e-Book Tools - a Word add-in - Page 44

BetterRed · 09-12-2018, 06:49 PM

Quote:

Originally Posted by Toxaris

Well, on the good side, I rewrote the complete settings part in the new version and it is much, much stable and resilient. Downside is that this version is not done yet and it will be a while. However, I will look at the old code and see if I can quickly (and correctly) fix it for you.

Thanks - no rush, I now have a macro to delete the pesky [###], [/###] tags.

I'm still puzzled as to how I had it working for so long without tags

.

BR

kbaerwald · 09-15-2018, 02:37 PM

Hi

after a longer pause with epub editing (2 months) I started some new projects last Monday. Most of them are pdf ebooks which I load directly into Word (MS Office 365 V1808) for conversion. After some basic corrections I usually start Postprocess OCR. Currently this module runs into step 3/15 (tag markup) then crashes Word, repeatedly. Tried different pdf and also plain docx files. Same story.

Nothing changed during the recent 2 months except for some updates of Word. I did not touch the ePubTools.

All other functions (e.g. search/replace or epub import) are working perfectly!

Anyone with such problems here around?

Klaus

Toxaris · 09-16-2018, 03:37 AM

Well, yesterday I did a book myself and had issues in that department as well. I also got a report from someone else. I am not quite sure why yet, as this does not happen on older version of Word. They did make a change, but it is not clear what it is and it is also not consistent. I am looking into it though.

Toxaris · 09-16-2018, 05:42 AM

I can confirm there is something broken in at least the latest version of Word (Office 365) in the search and replace engine of Word itself. I can repeat the observed behavior manually. Unfortunately this cannot be fixed by me of course. I did think of a way to circumvent this, but it will have a performance impact and I need time to program it as well.
I will post updates every now and then.

kbaerwald · 09-16-2018, 06:20 AM

Can you specify what is broken in the search/replace engine? Does it appear directly in the user interface or just in the programmatic interface that you use?

If it is not a permanent change by MS we should better wait for the next update to come. It would be too much a burden for you to find an intermediate fix for that problem.

Anyway - thank you for your help!

Klaus

BetterRed · 09-16-2018, 06:59 AM

Quote:

Originally Posted by kbaerwald

Hi

after a longer pause with epub editing (2 months) I started some new projects last Monday. Most of them are pdf ebooks which I load directly into Word (MS Office 365 V1808) for conversion. After some basic corrections I usually start Postprocess OCR. Currently this module runs into step 3/15 (tag markup) then crashes Word, repeatedly. Tried different pdf and also plain docx files. Same story.

Nothing changed during the recent 2 months except for some updates of Word. I did not touch the ePubTools.

All other functions (e.g. search/replace or epub import) are working perfectly!

Anyone with such problems here around?

Klaus

I may have had that OCR PP behaviour on Word 2007 and a much older ePubTools. But I know for sure I had a very similar problem in the HTML/EPUB creation tools. IIRC the problems went away if I disabled/un-installed another add-on I use - Office Tabs. I decided I couldn't live without Office Tabs, but I could live without those features of ePubTools.

BR

BetterRed · 09-16-2018, 07:16 AM

Quote:

Originally Posted by kbaerwald

Can you specify what is broken in the search/replace engine? Does it appear directly in the user interface or just in the programmatic interface that you use?

I've not encountered any problems in Word 2016, using it manually, or via EPubTools Search and Replace.

BR

kbaerwald · 09-16-2018, 07:46 AM

BR

most of the time I am using the Postprocess OCR for its very good scene detection feature (variable 'detection distance'). The other features are more a by-catch for me - they may be important for others however.

If Postprocess OCR remains unusuable I have to look for other ways - unfortunately the standard Word Search/Replace UI ist not too flexible regarding Regex.

Klaus

Toxaris · 09-16-2018, 09:27 AM

Sure I can explain more. If you for example do the following search:

Find: empty
Find formatting: italic
Replace: [dummy]^&[/dummy]
Replace formatting: non italic

If you now do a S&R, it should just set the dummy tags before and after the italic part and also remove the italic formatting. That is how it should be and always was. However, now the tags are placed at the correct place, but not everything is set to non-italic (including the end-tag!). That causes the same part of the block (including tag) to be selected again, and again, and again. That is not how it should work and this was recently introduced. I do expect it to be solved soon, as this is clearly an issue in Word itself as I can reproduce it manually.

kbaerwald · 09-16-2018, 04:05 PM

Quote:

Originally Posted by Toxaris

Sure I can explain more. If you for example do the following search:

Find: empty
Find formatting: italic
Replace: [dummy]^&[/dummy]
Replace formatting: non italic

If you now do a S&R, it should just set the dummy tags before and after the italic part and also remove the italic formatting. That is how it should be and always was. However, now the tags are placed at the correct place, but not everything is set to non-italic (including the end-tag!). That causes the same part of the block (including tag) to be selected again, and again, and again. That is not how it should work and this was recently introduced. I do expect it to be solved soon, as this is clearly an issue in Word itself as I can reproduce it manually.

Hi Toxaris

true - can confirm this (see snap shot below: replaced the "Sch" of Schönheit only). Funny - well we have to wait that MS is taking notice of the problem and is solving it. Thanks for the demonstration.

Klaus

BetterRed · 09-16-2018, 09:29 PM

Yep - me too.

IMO empty Find shouldn't do anything, there should be an Any String (^@ or something)

Curious : does this relate directly to Klaus' OCR PP problem.

BR

Toxaris · 09-17-2018, 02:33 PM

Yes, an empty find sounds strange but it works fine. Basically you say find anything. There is no Any String command unfortunately.

It is absolutely related to Klaus' OCR PP problem. It is actually doing that to tag the formatting in that step. As not all the text is de-formatted, it is again selected and tagged and again, again, etc. So, it appears to hang. If you wait long enough, it will end, but you will have a big mess with half tags all over. I actually had the same problem myself as well as someone else I know. I am not too optimistic that MS will solve this soon, they have the most atrocious bug reporting system I know and you actually needs votes to have your bug handled sooner or so it seems.

Hence, I am thinking of an alternative method. I have designed one, but it will require some time to program and test. I hope to start tomorrow evening.

DaleDe · 09-17-2018, 06:56 PM

Reminds me of a time in history when people thought there was no need for a zero. It took centuries for people to trust a place based number system from roman numerals.

BetterRed · 09-17-2018, 07:56 PM

@Toxaris - Ouch!

Unsurprisingly it's Tag Formatting too, which I use occasionally. What else ?

BR

Toxaris · 09-18-2018, 12:46 PM

This functionality is used in Post OCR, Tag Formatting and HTML Conversion... So, it is quite essential unfortunately. That is why I am building an alternative.

09-15-2018, 02:37 PM	#647
kbaerwald BioReader Posts: 292 Karma: 42568 Join Date: Apr 2009 Location: Germany Device: Various	Hi after a longer pause with epub editing (2 months) I started some new projects last Monday. Most of them are pdf ebooks which I load directly into Word (MS Office 365 V1808) for conversion. After some basic corrections I usually start Postprocess OCR. Currently this module runs into step 3/15 (tag markup) then crashes Word, repeatedly. Tried different pdf and also plain docx files. Same story. Nothing changed during the recent 2 months except for some updates of Word. I did not touch the ePubTools. All other functions (e.g. search/replace or epub import) are working perfectly! Anyone with such problems here around? Klaus Attached Thumbnails Last edited by kbaerwald; 09-15-2018 at 02:46 PM. Reason: Attached snap shot

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Save any Document from Microsoft Word 2007 to EPUB using a Free Add-in from Aspose	romeok	ePub	45	05-03-2017 03:33 PM
Automatically add tag if word/phrase found in ebook?	eosrose	Calibre	3	11-16-2011 06:48 AM
Doc Splitter-Macro or Add-in For Word 2003	konrad	Workshop	0	03-08-2011 03:43 PM
Can't see lit add in for MS Word	ray100	Reading and Management	3	08-12-2009 02:59 AM

09-16-2018, 03:37 AM	#648
Toxaris Wizard Posts: 4,520 Karma: 121692313 Join Date: Oct 2009 Location: Heemskerk, NL Device: PRS-T1, Kobo Touch, Kobo Aura	Well, yesterday I did a book myself and had issues in that department as well. I also got a report from someone else. I am not quite sure why yet, as this does not happen on older version of Word. They did make a change, but it is not clear what it is and it is also not consistent. I am looking into it though.

09-16-2018, 05:42 AM	#649
Toxaris Wizard Posts: 4,520 Karma: 121692313 Join Date: Oct 2009 Location: Heemskerk, NL Device: PRS-T1, Kobo Touch, Kobo Aura	I can confirm there is something broken in at least the latest version of Word (Office 365) in the search and replace engine of Word itself. I can repeat the observed behavior manually. Unfortunately this cannot be fixed by me of course. I did think of a way to circumvent this, but it will have a performance impact and I need time to program it as well. I will post updates every now and then.

09-16-2018, 06:20 AM	#650
kbaerwald BioReader Posts: 292 Karma: 42568 Join Date: Apr 2009 Location: Germany Device: Various	Can you specify what is broken in the search/replace engine? Does it appear directly in the user interface or just in the programmatic interface that you use? If it is not a permanent change by MS we should better wait for the next update to come. It would be too much a burden for you to find an intermediate fix for that problem. Anyway - thank you for your help! Klaus

09-16-2018, 07:46 AM	#653
kbaerwald BioReader Posts: 292 Karma: 42568 Join Date: Apr 2009 Location: Germany Device: Various	BR most of the time I am using the Postprocess OCR for its very good scene detection feature (variable 'detection distance'). The other features are more a by-catch for me - they may be important for others however. If Postprocess OCR remains unusuable I have to look for other ways - unfortunately the standard Word Search/Replace UI ist not too flexible regarding Regex. Klaus

09-16-2018, 09:27 AM	#654
Toxaris Wizard Posts: 4,520 Karma: 121692313 Join Date: Oct 2009 Location: Heemskerk, NL Device: PRS-T1, Kobo Touch, Kobo Aura	Sure I can explain more. If you for example do the following search: Find: empty Find formatting: italic Replace: [dummy]^&[/dummy] Replace formatting: non italic If you now do a S&R, it should just set the dummy tags before and after the italic part and also remove the italic formatting. That is how it should be and always was. However, now the tags are placed at the correct place, but not everything is set to non-italic (including the end-tag!). That causes the same part of the block (including tag) to be selected again, and again, and again. That is not how it should work and this was recently introduced. I do expect it to be solved soon, as this is clearly an issue in Word itself as I can reproduce it manually.

09-16-2018, 09:29 PM	#656
BetterRed null operator (he/him) Posts: 20,620 Karma: 26960534 Join Date: Mar 2012 Location: Sydney Australia Device: none	Yep - me too. IMO empty Find shouldn't do anything, there should be an Any String (^@ or something) Curious : does this relate directly to Klaus' OCR PP problem. BR

09-17-2018, 02:33 PM	#657
Toxaris Wizard Posts: 4,520 Karma: 121692313 Join Date: Oct 2009 Location: Heemskerk, NL Device: PRS-T1, Kobo Touch, Kobo Aura	Yes, an empty find sounds strange but it works fine. Basically you say find anything. There is no Any String command unfortunately. It is absolutely related to Klaus' OCR PP problem. It is actually doing that to tag the formatting in that step. As not all the text is de-formatted, it is again selected and tagged and again, again, etc. So, it appears to hang. If you wait long enough, it will end, but you will have a big mess with half tags all over. I actually had the same problem myself as well as someone else I know. I am not too optimistic that MS will solve this soon, they have the most atrocious bug reporting system I know and you actually needs votes to have your bug handled sooner or so it seems. Hence, I am thinking of an alternative method. I have designed one, but it will require some time to program and test. I hope to start tomorrow evening.

09-17-2018, 06:56 PM	#658
DaleDe Grand Sorcerer Posts: 11,470 Karma: 13095790 Join Date: Aug 2007 Location: Grass Valley, CA Device: EB 1150, EZ Reader, Literati, iPad 2 & Air 2, iPhone 7	Reminds me of a time in history when people thought there was no need for a zero. It took centuries for people to trust a place based number system from roman numerals.

09-17-2018, 07:56 PM	#659
BetterRed null operator (he/him) Posts: 20,620 Karma: 26960534 Join Date: Mar 2012 Location: Sydney Australia Device: none	@Toxaris - Ouch! Unsurprisingly it's Tag Formatting too, which I use occasionally. What else ? BR

09-18-2018, 12:46 PM	#660
Toxaris Wizard Posts: 4,520 Karma: 121692313 Join Date: Oct 2009 Location: Heemskerk, NL Device: PRS-T1, Kobo Touch, Kobo Aura	This functionality is used in Post OCR, Tag Formatting and HTML Conversion... So, it is quite essential unfortunately. That is why I am building an alternative.

Advert

Advert