Hi, I have been using "degenerate" inlinetasks with empty titles but many tags for implementing a kind of coding scheme for coding texts for qualitative data analysis. Like this: ----- Some text that I want to tag (The inlinetask in my scheme refers to the paragraph above it) *************** :tag1:tag2:tag3: Other text (no inlinetask-END, mostly) ----- Building org from the master branch, I recently noticed a changed behaviour in that these tags as are not parsed as tags but instead as the title, meaning my exports don't work as expected (and possibly other things, but searching for tags etc.doesn't seem to be affected. Those functions don't use org-element perhaps?). As far as I could see, this comes from the changes in commit 98ee73: org-element: Avoid `org-element-parse-secondary-string', where tags are matched with the regexp: (org-re "[ \t]+\\(:[[:alnum:]_@#%:]+:\\)[ \t]*$") which needs whitespace "after" the non-existent title. I haven't checked all the different changes going on in org-element though. I don't know if this changed behaviour is intended. Otherwise I guess it's a bug. Cheers,
Hello, Anders Johansson <mejlaandersj@gmail.com> writes: > I have been using "degenerate" inlinetasks with empty titles but many > tags for implementing a kind of coding scheme for coding texts for > qualitative data analysis. Like this: > ----- > > Some text that I want to tag (The inlinetask in my scheme refers to > the paragraph above it) > *************** :tag1:tag2:tag3: > > Other text (no inlinetask-END, mostly) > ----- > > Building org from the master branch, I recently noticed a changed > behaviour in that these tags as are not parsed as tags but instead as > the title, meaning my exports don't work as expected (and possibly > other things, but searching for tags etc.doesn't seem to be affected. > Those functions don't use org-element perhaps?). Indeed. > As far as I could see, this comes from the changes in commit > > 98ee73: org-element: Avoid `org-element-parse-secondary-string', > > where tags are matched with the regexp: > (org-re "[ \t]+\\(:[[:alnum:]_@#%:]+:\\)[ \t]*$") > > which needs whitespace "after" the non-existent title. > > I haven't checked all the different changes going on in org-element though. > > I don't know if this changed behaviour is intended. Otherwise I guess > it's a bug. Empty headings are a pathological case. What if I want to write *************** :title: ? Your interpretation prevents that. Of course, Org is expected to be consistent. However I'm not convinced supporting empty headlines with tags is a good thing. Regards, -- Nicolas Goaziou
Den 2015-04-10 18:24, Nicolas Goaziou skrev:
> Hello,
>
> Anders Johansson <mejlaandersj@gmail.com> writes:
>
>> I have been using "degenerate" inlinetasks with empty titles but many
>> tags for implementing a kind of coding scheme for coding texts for
>> qualitative data analysis. Like this:
>> -----
>>
>> Some text that I want to tag (The inlinetask in my scheme refers to
>> the paragraph above it)
>> *************** :tag1:tag2:tag3:
>>
>> Other text (no inlinetask-END, mostly)
>> -----
>>
>> Building org from the master branch, I recently noticed a changed
>> behaviour in that these tags as are not parsed as tags but instead as
>> the title, meaning my exports don't work as expected (and possibly
>> other things, but searching for tags etc.doesn't seem to be affected.
>> Those functions don't use org-element perhaps?).
>
> Indeed.
>
>> As far as I could see, this comes from the changes in commit
>>
>> 98ee73: org-element: Avoid `org-element-parse-secondary-string',
>>
>> where tags are matched with the regexp:
>> (org-re "[ \t]+\\(:[[:alnum:]_@#%:]+:\\)[ \t]*$")
>>
>> which needs whitespace "after" the non-existent title.
>>
>> I haven't checked all the different changes going on in org-element though.
>>
>> I don't know if this changed behaviour is intended. Otherwise I guess
>> it's a bug.
>
> Empty headings are a pathological case. What if I want to write
>
> *************** :title:
>
> ? Your interpretation prevents that.
>
> Of course, Org is expected to be consistent. However I'm not convinced
> supporting empty headlines with tags is a good thing.
>
> Regards,
>
Great, that sounds reasonable.
I guess I'll change my practice then.
Cheers,