Thinking about what we can add

[tperry2x]

In the script of fld "Cerca" (the search field), at the bottom:
on returnInField
if me = empty then textChanged --display all entries
end returnInField

Hi TerryL - I've tried this tonight, and all the entries appear, but unfortunately I get an error on the first entry which comes through as blank. When clicked on, it throws this error:

error-blank.png (120.81 KiB) Viewed 819 times

[TerryL]

I get the same error for '$_Post Binary', '$_Post Raw', and '$_Get Raw' too. Typing in the specific entry throws the same error. Maybe these entries have no accompanying text to extract? No ideas.

[tperry2x]

Hi TerryL,
It is a weird error.
The script I put together, I should clarify, was created in PHP (locally run instance) - and runs the SQL export query together with a bit of xDoTool for good measure.

It seems like there is indeed an entry for $_post binary as my stack reads it - no matter, as I'm moving away from the database approach:

post-binary-post.png (42.32 KiB) Viewed 791 times

Having said that, would be good to solve it and work out why for other xTalk version(s) that might want to retain the SQL dictionary method.
(edit: until Paul mentioned here, I had no idea it used to be done without a sql database)

[TerryL]

Got it. The entries lack the second underline char. I changed to '$_Post_Binary' and it returned the correct description text without error. Now, how to change them and remove the stray empty line too? Maybe Paul can do this with his github tool?

'$_Post Binary', '$_Post Raw', and '$_Get Raw' should be:
'$_Post_Binary', '$_Post_Raw', and '$_Get_Raw'

[OpenXTalkPaul]

To bad that Björnkes BvG Docu 2 doesn't work on LC 8.
No dict.index array any more.

I don't entirely believe that:
-

Yes 'what to do'... I've tried to explained twice previously how to build the dictionary from that folder full of .lcdoc format markdown files (essentially they're plain text with basic tagging system). Briefly, there's a stack in the repo that does push-button building of the dictionary files, that stack uses the revDocsParser script only stack (which I de-branded the scripts) to parse those lcdoc files.

This file:

dict.index

I believe is the index file to the whole bunch of stacks that made up the dictionary in v6.x / 7.x which was before they completely redesigned the Dictionary to use HTML/JS for display. So my guess is that if you ran that exporter stack in v6 or v7 it would work as expected.