e-Book Tools - a Word add-in

[BetterRed]

@Toxaris - ePUBTools 1.23.1 disappeared from the ribbon menu today. It was in Options->Add-ins->COM Add-ins, but it was unchecked and checking it had no effect. Tried uninstall/install of ePUBTools, and the 2010 Office runtime tools, repair Office, sfc/scannow, reboots etc --- all to no avail.

Then I uninstalled it again, and renamed the %APPDATA%\Toxaris directory to %APPDATA%\ToxarisSave and reinstalled - et voilà ePUBTools was back.

I had a look in ToxarisSave and noticed that ePUBTools.cfg was a mere 53 bytes; the .backup and .old files were 16,193 and 16.194 respectively, my memory is that the three files are normally about the same size (as they are now - ~15,800 bytes). This is what was in the .cfg file

<?xml version="1.0" standalone="yes"?>
<ConfigOpt />

Had a look at the log files - nothing out of the ordinary.

Weird

BR

[Tex2002ans]

Bug with the new "File-As" Box

Steps to Reproduce

In the "Generate ePUB" dialog box, type "F.A. Hayek" into the Author field.



click into another box.

File-As automatically shows "Hayek, F.a.":



If you type "F. A. Hayek" into the Author field, the File-As turns out as intended ("Hayek, F. A.").

Expected Functionality

Author: "F.A. Hayek" -> File-As: "Hayek, F.A."

[Toxaris]

I don't know if I can count that as a real bug to be honest. The file-as tries to guess the structure of the name and usually does it pretty well. That being said, the example in your case should be fine. I will take a look at it.
Good thing I didn't make the file-as field read-only, right?

[BetterRed]

@Toxaris - FYI - after I first installed version 1.23.1, I was still getting the spurious counts at end of Mark process that I reported in post 489. Since creating a fresh configuration, see post 496, I'm pretty sure I'm getting accurate counts - including the most important value - 0 (zero).

BR

[Toxaris]

Originally Posted by BetterRed

@Toxaris - FYI - after I first installed version 1.23.1, I was still getting the spurious counts at end of Mark process that I reported in post 489. Since creating a fresh configuration, see post 496, I'm pretty sure I'm getting accurate counts - including the most important value - 0 (zero).

BR

Still strange, since the configuration has zero influence on the count. It does not use parameters at all. Anyway, I have your other request ready now (support of multiple mark files to be controlled by the execute mark button. You can add mark files to a list and those files will show up on the button. Before executing it is checked if the mark file is there and if it validates.

[BetterRed]

Originally Posted by Toxaris

Still strange, since the configuration has zero influence on the count. It does not use parameters at all.

All I know is that the brute force solution I used to fix the serious problem reported in post 496, appears to have also fixed the annoyance reported in post 489. I can't see the connection either. My intuition says its probably an instance of unintentional pathological coupling, errors of that sort often manifest in apparently random and disconnected code paths.

Originally Posted by Toxaris

Anyway, I have your other request ready now (support of multiple mark files to be controlled by the execute mark button. You can add mark files to a list and those files will show up on the button. Before executing it is checked if the mark file is there and if it validates.

Do you mean what's in 1.23.1? I had the impression the feature wasn't 'finished'. The release notes has:

New: Started working on supporting multiple Mark files

Should I try to use it?

BR

[Toxaris]

Originally Posted by BetterRed

All I know is that the brute force solution I used to fix the serious problem reported in post 496, appears to have also fixed the annoyance reported in post 489. I can't see the connection either. My intuition says its probably an instance of unintentional pathological coupling, errors of that sort often manifest in apparently random and disconnected code paths.

Originally Posted by BetterRed

Do you mean what's in 1.23.1? I had the impression the feature wasn't 'finished'. The release notes has:

New: Started working on supporting multiple Mark files

Should I try to use it?
BR

No, I was just started then, but the bug reported about the ePUB generation was serious enough to warrant a bugfix. The code is now finished, but not released yet. You can already add your files into the list, but it doesn't do anything yet, besides adding them.

[BetterRed]

Sorry another bug - I cant Select my XML S&R Rules file, when I do this:



I get this



Happens with 1.23.1 and 1.23, I don't want to go back to 1.22 because then I would lose the changes you made to dialogue checker. I have been using that file for months. I downloaded the sample XML from toxaris.nl/en and got the same result

Can you tell me what I need to hand code into the epubtools.cfg file so that I can use 'myreplacerules.xml' file Ψ²

BR

[Toxaris]

Originally Posted by BetterRed

Sorry another bug - I cant Select my XML S&R Rules file, when I do this:

Attachment 149205

I get this

Attachment 149206

Happens with 1.23.1 and 1.23, I don't want to go back to 1.22 because then I would lose the changes you made to dialogue checker. I have been using that file for months. I downloaded the sample XML from toxaris.nl/en and got the same result

Can you tell me what I need to hand code into the epubtools.cfg file so that I can use 'myreplacerules.xml' file Ψ²

BR

I will fix this of course, this is silly. I typed OR instead of AND.

What you must change in the configuration file specify the correct path in the parameter "locVervang" while Word is not running. When you start Word, that file should be used.

[BetterRed]

Originally Posted by Toxaris

I will fix this of course, this is silly. I typed OR instead of AND.

What you must change in the configuration file specify the correct path in the parameter "locVervang" while Word is not running. When you start Word, that file should be used.

Didn't work - same result as yesterday - are you sure the tag name is right?

the Select file for Mark Rules works OK - here's what I entered



Solved - I added a new locVervang, after the existing empty one I wasn't sure if I needed to wrap the path in quotes, so I used the Mark option as a guide, hence the adjacency.

After I deleted the empty one its OK

BR

[Toxaris]

New release (1.24.0) with several fixes and new options. As usual, see the changelog for all changes.

It also fixes the issues BR and Tex mentioned earlier. There is one critical fix, since in certain situations it could happen that not all endnotes were exported correctly or not exported at all.

[BetterRed]

Sorry for previous false alarm

Enhanced Mark feature with multiple file is good. But I am still getting instances of spurious counts. I'm no longer sure they did go away when I thought they did.

Importantly it can show a +ve number (e.g. 5) but when I search for ^? Highlight nothing is found. If I search for the words in the list manually, nothing is found. So the highlighting is correct, it's the counts that are sometimes wrong. I am only interested in two cases - some words were highlighted or no words were highlighted. I can't see a pattern, maybe I should try again after emptying the Toxaris\Epubtools folder, that seem to fix it temporarily.

And . . . although I have checked 'Field codes' in Configure Ribbon to hide the item, (A) Fieldcodes still shows in the Ribbon, if I uncheck and recheck it in Configure Ribbon it goes away, but it comes back after a while. Its been like this for several releases - I've not mentioned it before because its non-intrusive, but . . . who knows. I don't use Word's Field Code feature.

For the record I have three outstanding defects listed for ePUBTools

1. All ribbon items -- except Configure Ribbon and (A) Fieldcodes -- sometimes show as disabled, even though a document is open. Work around - close documents and restart Word.
2. Counts in Mark Words are sometimes spurious. Work around - none, apart from ignoring the counts.
3. The Field Codes setting in Configure ribbon sometimes not persisting. Work around - because it's unobtrusive, ignore it.

I can't help wondering if they are not related

BR

[Toxaris]

Originally Posted by BetterRed

Sorry for previous false alarm

Enhanced Mark feature with multiple file is good. But I am still getting instances of spurious counts. I'm no longer sure they did go away when I thought they did.

This is really, really strange. It uses the same code as the actual marking, but instead of marking, it counts the occurence. There should be no difference at all. I would like to do it in one go, but that is not supported with the API's. Irritating, since Word itself does know the number of changes.

Originally Posted by BetterRed

And . . . although I have checked 'Field codes' in Configure Ribbon to hide the item, (A) Fieldcodes still shows in the Ribbon, if I uncheck and recheck it in Configure Ribbon it goes away, but it comes back after a while. Its been like this for several releases - I've not mentioned it before because its non-intrusive, but . . . who knows. I don't use Word's Field Code feature.

I might have forgotten something to make it stick. I will double-check of course.

Originally Posted by BetterRed

For the record I have three outstanding defects listed for ePUBTools

1. All ribbon items -- except Configure Ribbon and (A) Fieldcodes -- sometimes show as disabled, even though a document is open. Work around - close documents and restart Word.
2. Counts in Mark Words are sometimes spurious. Work around - none, apart from ignoring the counts.
3. The Field Codes setting in Configure ribbon sometimes not persisting. Work around - because it's unobtrusive, ignore it.

I can't help wondering if they are not related

Ad 1: I have not witnessed this so far on any installation. There is a regular check (every few seconds) that checks the number of open documents. If there is at least one, then the buttons must show. It is a very simple check and I cannot imagine why this doesn't work. There is also only one location in the code where this procedure is run.
Ad 2: Like I said, I cannot really explain this. I will hunt more for a better method for this.
Ad 3: I think I know what is going on there. It will be fixed for sure.

[BetterRed]

@Toxaris - Item 1 happens very occasionally, but I haven't seen it in most recent releases. When it does happens I am, by definition, at a point where stop/start Word isn't a big deal. Item 3 - never bothered me, I only reported now for 'the record'.

Item 2 - all I really want to know is some or no words were marked. If some were marked I don't much care if it was 2 or a 102. What I do care about is if it reports 5 and my search for ^? Highlight finds none. The search is in a macro with a kbshort, so it's not a case of me mis-keying it.

I can do some testing if you need it - my 'apprentice' is doing what I thought I'd be doing this month My intuition is still that 1 & 3 may have the same underlying cause - something is storing something where it shouldn't - used to known as a 'scribbling'.

Request : at the end of Marking could you clear the Replace With Highlight Option - maybe best to clear everything. The eye balling of the highlighted words, sometimes triggers me doing a replace all. If I forget to clear it then the Highlight carries over into my replace...

on file selection in Mark - used it a few times today, much easier than selecting the file in Settings every time.

BR

[EditorOne]

I’m going to give this add-in a try. Thanks, Toxaris. I especially like the prospect of looking for all accented words. The manuscript I’m working with now has a good deal of Spanish in it, as well as some Apache language which uses accents, but also letters with a line through them (example: Daĵɨdɨl).

I do have a question about what fonts I should be using that will ensure the various e-readers will be able to produce such letters instead of inserting an empty box if they are not available in the reader. I need both a serif and a sans-serif font for this manuscript. The Apache parts will use the sans-serif font. Is it best to try to embed the fonts or to specify in the code serif or sans-serif and hope for the best?

Thanks.