Are there any diagnostics for when user tool import fails?

[Tim Campbell]

I have an issue importing my book notes into Accordance. The notes are in html format, but when I go to import the process runs for a while then fails silently. No user tool is created.

 

I suspect that there is something wrong somewhere in the HTM (or perhaps some funny character that is not allowed) and hence causing the HTML parser to barf, but I cannot isolate it thus far. Are there any log files or similar that I can use to work out what line number in the source file the problem is occurring at?

 

I have attached a file containing an excerpt of the overall data I am trying to import to demonstrate the problem, though if there are log files or other diagnostics they would be good to know about also in case this occurs again. At the moment I am stuck with doing binary search trying to find where the problem is, but the process is very error prone.

 

Thanks in advance for any assistance.

[Greg Terry]

Tim:

 

I don't see any file attached to your post. I don't believe there is a log file.

 

If you have access to an html checker like Tidy, I would run your file through that. It can clean up any code that may be a problem.

 

Did you create the code by hand or is it exported out of something?

 

I use a Mac and I have found that I can convert fairly easily from rtf to html by using the Mac Text editor or by using a word processor to export the html code.  MS Word creates a bunch of non-standard code that can be remove with a good html checker program like Tidy. Many html editors have Tidy included or it can be downloaded as a separate program also. There is even an online version of Tidy at: https://infohound.net/tidy/

 

Importing files in Accordance if often an adventure.

 

Hope this helps.

Edited May 15, 2015 by Greg Terry

[Fabian]

Look and try this here:

http://www.accordancebible.com/forums/topic/15717-help-with-using-import-user-tool/

 

Greetings

 

Fabian

[Tim Campbell]

Thanks Greg and Fabian, these are both helpful suggestions. I will investigate using the HTML tidy tool as a first option, as my HTML is hand-crafted and probably contains some typos. 

[Tim Campbell]

It seems the problem is not with the HTML, but with the scripture references in the text I am importing.

 

Basically I keep notes on books and articles that I read and import them into Accordance for searching. Unfortunately, it seems that where references are given in the form <Book> <Chap>.<Verse>, as in some academic and older works, problems start to arise. In particular, where such references occur in a list, the import fails silently. So, for example, a simple HTML file containing only the string "2 Cor. 1.22; 5.5; Eph 1.14" will fail to import.

 

Similarly, it seems that if a verse in this form appears at the start of a list, then it will fail if there is a semicolon and something it doesn't recognise, e.g. "2 Cor. 1.22; 2 Macc. 9" will fail, but "2 Cor. 1:22; 2 Macc. 9" will succeed.

 

This is very frustrating, as it means I am having to go through all of my notes taken over the last 8 years and convert references by hand. I have spent ~6hrs today doing this, and am about half way through. Because the failure is silent (no error message, no line number etc.) I am having to go article by article, manually scan through and, if that doesn't work, delete stuff until the problem goes away to try and localise the issue. Any help on the diagnostic side within Accordance would be greatly appreciated.

 

Ideally, of course, it would be better if the process was more robust - perhaps continuing and creating a log of unparseable references as warnings?

 

Anyway, I am now confident that this warrants a bug report. Is there some easy way of moving this thread into the bug reports section, or do I need to create a new thread?

[Joel Brown]

I apologize for the delay in response, Tim, but this is definitely a bug that we'll get fixed for the next rev.  I'm not 100% if we'll support the autolink of all of those references (I think we should, though), but it definitely shouldn't fail silently.  Thanks for the report!