[GUI Plugin] Multi-Column Search

[DaltonST]

Originally Posted by Tanjamuse

I've tried this:

Story URL: https://www.fanfiction.net/s/[0-9][0-9][0-9]+/|Story URL: http://www.fanfiction.net/s/[0-9][0-9][0-9]+/

I don't understand the number in the selection after story_id.

The only reason I inserted the 9 was because it was in your example image earlier.

If this is wrong, then I would appreciate some help.

Also some of the urls are http and some are https.

If you don't understand the #9 and what it does, then it is because you have not analyzed the image attached way above that shows an example of what you are doing. It has 4 images from left to right.

It is also because you have not read the ToolTips for the widgets in general, and the one that has a "9" in particular.

Whether some are http or https does not at all matter, since the regular expression handles both.

Please read the ToolTips in MCS, plus all of my previous posts about this.



DaltonST

[Tanjamuse]

It now finds books, except it finds all the books in the library instead of just the ones where the url in the text-file matches the id-number.

It even matches those books which has no id-number.

In my custom column for the story_url I don't have the Story URL: only https://www.fanfiction.net/s/7347050/1/

[DaltonST]

Originally Posted by Tanjamuse

It now finds books, except it finds all the books in the library instead of just the ones where the url in the text-file matches the id-number.

It even matches those books which has no id-number.

In my custom column for the story_url I don't have the Story URL: only https://www.fanfiction.net/s/7347050/1/

The screenprint shows that your "new" search query regular expression has an error again. Change it to:

https://www.fanfiction.net/s/[0-9][0-9][0-9]+/|http://www.fanfiction.net/s/[0-9][0-9][0-9]+/

It also finds all books because you did not check the checkbox called "When Found:". That means that the new functionality is not being used at all.


DaltonST

[Tanjamuse]

I've read every tool tip in the image and the one in the large field in MCS. In the image example (and the tool tip) you have this version of the search:

https://www.fanfiction.net/s/[0-9]+/[0-9]/|http://www.fanfiction.net/s/[0-9]+/[0-9]

but when I look at the tool tip for the MCS Regular Expression Tab, the example shown has:

[0-9][0-9][0-9]+/

Also the number in bold in the url below is always the same.

http://www.fanfiction.net/s/3507422/1/

Does the number 9 in the MCS Regular Expression Tab tell it to match the entire string of numbers?

[DaltonST]

See the attached example that I just created to answer your question.

I highly recommend that you personally use https://pythex.org/ to learn how regular expressions work.




DaltonST

[Tanjamuse]

I've added this to the regular expression tab in #9:

[0-9][0-9][0-9]+

and this in the txt queries:

https://www.fanfiction.net/s/[0-9][0-9][0-9]+/|http://www.fanfiction.net/s/[0-9][0-9][0-9]+/

There's only one book with any number and no other metadata and I've used those books I sent to you.

I haven't added the story_url or anything else, only the Story_ID to test if it can be found, but I still get nothing.

I'm sorry to be such a bother. I have looked at the site and I understand at least that [0-9][0-9][0-9]+/ only extract the number.

[DaltonST]

I copied exactly what you used in your screenprints, and I successfully matched #story_id (a long-text, like comments CC interpreted like short-text, like a title) and updated #story_url (a long-text, like comments CC interpreted like short-text, like a title) all 20 TXT format ebooks that you gave me as test data. It is perfect.

What you did not attach was a screenprint of your Calibre library data showing what it was looking for, plus proof that you are searching TXT formats and not EPUB formats. Verify that your #story_id CC is an appropriate textual CC.


My test results per your last post #216 from this morning are attached. They are perfect.





DaltonST

[Tanjamuse]

I've sent you a PM with a copy of the library.

[DaltonST]

Originally Posted by Tanjamuse

I've sent you a PM with a copy of the library.

The problem is that the #story_id custom column is a comma-separated "Tag" type, rather than a simple plain-text type. An ID is unique to a book, and not generic like Tags are. Story IDs are unique, and each book can have only one (1) of them, so it really should not be a Tag-type, but a simple plain-text type. That is why ISBN is not a Tag, but a textual Identifier.

Test results:

Match found but re_text_found_final <> compare_to_value: 2131277 [u'2131277']

However, I can easily add a little code to account for that scenario, which obviously I did not test for. My bad.

I will upload a new version later today.



DaltonST

[Tanjamuse]

I need it to be a tag column because when I merge the same story from different archives it has to stay as separate tags.

[DaltonST]

Version 1.0.74 - 13 October 2017 Full-Text Search enhancements for Tag-like Custom Columns to be updated by merging existing values with the new final search value. There can be no duplicates, of course, just as for real Tags.



DaltonST

[Tanjamuse]

Can you send me back a version of my library where it works, because even though I've updated the plugin I still get nothing.

[DaltonST]

I used the library you gave me as-is. I changed nothing. I would just be sending back what you already have.

The MCS preferences are Calibre-wide, not specific to a particular library. When I opened your library, my MCS still had my settings from before.

Look at the attachment from my post this morning of #217. Your settings should be those, since those are what you were using.


DaltonST

[DaltonST]

Attached is your Library you gave me updated by me.

I first updated Story URL, and then updated Story ID both with MCS and Mass Edit to add test tags.

[Tanjamuse]

My mistake, I hadn't noticed that the regular expression wasn't saved.

It now works. Thank you so much for your patience.