[NTG-context] [entry not flushed] problem since August 17 update

[NTG-context] [entry not flushed] problem since August 17 update

[Rik Kabel]

[-- Attachment #1.1: Type: text/plain, Size: 629 bytes --]

Hello list,

Since the August 17 update, and continuing to the current August 21 
release, my indexes show

    [entry not flushed],

on many entries where the page number should be suppressed and an 
endnote number, provided with the register command, displayed. This was 
not an issue with the July  21 release and earlier.

The indexing code for this document is a bit complex, and I have not yet 
been able to isolate the issue further.

Has anybody else seen similar problems?

Is there something in particular I should look for?

I am trying to put together an MWE, but have not got it yet.

-- 
Rik

[-- Attachment #1.2: Type: text/html, Size: 961 bytes --]

[-- Attachment #2: Type: text/plain, Size: 511 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________

[NTG-context] Re: [entry not flushed] problem since August 17 update

[Henning Hraban Ramm]

Am 08.09.25 um 02:40 schrieb Rik Kabel:
> Hello list,
> 
> Since the August 17 update, and continuing to the current August 21 
> release, my indexes show
> 
>     [entry not flushed],
> 
> on many entries where the page number should be suppressed and an 
> endnote number, provided with the register command, displayed. This was 
> not an issue with the July  21 release and earlier.
> 
> The indexing code for this document is a bit complex, and I have not yet 
> been able to isolate the issue further.
> 
> Has anybody else seen similar problems?
> 
> Is there something in particular I should look for?
> 
> I am trying to put together an MWE, but have not got it yet.

Maybe index entries in footnotes (in tables, floats, postponed pages) 
that disappear?

Hraban
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________

[NTG-context] Re: [entry not flushed] problem since August 17 update

[Rik Kabel]

[-- Attachment #1.1: Type: text/plain, Size: 2045 bytes --]


On 2025-09-07 20:40, Rik Kabel wrote:
>
> Hello list,
>
> Since the August 17 update, and continuing to the current August 21 
> release, my indexes show
>
>     [entry not flushed],
>
> on many entries where the page number should be suppressed and an 
> endnote number, provided with the register command, displayed. This 
> was not an issue with the July  21 release and earlier.
>
> The indexing code for this document is a bit complex, and I have not 
> yet been able to isolate the issue further.
>
> Has anybody else seen similar problems?
>
> Is there something in particular I should look for?
>
> I am trying to put together an MWE, but have not got it yet.
>
Okay, I have created an MWE. The issues appears to be tagging when 
interaction is active -- disabling either or both of the first two lines 
below resolves the error, but that should not be necessary.

    \setupinteraction[state=start]
    \setuptagging[state=start]

    \starttexdefinition IndexPageCommand #1
       \doifsomethingelse{\currentregisterpageuserdata{note}}
         {n\currentregisterpageuserdata{note}}
         #1
       \doifsomething{\currentregisterpageuserdata{id}}
         {\in[\currentregisterpageuserdata{id}]}
    \stoptexdefinition

    \defineregister[Pindex]
    \setupregister [Pindex][pagecommand=\IndexPageCommand]

    \starttexdefinition tolerant protected unexpanded PIndex
    [#1]#:#2#*#=#*#=
       \doifelse{#3}{}
         {\setregisterentry[Pindex][keys={#1},entries={#2}]}
         {\setregisterentry[Pindex][keys={#1},entries={#2}][#3]}
    \stoptexdefinition

    \starttext
         \PIndex{OK, Entry}
       Here is some text.
    \startendnote[en:lentz]
         \PIndex{Bad, Entry1}
         \PIndex{Bad, Entry2}[note={\in[en:lentz]}]%
       Here is an endnote.
    \stopendnote
    \page
       \placenotes[endnote]
    \page
       \placeregister[Pindex]
    \stoptext

Please let me know if this is user error or a bug.

-- 
Rik

[-- Attachment #1.2: Type: text/html, Size: 3102 bytes --]

[-- Attachment #2: Type: text/plain, Size: 511 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________

[NTG-context] Re: [entry not flushed] problem since August 17 update

[Rik Kabel]

[-- Attachment #1.1: Type: text/plain, Size: 1255 bytes --]

Slightly simpler MWE:

    \setuppapersize[A6,landscape]

    \setupinteraction[state=start]
    \setuptagging[state=start]

    \starttexdefinition IndexPageCommand #1
       \doifsomethingelse{\currentregisterpageuserdata{note}}
         {n\currentregisterpageuserdata{note}}
         #1
       \doifsomething{\currentregisterpageuserdata{id}}
         {\in[\currentregisterpageuserdata{id}]}
    \stoptexdefinition

    \defineregister[Pindex]
    \setupregister [Pindex][pagecommand=\IndexPageCommand]

    \starttexdefinition tolerant protected unexpanded PIndex
    [#1]#:#2#*#=#*#=
       \doifelse{#3}{}
         {\setregisterentry[Pindex][keys={#1},entries={#2}]}
         {\setregisterentry[Pindex][keys={#1},entries={#2}][#3]}
    \stoptexdefinition

    \starttext
         \PIndex{OK, Entry}
         \PIndex{OK, Entry1}[note={ote}]%
       Here is some text.
    \startendnote
         \PIndex{Bad, Entry2}
         \PIndex{Bad, Entry3}[note={ote}]%
       Here is an endnote.
    \stopendnote
    \page
       \placenotes[endnote]
    \page
       \placeregister[Pindex]
    \stoptext

Fails with tagging and/or interaction enabled. Succeeds with both disabled.

-- 
Rik

[-- Attachment #1.2: Type: text/html, Size: 1885 bytes --]

[-- Attachment #2: Type: text/plain, Size: 511 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________

[NTG-context] Re: [entry not flushed] problem since August 17 update

[Hans Hagen via ntg-context]

On 9/12/2025 5:33 PM, Rik Kabel wrote:
> 
> On 2025-09-07 20:40, Rik Kabel wrote:
>>
>> Hello list,
>>
>> Since the August 17 update, and continuing to the current August 21 
>> release, my indexes show
>>
>>     [entry not flushed],
>>
>> on many entries where the page number should be suppressed and an 
>> endnote number, provided with the register command, displayed. This 
>> was not an issue with the July  21 release and earlier.
>>
>> The indexing code for this document is a bit complex, and I have not 
>> yet been able to isolate the issue further.
>>
>> Has anybody else seen similar problems?
>>
>> Is there something in particular I should look for?
>>
>> I am trying to put together an MWE, but have not got it yet.
>>
> Okay, I have created an MWE. The issues appears to be tagging when 
> interaction is active -- disabling either or both of the first two lines 
> below resolves the error, but that should not be necessary.
> 
>     \setupinteraction[state=start]
>     \setuptagging[state=start]
> 
>     \starttexdefinition IndexPageCommand #1
>        \doifsomethingelse{\currentregisterpageuserdata{note}}
>          {n\currentregisterpageuserdata{note}}
>          #1
>        \doifsomething{\currentregisterpageuserdata{id}}
>          {\in[\currentregisterpageuserdata{id}]}
>     \stoptexdefinition
> 
>     \defineregister[Pindex]
>     \setupregister [Pindex][pagecommand=\IndexPageCommand]
> 
>     \starttexdefinition tolerant protected unexpanded PIndex
>     [#1]#:#2#*#=#*#=
>        \doifelse{#3}{}
>          {\setregisterentry[Pindex][keys={#1},entries={#2}]}
>          {\setregisterentry[Pindex][keys={#1},entries={#2}][#3]}
>     \stoptexdefinition
> 
>     \starttext
>          \PIndex{OK, Entry}
>        Here is some text.
>     \startendnote[en:lentz]
>          \PIndex{Bad, Entry1}
>          \PIndex{Bad, Entry2}[note={\in[en:lentz]}]%
>        Here is an endnote.
>     \stopendnote
>     \page
>        \placenotes[endnote]
>     \page
>        \placeregister[Pindex]
>     \stoptext
> 
> Please let me know if this is user error or a bug.

more a side effect of the fact that we preroll that one in order to get 
some alt text or so

This ine is kind of general (add it to cont-new.mlxl):

\appendtoks
     \let\doflushatpar\gobbleoneargument
\to \everypreroll

and that one more specific (protected == unexpanded)

\starttexdefinition tolerant protected PIndex [#1]#:#2#*#=#*#=
     \ifprerolling \else
         \doifelsesomething {#3} {
             \setregisterentry[Pindex][keys={#1},entries={#2}][#3]
         } {
             \setregisterentry[Pindex][keys={#1},entries={#2}]
         }
     \fi
\stoptexdefinition

or a bit more texnic:

\starttexdefinition tolerant protected unexpanded PIndex [#1]#:#2#*#=#*#=
     \ifprerolling
         % not this time
     \orelse\ifparameter#3\or
         \setregisterentry[Pindex][keys={#1},entries={#2}][#3]
     \else
         \setregisterentry[Pindex][keys={#1},entries={#2}]
     \fi
\stoptexdefinition

but then you don't get that 'n' (not sure why)

Hans



-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________

[NTG-context] Re: [entry not flushed] problem since August 17 update

[Rik Kabel]

[-- Attachment #1.1: Type: text/plain, Size: 3938 bytes --]


On 2025-09-12 12:48, Hans Hagen via ntg-context wrote:
> On 9/12/2025 5:33 PM, Rik Kabel wrote:
>>
>> On 2025-09-07 20:40, Rik Kabel wrote:
>>>
>>> Hello list,
>>>
>>> Since the August 17 update, and continuing to the current August 21 
>>> release, my indexes show
>>>
>>>     [entry not flushed],
>>>
>>> on many entries where the page number should be suppressed and an 
>>> endnote number, provided with the register command, displayed. This 
>>> was not an issue with the July  21 release and earlier.
>>>
>>> The indexing code for this document is a bit complex, and I have not 
>>> yet been able to isolate the issue further.
>>>
>>> Has anybody else seen similar problems?
>>>
>>> Is there something in particular I should look for?
>>>
>>> I am trying to put together an MWE, but have not got it yet.
>>>
>> Okay, I have created an MWE. The issues appears to be tagging when 
>> interaction is active -- disabling either or both of the first two 
>> lines below resolves the error, but that should not be necessary.
>>
>>     \setupinteraction[state=start]
>>     \setuptagging[state=start]
>>
>>     \starttexdefinition IndexPageCommand #1
>>        \doifsomethingelse{\currentregisterpageuserdata{note}}
>>          {n\currentregisterpageuserdata{note}}
>>          #1
>>        \doifsomething{\currentregisterpageuserdata{id}}
>>          {\in[\currentregisterpageuserdata{id}]}
>>     \stoptexdefinition
>>
>>     \defineregister[Pindex]
>>     \setupregister [Pindex][pagecommand=\IndexPageCommand]
>>
>>     \starttexdefinition tolerant protected unexpanded PIndex
>>     [#1]#:#2#*#=#*#=
>>        \doifelse{#3}{}
>>          {\setregisterentry[Pindex][keys={#1},entries={#2}]}
>>          {\setregisterentry[Pindex][keys={#1},entries={#2}][#3]}
>>     \stoptexdefinition
>>
>>     \starttext
>>          \PIndex{OK, Entry}
>>        Here is some text.
>>     \startendnote[en:lentz]
>>          \PIndex{Bad, Entry1}
>>          \PIndex{Bad, Entry2}[note={\in[en:lentz]}]%
>>        Here is an endnote.
>>     \stopendnote
>>     \page
>>        \placenotes[endnote]
>>     \page
>>        \placeregister[Pindex]
>>     \stoptext
>>
>> Please let me know if this is user error or a bug.
>
> more a side effect of the fact that we preroll that one in order to 
> get some alt text or so
>
> This ine is kind of general (add it to cont-new.mlxl):
>
> \appendtoks
>     \let\doflushatpar\gobbleoneargument
> \to \everypreroll
>
> and that one more specific (protected == unexpanded)
>
> \starttexdefinition tolerant protected PIndex [#1]#:#2#*#=#*#=
>     \ifprerolling \else
>         \doifelsesomething {#3} {
>             \setregisterentry[Pindex][keys={#1},entries={#2}][#3]
>         } {
>             \setregisterentry[Pindex][keys={#1},entries={#2}]
>         }
>     \fi
> \stoptexdefinition
>
> or a bit more texnic:
>
> \starttexdefinition tolerant protected unexpanded PIndex [#1]#:#2#*#=#*#=
>     \ifprerolling
>         % not this time
>     \orelse\ifparameter#3\or
>         \setregisterentry[Pindex][keys={#1},entries={#2}][#3]
>     \else
>         \setregisterentry[Pindex][keys={#1},entries={#2}]
>     \fi
> \stoptexdefinition
>
> but then you don't get that 'n' (not sure why)
>
> Hans


Well, the 'n', followed by the note number, is pretty important, so 
losing it is not an option. (The design of the index calls for citing 
the endnote number instead of the page number for items in the endnotes.)

The suggested change does eliminate the '[entry not flushed],' problem 
but also introduces another problem: the endnote is now preceded 
'note=1', as in:

-- 
Rik


[-- Attachment #1.2.1: Type: text/html, Size: 6112 bytes --]

[-- Attachment #1.2.2: ySVdjLd3096CXTcs.png --]
[-- Type: image/png, Size: 3312 bytes --]

[-- Attachment #2: Type: text/plain, Size: 511 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________

[NTG-context] Re: [entry not flushed] problem since August 17 update

[Wolfgang Schuster]

Am 12.09.2025 um 21:25 schrieb Rik Kabel:
> 
> On 2025-09-12 12:48, Hans Hagen via ntg-context wrote:
>> [...]
>>
>> but then you don't get that 'n' (not sure why)
>>
>> Hans
> 
> 
> Well, the 'n', followed by the note number, is pretty important, so 
> losing it is not an option. (The design of the index calls for citing 
> the endnote number instead of the page number for items in the endnotes.)

Riks command definition

   \PIndex [<OPT>] {...} {<OPT>} {<OPT>}

and what he uses in the document.

   \PIndex{OK, Entry1}[note={ote}]

He just got lucky because \setregisterentry picks up the following argument.

Wolfgang

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________

[NTG-context] Re: [entry not flushed] problem since August 17 update

[Rik Kabel]

[-- Attachment #1.1: Type: text/plain, Size: 1394 bytes --]


On 2025-09-12 15:47, Wolfgang Schuster wrote:
> Am 12.09.2025 um 21:25 schrieb Rik Kabel:
>>
>> On 2025-09-12 12:48, Hans Hagen via ntg-context wrote:
>>> [...]
>>>
>>> but then you don't get that 'n' (not sure why)
>>>
>>> Hans
>>
>>
>> Well, the 'n', followed by the note number, is pretty important, so 
>> losing it is not an option. (The design of the index calls for citing 
>> the endnote number instead of the page number for items in the 
>> endnotes.)
>
> Riks command definition
>
>   \PIndex [<OPT>] {...} {<OPT>} {<OPT>}
>
> and what he uses in the document.
>
>   \PIndex{OK, Entry1}[note={ote}]
>
> He just got lucky because \setregisterentry picks up the following 
> argument.
>
> Wolfgang
>
> ___________________________________________________________________________________ 
>
> If your question is of interest to others as well, please add an entry 
> to the Wiki!
>
> maillist : ntg-context@ntg.nl / 
> https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
> webpage  : https://www.pragma-ade.nl / https://context.aanhet.net 
> (mirror)
> archive  : https://github.com/contextgarden/context
> wiki     : https://wiki.contextgarden.net
> ___________________________________________________________________________________ 
>

Thank you, Wolfgang. With that corrected, the fix Hans provided indeed 
does the job.


[-- Attachment #1.2: Type: text/html, Size: 2806 bytes --]

[-- Attachment #2: Type: text/plain, Size: 511 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________

[NTG-context] Re: [entry not flushed] problem since August 17 update FOLLOWUP

[Rik Kabel]

[-- Attachment #1.1: Type: text/plain, Size: 704 bytes --]

First, thank you again to Hans and Wolfgang for their parts of the solution.

As Wolfgang pointed out, my macro expected one type of argument and I 
was providing another.  A stricter parser might have caught it earlier, 
but it is now fixed.

The fix Hans provided cured the flushing problem, but only the first 
part was needed, and that needs some additional work. It was not 
necessary to test for prerolling in my indexing macros.

The fix does fail when overload protection is set to error, since it is 
redefining \doflushatpar. I imagine that this will be addressed when it 
is incorporated into a new release. For now I have dropped down to 
warnings and all is well.

-- 
Rik

[-- Attachment #1.2: Type: text/html, Size: 953 bytes --]

[-- Attachment #2: Type: text/plain, Size: 511 bytes --]

___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!

maillist : ntg-context@ntg.nl / https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage  : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive  : https://github.com/contextgarden/context
wiki     : https://wiki.contextgarden.net
___________________________________________________________________________________