Function Mode - Not a feature request

phossler · 11-22-2014, 08:14 AM

Don't know any python so a small request

Cook booking the example for auto numbering (changed to H1's) and if I re-run it then it adds another number (I was even able to figure out how to add the word 'Chapter' all by myself

)

Code:

def replace(match, number, file_name, metadata, dictionaries, data, functions, *args, **kwargs):
    section_number = '%d. ' % number
    return match.group(1) + 'Chapter ' +  section_number + match.group(2)

Q - Can someone tell me the tweak to have the function to delete any leading text up to and including the space after the '.' first? If files are rearranged or added or deleted then numbers will be off and re-running the RegEx-Function just adds more. I could do a 'regular' regular expression, but adding the logic to the function would be much handier

Code:

<?xml version='1.0' encoding='utf-8'?>
<html xmlns="http://www.w3.org/1999/xhtml" lang="en">

<head>
  <title>blah</title>
</head>

<body>

  <h1>Chapter 4. Chapter 4. Chapter 4. blah</h1>

</body>

</html>

kovidgoyal · 11-22-2014, 08:40 AM

Code:

def replace(match, number, file_name, metadata, dictionaries, data, functions, *args, **kwargs):
    section_number = '%d. ' % number
    return match.group(1) + 'Chapter ' +  section_number + match.group(2).partition('.')[2][1:]

phossler · 11-22-2014, 11:00 AM

Thanks, but it didn't seem to keep the original text

Did I do it right?

kovidgoyal · 11-22-2014, 11:06 AM

That assumes that there is already a Chapter x. prefix. If you want to handle both the case where there is a prefix and one where there is not, use this:

Code:

import re

def replace(match, number, file_name, metadata, dictionaries, data, functions, *args, **kwargs):
    section_number = '%d. ' % number
    return match.group(1) + 'Chapter ' +  section_number + re.sub(r'^Chapter \d+\. ', '', match.group(2))

phossler · 11-22-2014, 11:38 AM

thanks

Another hopefully not too dumb question

I had the 4 files in 'normal' order (start.xhtml, two.xhtml, three.xhtml, four.xhtml) and ran the code from post#4 several times and it did not keep adding so that works just fine

If I re-arrange the files (four.xhtml, two.xhtml, three.xhtml, start.xhtml) and rerun the F&R again, the first file in logical order (four.xhtml) keeps the Chapter 4, but the sequence number resets to 1 with the second file in logical order (two.xhtml) and goes 2 and 3

Do I need to set something to 1 to get the counter's working 1,2,3,4?

Thanks (and sorry to be a pest)

kovidgoyal · 11-22-2014, 11:44 AM

If you want a consistent numbering across files, your best bet is to close all the open file tabs, otherwise the order will be somewhat random, because find and replace starts at the current cursor position and moves on to the end fo the file, then the next file and so on, until it wraps around to the start of the first file and goes from the top of the file to the cursor position.

IIRC, if none of the files are open, all text files are processed in 'spine' order, that is, the order they appear in in the Files Browser.

phossler · 11-22-2014, 01:24 PM

Yep, I was missing something

Thanks

eschwartz · 11-22-2014, 10:28 PM

Quote:

Originally Posted by kovidgoyal

If you want a consistent numbering across files, your best bet is to close all the open file tabs, otherwise the order will be somewhat random, because find and replace starts at the current cursor position and moves on to the end fo the file, then the next file and so on, until it wraps around to the start of the first file and goes from the top of the file to the cursor position.

IIRC, if none of the files are open, all text files are processed in 'spine' order, that is, the order they appear in in the Files Browser.

That seems like a somewhat confusing way of doing things. I assume there must be some good reason behind the scenes why you are doing it, but at least there should maybe be a warning in the manual in the Function Mode section.

Or maybe this was simply the easiest way of doing it before function mode when I guess it really didn't matter?

kovidgoyal · 11-22-2014, 10:38 PM

Search and replace always goes from the cursor onwards, in pretty much every implementation of S&R in an editor. It allows for much more control.

eschwartz · 11-22-2014, 10:55 PM

Quote:

Originally Posted by kovidgoyal

Search and replace always goes from the cursor onwards, in pretty much every implementation of S&R in an editor. It allows for much more control.

Well, I see the logic in starting form the cursor. But in an ebook where there is a particular order to the files, I can see a certain logic in going to the next file in the spine instead of the next tab.

I don't particularly think it must be changed though, because I haven't really thought about the pros and cons, and I bow to your superior knowledge.

I do think if that is the way it works, it should be noted in the documentation for Function Mode, as a potentially common pitfall.

kovidgoyal · 11-22-2014, 10:58 PM

The proper solution is to add a function annotation, like call_after_last_match that tells the editor to run the search in spine order or reverse spine order, when searching on multiple files.

phossler · 11-23-2014, 09:51 AM

I was thinking maybe a new F&R checkbox

[X] Spine Order

that just started the F&R at file#1 instead of the current file

Up/Down and Wrap would be unchanged (probably)

eschwartz · 11-23-2014, 05:35 PM

Thanks for the change!

https://github.com/kovidgoyal/calibr...e5dc6b5f2708bd

phossler · 11-24-2014, 08:12 AM

So would it look like this when the capability is in the next version?

Code:

mport re

def replace(match, number, file_name, metadata, dictionaries, data, functions, *args, **kwargs):
    section_number = '%d. ' % number
    replace.file_order = 'spine'
    return match.group(1) + 'Chapter ' +  section_number + re.sub(r'^Chapter \d+\. ', '', match.group(2))

Great stuff -- thanks

kovidgoyal · 11-24-2014, 08:26 AM

Not quite

Code:

import re

def replace(match, number, file_name, metadata, dictionaries, data, functions, *args, **kwargs):
    section_number = '%d. ' % number    
    return match.group(1) + 'Chapter ' +  section_number + re.sub(r'^Chapter \d+\. ', '', match.group(2))

replace.file_order = 'spine'

11-22-2014, 11:00 AM	#3
phossler Wizard Posts: 1,090 Karma: 447222 Join Date: Jan 2009 Location: Valley Forge, PA, USA Device: Kindle Paperwhite	Thanks, but it didn't seem to keep the original text Did I do it right? Attached Thumbnails

11-23-2014, 09:51 AM	#12
phossler Wizard Posts: 1,090 Karma: 447222 Join Date: Jan 2009 Location: Valley Forge, PA, USA Device: Kindle Paperwhite	I was thinking maybe a new F&R checkbox [X] Spine Order that just started the F&R at file#1 instead of the current file Up/Down and Wrap would be unchanged (probably) Attached Thumbnails Last edited by phossler; 11-23-2014 at 10:09 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Function mode: Feature Request	jbacelar	Editor	2	11-22-2014 02:36 AM
Function mode in editor S&R -- coming soon	eschwartz	Editor	12	11-21-2014 08:26 AM
Error in function mode in editor S&R	jbacelar	Editor	3	11-21-2014 05:34 AM
Feature Request at the import-function	ralphffm44	Calibre	4	04-02-2010 01:14 PM
Feature Request - cover browse in the "Fetch Metadata from server" function	nathander13	Calibre	1	01-30-2010 02:11 PM

11-22-2014, 08:14 AM	#1
phossler Wizard Posts: 1,090 Karma: 447222 Join Date: Jan 2009 Location: Valley Forge, PA, USA Device: Kindle Paperwhite	Function Mode - Not a feature request Don't know any python so a small request Cook booking the example for auto numbering (changed to H1's) and if I re-run it then it adds another number (I was even able to figure out how to add the word 'Chapter' all by myself ) Code: def replace(match, number, file_name, metadata, dictionaries, data, functions, args, *kwargs): section_number = '%d. ' % number return match.group(1) + 'Chapter ' + section_number + match.group(2) Q - Can someone tell me the tweak to have the function to delete any leading text up to and including the space after the '.' first? If files are rearranged or added or deleted then numbers will be off and re-running the RegEx-Function just adds more. I could do a 'regular' regular expression, but adding the logic to the function would be much handier Code: <?xml version='1.0' encoding='utf-8'?> <html xmlns="http://www.w3.org/1999/xhtml" lang="en"> <head> <title>blah</title> </head> <body> <h1>Chapter 4. Chapter 4. Chapter 4. blah</h1> </body> </html>

11-22-2014, 11:06 AM	#4
kovidgoyal creator of calibre Posts: 45,604 Karma: 28548974 Join Date: Oct 2006 Location: Mumbai, India Device: Various	That assumes that there is already a Chapter x. prefix. If you want to handle both the case where there is a prefix and one where there is not, use this: Code: import re def replace(match, number, file_name, metadata, dictionaries, data, functions, args, *kwargs): section_number = '%d. ' % number return match.group(1) + 'Chapter ' + section_number + re.sub(r'^Chapter \d+\. ', '', match.group(2))

11-22-2014, 11:44 AM	#6
kovidgoyal creator of calibre Posts: 45,604 Karma: 28548974 Join Date: Oct 2006 Location: Mumbai, India Device: Various	If you want a consistent numbering across files, your best bet is to close all the open file tabs, otherwise the order will be somewhat random, because find and replace starts at the current cursor position and moves on to the end fo the file, then the next file and so on, until it wraps around to the start of the first file and goes from the top of the file to the cursor position. IIRC, if none of the files are open, all text files are processed in 'spine' order, that is, the order they appear in in the Files Browser.

11-22-2014, 01:24 PM	#7
phossler Wizard Posts: 1,090 Karma: 447222 Join Date: Jan 2009 Location: Valley Forge, PA, USA Device: Kindle Paperwhite	Yep, I was missing something Thanks

11-22-2014, 10:38 PM	#9
kovidgoyal creator of calibre Posts: 45,604 Karma: 28548974 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Search and replace always goes from the cursor onwards, in pretty much every implementation of S&R in an editor. It allows for much more control.

11-22-2014, 10:58 PM	#11
kovidgoyal creator of calibre Posts: 45,604 Karma: 28548974 Join Date: Oct 2006 Location: Mumbai, India Device: Various	The proper solution is to add a function annotation, like call_after_last_match that tells the editor to run the search in spine order or reverse spine order, when searching on multiple files.

11-23-2014, 05:35 PM	#13
eschwartz Ex-Helpdesk Junkie Posts: 19,421 Karma: 85400180 Join Date: Nov 2012 Location: The Beaten Path, USA, Roundworld, This Side of Infinity Device: Kindle Touch fw5.3.7 (Wifi only)	Thanks for the change! https://github.com/kovidgoyal/calibr...e5dc6b5f2708bd

Advert

Advert