alex On 06/26/2005 11:42:30 AM, Julio Sanchez wrote: >=20 > How are these methods and the associated data meant to be used? I > have problems because of the result of get_document() being bad. It > happens with several reports (IndivComplete, Timeline, ...), but I am > not sure what is the triggering condition, seems to be related to the > particular filter chosen: Everyone works, one single person works, > ascendants fail and form my custom filters some work and some don't > and I do not understand why. >=20 > One sample message is: >=20 > GRAMPS : 2.0.4-0.CVS20050619 > LANG : es_ES.UTF-8 > Python : 2.4.1 final > GTK : 2.6.7 > PyGTK : 2.6.0 > OS : Fedora Core release 4 (Stentz) >=20 > Traceback (most recent call last): > File "/home/jsanchez/devel/gramps/cvs/gramps2-gramps20-ro/src/Report.py= ", > line 1766, in report > MyReport =3D report_class(dialog.db,dialog.person,dialog.options) > File "./plugins/IndivComplete.py", line 76, in __init__ > Report.Report.__init__(self,database,person,options_class) > File "/home/jsanchez/devel/gramps/cvs/gramps2-gramps20-ro/src/Report.py= ", > line 215, in __init__ > self.doc.creator(creator) > AttributeError: 'NoneType' object has no attribute 'creator' >=20 > Here, self.doc was received from a call to get_document so it seems a > value is not provided, but it is unclear to me what part of GRAMPS is > responsible for this. --=20 Alexander Roitman http://www.gramps-project.org |
From: Julio S. <jul...@gm...> - 2005-07-02 08:51:10
|
Alex, I cannot reproduce this anymore, it seems, so probably some of the other changes did this one in. What bugs me now are the ooo crashes. Perhaps it is triggered by improper files created by GRAMPS, so the fix in ooo, if it ever comes, may be from crashing to a dialog saying "broken file on input". No one else has seen ooo crash on input from GRAMPS recently? It appears many ooo versions hate on the test file I gave. All my output files now exhibit the same problem. I do not see any relevant change in GRAMPS, though, so I do not know what to think. Fact is I cannot produce any output right now. There was another thing pending when you left and that I don't think has been touched, see thread with subject "strange code for merging sources" or some such. Regards, Julio What kills=20 2005/7/2, Alex Roitman <sh...@gr...>: > Julio, >=20 > Is this still the issue with 2.0.4? I think the failure with the > filters leads to self.doc being None instead of a BaseDoc instance. > Seems that you fixed some filter issues, so please let me know if this > still needs attention. >=20 > Thanks, > Alex >=20 > On 06/26/2005 11:42:30 AM, Julio Sanchez wrote: > > > > How are these methods and the associated data meant to be used? I > > have problems because of the result of get_document() being bad. It > > happens with several reports (IndivComplete, Timeline, ...), but I am > > not sure what is the triggering condition, seems to be related to the > > particular filter chosen: Everyone works, one single person works, > > ascendants fail and form my custom filters some work and some don't > > and I do not understand why. > > > > One sample message is: > > > > GRAMPS : 2.0.4-0.CVS20050619 > > LANG : es_ES.UTF-8 > > Python : 2.4.1 final > > GTK : 2.6.7 > > PyGTK : 2.6.0 > > OS : Fedora Core release 4 (Stentz) > > > > Traceback (most recent call last): > > File "/home/jsanchez/devel/gramps/cvs/gramps2-gramps20-ro/src/Report.= py", > > line 1766, in report > > MyReport =3D report_class(dialog.db,dialog.person,dialog.options) > > File "./plugins/IndivComplete.py", line 76, in __init__ > > Report.Report.__init__(self,database,person,options_class) > > File "/home/jsanchez/devel/gramps/cvs/gramps2-gramps20-ro/src/Report.= py", > > line 215, in __init__ > > self.doc.creator(creator) > > AttributeError: 'NoneType' object has no attribute 'creator' > > > > Here, self.doc was received from a call to get_document so it seems a > > value is not provided, but it is unclear to me what part of GRAMPS is > > responsible for this. >=20 > -- > Alexander Roitman http://www.gramps-project.org >=20 >=20 > |
From: Don A. <don...@co...> - 2005-07-02 11:34:23
|
Julio, If you get broken OOo file, can you send it to me? I can dig into the generated XML and I might be able to figure out what is happening. Don On Sat, 2005-07-02 at 10:51 +0200, Julio Sanchez wrote: > Alex, > > I cannot reproduce this anymore, it seems, so probably some of the > other changes did this one in. > > What bugs me now are the ooo crashes. Perhaps it is triggered by > improper files created by GRAMPS, so the fix in ooo, if it ever comes, > may be from crashing to a dialog saying "broken file on input". No > one else has seen ooo crash on input from GRAMPS recently? It appears > many ooo versions hate on the test file I gave. All my output files > now exhibit the same problem. I do not see any relevant change in > GRAMPS, though, so I do not know what to think. Fact is I cannot > produce any output right now. > > There was another thing pending when you left and that I don't think > has been touched, see thread with subject "strange code for merging > sources" or some such. > > Regards, > > Julio > > What kills > > 2005/7/2, Alex Roitman <sh...@gr...>: > > Julio, > > > > Is this still the issue with 2.0.4? I think the failure with the > > filters leads to self.doc being None instead of a BaseDoc instance. > > Seems that you fixed some filter issues, so please let me know if this > > still needs attention. > > > > Thanks, > > Alex > > > > On 06/26/2005 11:42:30 AM, Julio Sanchez wrote: > > > > > > How are these methods and the associated data meant to be used? I > > > have problems because of the result of get_document() being bad. It > > > happens with several reports (IndivComplete, Timeline, ...), but I am > > > not sure what is the triggering condition, seems to be related to the > > > particular filter chosen: Everyone works, one single person works, > > > ascendants fail and form my custom filters some work and some don't > > > and I do not understand why. > > > > > > One sample message is: > > > > > > GRAMPS : 2.0.4-0.CVS20050619 > > > LANG : es_ES.UTF-8 > > > Python : 2.4.1 final > > > GTK : 2.6.7 > > > PyGTK : 2.6.0 > > > OS : Fedora Core release 4 (Stentz) > > > > > > Traceback (most recent call last): > > > File "/home/jsanchez/devel/gramps/cvs/gramps2-gramps20-ro/src/Report.py", > > > line 1766, in report > > > MyReport = report_class(dialog.db,dialog.person,dialog.options) > > > File "./plugins/IndivComplete.py", line 76, in __init__ > > > Report.Report.__init__(self,database,person,options_class) > > > File "/home/jsanchez/devel/gramps/cvs/gramps2-gramps20-ro/src/Report.py", > > > line 215, in __init__ > > > self.doc.creator(creator) > > > AttributeError: 'NoneType' object has no attribute 'creator' > > > > > > Here, self.doc was received from a call to get_document so it seems a > > > value is not provided, but it is unclear to me what part of GRAMPS is > > > responsible for this. > > > > -- > > Alexander Roitman http://www.gramps-project.org > > > > > > > > > ------------------------------------------------------- > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > from IBM. Find simple to follow Roadmaps, straightforward articles, > informative Webcasts and more! Get everything you need to get up to > speed, fast. http://ads.osdn.com/?ad_idt77&alloc_id492&op=click > _______________________________________________ > Gramps-devel mailing list > Gra...@li... > https://lists.sourceforge.net/lists/listinfo/gramps-devel > -- Don Allingham <don...@co...> |
From: Julio S. <jul...@gm...> - 2005-07-02 11:55:57
|
2005/7/2, Don Allingham <don...@co...>: > Julio, >=20 > If you get broken OOo file, can you send it to me? I can dig into the > generated XML and I might be able to figure out what is happening. You have it here: http://www.openoffice.org/nonav/issues/showattachment.cgi/27537/book.sxw It is an attachment to the bug report made by someone from Fedora when they reproduced my report. It is the file I sent them (I destroyed my original file by accident after attaching it to the Fedora bug report, sorry). The bug tracker itself is: http://qa.openoffice.org/issues/show_bug.cgi?id=3D51301 It would be great if you had a look at that file, I have extracted the pieces from it and I could not see anything obviously wrong but, then, I do not know the format so I just perused the XML. Please, ask for anything else you might need. I am going to be plugged to this keyboard the whole weekend. For real-life work, I am afraid :-( Thanks, Julio |
From: Alex R. <sh...@gr...> - 2005-07-02 18:58:25
|
Julio, On 07/02/2005 06:55:46 AM, Julio Sanchez wrote: >=20 > You have it here: >=20 > http://www.openoffice.org/nonav/issues/showattachment.cgi/27537/book.sxw >=20 > It is an attachment to the bug report made by someone from Fedora when > they reproduced my report. =20 It opens fine in my OOo, screenshot here: http://gramps-project.org/source/shots/shot.png My OOo is 1.1.4-3 from Debian unstable. Alex --=20 Alexander Roitman http://www.gramps-project.org |
From: Julio S. <jul...@gm...> - 2005-07-02 19:25:57
|
2005/7/2, Alex Roitman <sh...@gr...>: > Julio, >=20 >=20 > It opens fine in my OOo, screenshot here: > http://gramps-project.org/source/shots/shot.png >=20 > My OOo is 1.1.4-3 from Debian unstable. Good to know, it seems to affect OOo 2.x only. The file is broken. This I have seen. So it has to be fixed in GRAMPS, at least until it produces DTD-conforming files. Now, OOo 2.x is broken, it is unacceptable for any program to die this way, especially a program that might be fed data not built by its user. There may be a security hole hidden there. So they have to fix it until it does no die. Now, simple-minded cleanup of the XML does not keep that small file from breaking OOo. The question is: is the result of the cleanup a completely legal OOo file, or is it still broken? Who will need to run this last mile? Is the new OOo broken if it does not take the file, or has GRAMPS always produced broken output and the old OOo was more permissive? That, I do not know yet. Julio |
From: Alex R. <sh...@gr...> - 2005-07-02 20:29:54
|
On 07/02/2005 02:25:48 PM, Julio Sanchez wrote: >=20 > The file is broken. This I have seen. So it has to be fixed in > GRAMPS, at least until it produces DTD-conforming files. Agreed. However, as you say, the dtd-conforming file still crashes your newer OOo. We will fix dtd issues, but this does not seem to be the show stopper. > Now, OOo 2.x is broken, it is unacceptable for any program to die this > way, especially a program that might be fed data not built by its > user. There may be a security hole hidden there. So they have to fix > it until it does no die. Agreed. Whie they're at it, they may want to give a message as to what is wrong with the file. E.g. if you give OOo a badly-formed XML with mismatched tags, it will tell you line and column number where the error is detected. This would be very valuable for us, if indeed we create a broken file (besides DTD issues). > Now, simple-minded cleanup of the XML does not keep that small file > from breaking OOo. The question is: is the result of the cleanup a > completely legal OOo file, or is it still broken? Who will need to run > this last mile? Is the new OOo broken if it does not take the file, > or has GRAMPS always produced broken output and the old OOo was more > permissive? I think OOo people are best suited to answer this. If the file causes their parser to break, they could easily find out which call breaks it and why. Let's see what they do with the report. Alex --=20 Alexander Roitman http://www.gramps-project.org |
From: Julio S. <jul...@gm...> - 2005-07-04 22:19:57
Attachments:
docgen.patch
|
2005/7/2, Alex Roitman <sh...@gr...>: > On 07/02/2005 02:25:48 PM, Julio Sanchez wrote: > > > > The file is broken. This I have seen. So it has to be fixed in > > GRAMPS, at least until it produces DTD-conforming files. >=20 > Agreed. However, as you say, the dtd-conforming file still crashes > your newer OOo. We will fix dtd issues, but this does not seem to be > the show stopper. I attach a patch that makes the output DTD-compliant. In every case an attribute was wrong I have tried to guess which, among those allowed, seemed more likely. On a few cases, however, I have had to resort to removing attributes. Please review this patch. Regards, Julio |
From: Don A. <don...@co...> - 2005-07-03 01:53:01
|
This could very well be an OpenOffice.org 2.X issue. I know they changed to the OASIS format. Could there be an issue with this? Don On 07/02/2005 12:57:41 PM, Alex Roitman wrote: > Julio, > > On 07/02/2005 06:55:46 AM, Julio Sanchez wrote: > > > > You have it here: > > > > > http://www.openoffice.org/nonav/issues/showattachment.cgi/27537/book.sxw > > > > It is an attachment to the bug report made by someone from Fedora > when > > they reproduced my report. > > It opens fine in my OOo, screenshot here: > http://gramps-project.org/source/shots/shot.png > > My OOo is 1.1.4-3 from Debian unstable. > > Alex > > -- > Alexander Roitman http://www.gramps-project.org > > |
From: Don A. <don...@co...> - 2005-07-02 12:26:49
|
I'm on vacation this week, and traveling today, but I should be able to find some time to take a look at it. The first thing I will do is unzip the file and run "xmllint" on the XML files contained within. This usually identifies the problem. Don On Sat, 2005-07-02 at 13:55 +0200, Julio Sanchez wrote: > 2005/7/2, Don Allingham <don...@co...>: > > Julio, > > > > If you get broken OOo file, can you send it to me? I can dig into the > > generated XML and I might be able to figure out what is happening. > > You have it here: > > http://www.openoffice.org/nonav/issues/showattachment.cgi/27537/book.sxw > > It is an attachment to the bug report made by someone from Fedora when > they reproduced my report. It is the file I sent them (I destroyed my > original file by accident after attaching it to the Fedora bug report, > sorry). The bug tracker itself is: > > http://qa.openoffice.org/issues/show_bug.cgi?id=51301 > > It would be great if you had a look at that file, I have extracted the > pieces from it and I could not see anything obviously wrong but, then, > I do not know the format so I just perused the XML. Please, ask for > anything else you might need. I am going to be plugged to this > keyboard the whole weekend. For real-life work, I am afraid :-( > > Thanks, > > Julio > -- Don Allingham <don...@co...> |
From: Julio S. <jul...@gm...> - 2005-07-02 14:32:26
|
2005/7/2, Don Allingham <don...@co...>: > I'm on vacation this week, and traveling today, but I should be able to > find some time to take a look at it. >=20 > The first thing I will do is unzip the file and run "xmllint" on the XML > files contained within. This usually identifies the problem. Good hint. Let's try. I run it like this: xmllint --noout --noent --load-trace --dtdvalid /usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/office.dtd --valid content.xml and I get lots of errors. Are they new or have they always been there? I get this: Loaded URL=3D"content.xml" ID=3D"(null)" content.xml:2: validity error : Validation failed: no DTD found ! ipt=3D"http://openoffice.org/2000/script" office:class=3D"text" office:vers= ion=3D"0.9" = ^ Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/offic= e.dtd" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/dtype= s.mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/nmspa= ce.mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/defs.= mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/offic= e.mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/style= .mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/meta.= mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/scrip= t.mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/drawi= ng.mod" ID=3D"(null)" /usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/drawing.mod:263: validity warning : Attribute extrusion-origin-x of element draw:enhanced-geometry: already defined <!ATTLIST draw:enhanced-geometry draw:extrusion-origin-x %percentage; "0"> ^ Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/text.= mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/table= .mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/chart= .mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/datas= tyl.mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/form.= mod" ID=3D"(null)" Loaded URL=3D"/usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/setti= ngs.mod" ID=3D"(null)" content.xml:16: element properties: validity error : No declaration for attribute color of element properties content.xml:22: element properties: validity error : No declaration for attribute color of element properties content.xml:28: element properties: validity error : No declaration for attribute color of element properties content.xml:34: element properties: validity error : No declaration for attribute color of element properties content.xml:40: element properties: validity error : No declaration for attribute color of element properties content.xml:46: element properties: validity error : No declaration for attribute color of element properties content.xml:52: element properties: validity error : No declaration for attribute color of element properties content.xml:84: element style: validity error : No declaration for attribute parent-name of element style content.xml:84: element properties: validity error : No declaration for attribute numer-wrapped-paragraphs of element properties content.xml:85: element style: validity error : No declaration for attribute parent-name of element style content.xml:85: element properties: validity error : No declaration for attribute numer-wrapped-paragraphs of element properties content.xml:86: element style: validity error : No declaration for attribute parent-name of element style content.xml:87: element style: validity error : No declaration for attribute parent-name of element style content.xml:99: element line: validity error : No declaration for attribute style of element line content.xml:103: element line: validity error : No declaration for attribute style of element line content.xml:107: element line: validity error : No declaration for attribute style of element line content.xml:111: element line: validity error : No declaration for attribute style of element line content.xml:115: element line: validity error : No declaration for attribute style of element line content.xml:119: element line: validity error : No declaration for attribute style of element line Document content.xml does not validate against /usr/lib/openoffice.org2.0/share/dtd/officedocument/1_0/office.dtd I have checked a few in the DTD's and at least some of them seem to me genuine errors. For instance, style:parent-name should be style:parent-style-name every time. Others, however, I am not so sure. What should be used instead of "draw:color"? According to the DTD, it seems they can only be attributes to draw:hatch and presentation:dim, Some other color attribute is meant, but I am not sure which. Regards, Julio |
From: Julio S. <jul...@gm...> - 2005-07-02 15:16:19
|
2005/7/2, Julio Sanchez <jul...@gm...>: > 2005/7/2, Don Allingham <don...@co...>: > > I'm on vacation this week, and traveling today, but I should be able to > > find some time to take a look at it. > > > > The first thing I will do is unzip the file and run "xmllint" on the XM= L > > files contained within. This usually identifies the problem. >=20 > Good hint. Let's try. Well. I think I cleared all problems found by xmllint by hand, but ooo still crashes. The problem is at some other level. I'l try to make room next week to commit changes to ensure that xmllint is happy with the output. Regards, Julio |