BUG: export options properties drawer position and planning dates

BUG: export options properties drawer position and planning dates

[Michael Dauer]

[-- Attachment #1: Type: text/plain, Size: 1375 bytes --]

Hi,

There seems to be a bug in Org mode version 9.4.6 (9.4.6-gcf30f7:
EXPORT_OPTIONS (at least p for with-planning) is only respected if there is
no planning date placed above the properties drawer.

>>>
* TODO export options
:PROPERTIES:
:EXPORT_OPTIONS: p:nil
:END:
SCHEDULED: <2021-06-08 Di.>
** TODO l1
:PROPERTIES:
:EXPORT_OPTIONS: p:t
:END:
SCHEDULED: <2021-06-08 Di.>
*** TODO l2
SCHEDULED: <2021-06-08 Di.>
**** TODO l3
:PROPERTIES:
:EXPORT_OPTIONS: p:nil
:END:
SCHEDULED: <2021-06-08 Di.>
***** TODO l4
SCHEDULED: <2021-06-08 Di.>
<<<

produces the somehow expected behavior:
>>>
SCHEDULED: <2021-06-08>


TODO l1
=======

  SCHEDULED: <2021-06-08>


TODO l2
~~~~~~~

TODO l3
-------

  SCHEDULED: <2021-06-08>


* TODO l4
<<<

But it is syntactically incorrect since the planning dates have to be the
first line below the heading. For the following syntactically correct
snippet the export_options are ignored:
>>>
* TODO export options
SCHEDULED: <2021-06-08 Di.>
:PROPERTIES:
:EXPORT_OPTIONS: p:nil
:END:
** TODO l1
SCHEDULED: <2021-06-08 Di.>
:PROPERTIES:
:EXPORT_OPTIONS: p:t
:END:
*** TODO l2
SCHEDULED: <2021-06-08 Di.>
**** TODO l3
SCHEDULED: <2021-06-08 Di.>
:PROPERTIES:
:EXPORT_OPTIONS: p:nil
:END:
***** TODO l4
SCHEDULED: <2021-06-08 Di.>
<<<

>>>
TODO l1
=======

TODO l2
~~~~~~~

TODO l3
-------

* TODO l4
<<<

Same behavior I see with HTML export.

[-- Attachment #2: Type: text/html, Size: 2029 bytes --]

Re: BUG: export options properties drawer position and planning dates

[Nicolas Goaziou]

Hello,

Michael Dauer <mick.dauer@gmail.com> writes:

> There seems to be a bug in Org mode version 9.4.6 (9.4.6-gcf30f7:
> EXPORT_OPTIONS (at least p for with-planning) is only respected if there is
> no planning date placed above the properties drawer.
>
>>>>
> * TODO export options
> :PROPERTIES:
> :EXPORT_OPTIONS: p:nil
> :END:
> SCHEDULED: <2021-06-08 Di.>
> ** TODO l1
> :PROPERTIES:
> :EXPORT_OPTIONS: p:t
> :END:
> SCHEDULED: <2021-06-08 Di.>
> *** TODO l2
> SCHEDULED: <2021-06-08 Di.>
> **** TODO l3
> :PROPERTIES:
> :EXPORT_OPTIONS: p:nil
> :END:
> SCHEDULED: <2021-06-08 Di.>
> ***** TODO l4
> SCHEDULED: <2021-06-08 Di.>
> <<<
>
> produces the somehow expected behavior:
>>>>
> SCHEDULED: <2021-06-08>
>
>
> TODO l1
> =======
>
>   SCHEDULED: <2021-06-08>
>
>
> TODO l2
> ~~~~~~~
>
> TODO l3
> -------
>
>   SCHEDULED: <2021-06-08>
>
>
> * TODO l4
> <<<
>
> But it is syntactically incorrect since the planning dates have to be the
> first line below the heading.

This is correct, because the SCHEDULED lines are exported as paragraphs,
which is what they are.

> For the following syntactically correct
> snippet the export_options are ignored:
>>>>
> * TODO export options
> SCHEDULED: <2021-06-08 Di.>
> :PROPERTIES:
> :EXPORT_OPTIONS: p:nil
> :END:
> ** TODO l1
> SCHEDULED: <2021-06-08 Di.>
> :PROPERTIES:
> :EXPORT_OPTIONS: p:t
> :END:
> *** TODO l2
> SCHEDULED: <2021-06-08 Di.>
> **** TODO l3
> SCHEDULED: <2021-06-08 Di.>
> :PROPERTIES:
> :EXPORT_OPTIONS: p:nil
> :END:
> ***** TODO l4
> SCHEDULED: <2021-06-08 Di.>
> <<<

>>>>
> TODO l1
> =======
>
> TODO l2
> ~~~~~~~
>
> TODO l3
> -------
>
> * TODO l4
> <<<
>
> Same behavior I see with HTML export.

Did you use subtree export? Where was the point when you exported it?

For example, if I use a subtree export in the first section above, I get
no planning line, you if I subtree-export from the second section, i.e.,
"l1", all subsequent planning lines appear.

IOW, I cannot reproduce your issue.

Regards,
-- 
Nicolas Goaziou

Re: BUG: export options properties drawer position and planning dates

[Michael Dauer]

[-- Attachment #1: Type: text/plain, Size: 3132 bytes --]

Hi Nicolas,

I would understand that the export would take the export settings of the
current heading to control the export of the complete subtree. This would
be the case for the second example, where the scheduled date is (correctly)
placed above the properties drawer.

BUT:
1. The much better logic would be that each node determines e.g. the
with-planning by its own (or inherited) properties.
2. This actually works when the scheduled date is (incorrectly) placed
below the drawer. It is not just treated as the first paragraph, but
omitted when the with-planning property of its node is nil, while normal
text would be exported.

I did not check the code. I assume that the better logic was actually
implemented with the mistake of assuming an incorrect order of planning
line and properties drawer.

Regards,
Michael

Am Di., 8. Juni 2021 um 22:22 Uhr schrieb Nicolas Goaziou <
mail@nicolasgoaziou.fr>:

> Hello,
>
> Michael Dauer <mick.dauer@gmail.com> writes:
>
> > There seems to be a bug in Org mode version 9.4.6 (9.4.6-gcf30f7:
> > EXPORT_OPTIONS (at least p for with-planning) is only respected if there
> is
> > no planning date placed above the properties drawer.
> >
> >>>>
> > * TODO export options
> > :PROPERTIES:
> > :EXPORT_OPTIONS: p:nil
> > :END:
> > SCHEDULED: <2021-06-08 Di.>
> > ** TODO l1
> > :PROPERTIES:
> > :EXPORT_OPTIONS: p:t
> > :END:
> > SCHEDULED: <2021-06-08 Di.>
> > *** TODO l2
> > SCHEDULED: <2021-06-08 Di.>
> > **** TODO l3
> > :PROPERTIES:
> > :EXPORT_OPTIONS: p:nil
> > :END:
> > SCHEDULED: <2021-06-08 Di.>
> > ***** TODO l4
> > SCHEDULED: <2021-06-08 Di.>
> > <<<
> >
> > produces the somehow expected behavior:
> >>>>
> > SCHEDULED: <2021-06-08>
> >
> >
> > TODO l1
> > =======
> >
> >   SCHEDULED: <2021-06-08>
> >
> >
> > TODO l2
> > ~~~~~~~
> >
> > TODO l3
> > -------
> >
> >   SCHEDULED: <2021-06-08>
> >
> >
> > * TODO l4
> > <<<
> >
> > But it is syntactically incorrect since the planning dates have to be the
> > first line below the heading.
>
> This is correct, because the SCHEDULED lines are exported as paragraphs,
> which is what they are.
>
> > For the following syntactically correct
> > snippet the export_options are ignored:
> >>>>
> > * TODO export options
> > SCHEDULED: <2021-06-08 Di.>
> > :PROPERTIES:
> > :EXPORT_OPTIONS: p:nil
> > :END:
> > ** TODO l1
> > SCHEDULED: <2021-06-08 Di.>
> > :PROPERTIES:
> > :EXPORT_OPTIONS: p:t
> > :END:
> > *** TODO l2
> > SCHEDULED: <2021-06-08 Di.>
> > **** TODO l3
> > SCHEDULED: <2021-06-08 Di.>
> > :PROPERTIES:
> > :EXPORT_OPTIONS: p:nil
> > :END:
> > ***** TODO l4
> > SCHEDULED: <2021-06-08 Di.>
> > <<<
>
> >>>>
> > TODO l1
> > =======
> >
> > TODO l2
> > ~~~~~~~
> >
> > TODO l3
> > -------
> >
> > * TODO l4
> > <<<
> >
> > Same behavior I see with HTML export.
>
> Did you use subtree export? Where was the point when you exported it?
>
> For example, if I use a subtree export in the first section above, I get
> no planning line, you if I subtree-export from the second section, i.e.,
> "l1", all subsequent planning lines appear.
>
> IOW, I cannot reproduce your issue.
>
> Regards,
> --
> Nicolas Goaziou
>

[-- Attachment #2: Type: text/html, Size: 4365 bytes --]

Re: BUG: export options properties drawer position and planning dates

[Nicolas Goaziou]

Hello,

Michael Dauer <mick.dauer@gmail.com> writes:

> I would understand that the export would take the export settings of the
> current heading to control the export of the complete subtree.

That's correct.

> 1. The much better logic would be that each node determines e.g. the
> with-planning by its own (or inherited) properties.

This is not how it is implemented. Export options are per export
process, not per node. Besides, the above would not make sense for
one-off items, like title:nil.

I guess the much better logic would first need to distinguish global
from local export options. But I don't think this is worth the trouble.

> 2. This actually works when the scheduled date is (incorrectly) placed
> below the drawer. It is not just treated as the first paragraph, but
> omitted when the with-planning property of its node is nil, while normal
> text would be exported.

Would you mind providing an ECM for it? I'm not sure what example you're
referring to.

Regards,
-- 
Nicolas Goaziou

Re: BUG: export options properties drawer position and planning dates

[Michael Dauer]

[-- Attachment #1: Type: text/plain, Size: 1774 bytes --]

>>>
* TODO export options
:PROPERTIES:
:EXPORT_OPTIONS: p:nil
:END:
SCHEDULED: <2021-06-08 Di.>
normal text
** TODO l1
:PROPERTIES:
:EXPORT_OPTIONS: p:t
:END:
SCHEDULED: <2021-06-08 Di.>
normal text
*** TODO l2
SCHEDULED: <2021-06-08 Di.>
normal text
**** TODO l3
:PROPERTIES:
:EXPORT_OPTIONS: p:nil
:END:
SCHEDULED: <2021-06-08 Di.>
normal text
***** TODO l4
SCHEDULED: <2021-06-08 Di.>
normal text
<<<

>>>
SCHEDULED: <2021-06-08>
normal text


TODO l1
=======

  SCHEDULED: <2021-06-08>
  normal text


TODO l2
~~~~~~~

  normal text


TODO l3
-------

  SCHEDULED: <2021-06-08>
  normal text


* TODO l4

  normal text
<<<

Am Di., 15. Juni 2021 um 08:07 Uhr schrieb Nicolas Goaziou <
mail@nicolasgoaziou.fr>:

> Hello,
>
> Michael Dauer <mick.dauer@gmail.com> writes:
>
> > I would understand that the export would take the export settings of the
> > current heading to control the export of the complete subtree.
>
> That's correct.
>
> > 1. The much better logic would be that each node determines e.g. the
> > with-planning by its own (or inherited) properties.
>
> This is not how it is implemented. Export options are per export
> process, not per node. Besides, the above would not make sense for
> one-off items, like title:nil.
>
> I guess the much better logic would first need to distinguish global
> from local export options. But I don't think this is worth the trouble.
>
> > 2. This actually works when the scheduled date is (incorrectly) placed
> > below the drawer. It is not just treated as the first paragraph, but
> > omitted when the with-planning property of its node is nil, while normal
> > text would be exported.
>
> Would you mind providing an ECM for it? I'm not sure what example you're
> referring to.
>
> Regards,
> --
> Nicolas Goaziou
>

[-- Attachment #2: Type: text/html, Size: 2567 bytes --]

Re: BUG: export options properties drawer position and planning dates

[Nicolas Goaziou]

Hello,

Michael Dauer <mick.dauer@gmail.com> writes:

>>>>

[...]

>>>>

I think we are mis-communicating. I saw the document you're re-posting.
My question, however, is about the specific part demonstrating the
following:

>> > 2. This actually works when the scheduled date is (incorrectly) placed
>> > below the drawer. It is not just treated as the first paragraph, but
>> > omitted when the with-planning property of its node is nil, while normal
>> > text would be exported.

I'm trying to reproduce this, so I need minimal example, with
instruction.

Regards,
-- 
Nicolas Goaziou

Re: BUG: export options properties drawer position and planning dates

[Michael Dauer]

[-- Attachment #1: Type: text/plain, Size: 969 bytes --]

Yes, you're right. The problem was with my sample. Forget about my 2.

Any ideas how I can specify the inclusion of planning dates specifically on
sub-tree level (not just once for the whole export) without custom exporter
or export filter?

thx
m


Am Di., 15. Juni 2021 um 17:05 Uhr schrieb Nicolas Goaziou <
mail@nicolasgoaziou.fr>:

> Hello,
>
> Michael Dauer <mick.dauer@gmail.com> writes:
>
> >>>>
>
> [...]
>
> >>>>
>
> I think we are mis-communicating. I saw the document you're re-posting.
> My question, however, is about the specific part demonstrating the
> following:
>
> >> > 2. This actually works when the scheduled date is (incorrectly) placed
> >> > below the drawer. It is not just treated as the first paragraph, but
> >> > omitted when the with-planning property of its node is nil, while
> normal
> >> > text would be exported.
>
> I'm trying to reproduce this, so I need minimal example, with
> instruction.
>
> Regards,
> --
> Nicolas Goaziou
>

[-- Attachment #2: Type: text/html, Size: 1540 bytes --]

Re: BUG: export options properties drawer position and planning dates

[Nicolas Goaziou]

Hello,

Michael Dauer <mick.dauer@gmail.com> writes:

> Yes, you're right. The problem was with my sample. Forget about my 2.
>
> Any ideas how I can specify the inclusion of planning dates specifically on
> sub-tree level (not just once for the whole export) without custom exporter
> or export filter?

I think you cannot without using those tools.

Regards,
-- 
Nicolas Goaziou