How i can sort number from largest to smallest on notepad++

Mark Olson

@Alan-Kilborn
Awesome! I don’t mind being upstaged at all, and it has all the features I would have wanted. I’ll play around with this and see if I can come up with improvements.

TBugReporter

@Alan-Kilborn said in How i can sort number from largest to smallest on notepad++:

(Yep, my ever-so-cheesy-UI)

Nice of you to acknowledge this. 🤓

Mark Olson

OK, I’ve added two new options: lowercase culture-insensitive (Python can do that, it’s just rather obscure) and sorting unmatched lines to bottom.

The text:

foo: tone bar: 27.1
foo: bLue bar: 241.3
foo: taco bar: 66.5
foo: basst bar: 281.7
foo: öyster bar: 144.0
foo: oyster bar: 164.1
foo: Spb bar: 102.1
foo: blüe bar: 203.2
foo: spä bar: 121.2
foo: spb bar: 84.2
foo: blve bar: 183.2
foo: blue bar: 222.2
foo: baßk bar: 261.3
foo: täco bar: 47.3

Sorting unmatched lines to bottom (regex bar: (\d\d)\D, all other options default). Note that the lines with a 3-digit number all stay in the same order.

foo: tone bar: 27.1
foo: täco bar: 47.3
foo: taco bar: 66.5
foo: spb bar: 84.2
foo: bLue bar: 241.3
foo: basst bar: 281.7
foo: öyster bar: 144.0
foo: oyster bar: 164.1
foo: Spb bar: 102.1
foo: blüe bar: 203.2
foo: spä bar: 121.2
foo: blve bar: 183.2
foo: blue bar: 222.2
foo: baßk bar: 261.3

Ignorecase culture-sensitive (regex foo: (\w+), all other options default)

foo: basst bar: 281.7
foo: baßk bar: 261.3
foo: bLue bar: 241.3
foo: blue bar: 222.2
foo: blüe bar: 203.2
foo: blve bar: 183.2
foo: oyster bar: 164.1
foo: öyster bar: 144.0
foo: spä bar: 121.2
foo: Spb bar: 102.1
foo: spb bar: 84.2
foo: taco bar: 66.5
foo: täco bar: 47.3
foo: tone bar: 27.1

And finally, my modification of the script:

# -*- coding: utf-8 -*-
from __future__ import print_function

#########################################
#
#  SortLinesWithRegexGroup1AsSortKey (SLWRG1ASK)
#
#########################################

# references:
#  https://community.notepad-plus-plus.org/topic/24325

#-------------------------------------------------------------------------------

from Npp import *
import inspect
import os
import re
import threading
import unicodedata
import SendKeys as sk

#-------------------------------------------------------------------------------

def ignorecase_culture_invariant(x):
    return unicodedata.normalize('NFD', x.lower())

class SLWRG1ASK(object):

    def __init__(self):

        self.this_script_name = inspect.getframeinfo(inspect.currentframe()).filename.split(os.sep)[-1].rsplit('.', 1)[0]

        self.key_and_line_content_tup_list = []
        self.matched_line_numbers = set()
        self.key_did_not_match_specified_type__ERROR = False
        self.user_did_not_specify_group1__ERROR = False
        self.not_all_lines_matched__ERROR = False
        self.last_matched_line_num = None
        self.user_line_num_of_error = None
        self.zero_matches_for_regex = True
        self.key_transform = None
        self.do_descending_sort = False
        self.each_line_must_have_regex_group1_key_match = True
        self.remove_lines_that_dont_match_regex = False
        self.sort_lines_that_dont_match_to_bottom = False

        rect_sel_mode = editor.getSelectionMode() in [ SELECTIONMODE.RECTANGLE, SELECTIONMODE.THIN ]
        if rect_sel_mode or editor.getSelections() > 1:
            self.mb('Cannot operate on a column selection or multiple selections; aborting!')
            return

        if len(editor.getSelText()) == 0:
            if not self.mb_ok_cancel('No selection active; confirm desire to sort entire file by pressing OK.'):
                return

        (search_start_line, search_end_line) = editor.getUserLineSelection()
        if editor.positionFromLine(search_end_line) == editor.getSelectionEnd(): search_end_line -= 1
        num_lines_in_user_sel = search_end_line - search_start_line + 1
        if num_lines_in_user_sel < 2: return  # if selection only on 1 line, nothing to do; already sorted!

        user_regex_with_group_1 = ''
        while True:
            user_regex_with_group_1 = self.prompt('Enter regex with group 1 being the sort key', user_regex_with_group_1)
            if user_regex_with_group_1 is None: return  # user cancel
            if len(user_regex_with_group_1.strip()) == 0:
                self.mb('Cannot specify an empty regex!  Try again.')
                continue
            if '(' not in user_regex_with_group_1:
                self.mb('Need to specify capture group 1 (as the sort key) in the regex!  Try again.')
                continue
            regex_err_msg = self.search_regex_is_invalid_error_msg(user_regex_with_group_1)
            if len(regex_err_msg) > 0:
                self.mb('Bad regular expression!\r\n\r\n{e}\r\n\r\nTry again.'.format(e=regex_err_msg))
                continue
            break

        while True:
            threading.Timer(0.25, lambda : sk.SendKeys("{RIGHT}")).start()  # clear auto-selection of prompt text when box appears
            user_options = self.prompt(
                '    Regex specified: {r}\r\n'
                'Choose options by placing 1 x in each group (no need to x to get DEFAULT choice)'.format(r=user_regex_with_group_1),
                '\r\n'.join([
                    'Group 1 transformation:  [   ]Integer(DEFAULT)      [   ]Decimal      [   ]String      [   ]Ignorecase-cultural',
                    'Sort order:  [   ]Ascending(DEFAULT)    [   ]Descending',
                    'Other:',
                    '[   ]Sort-only-if-all-lines-have-group-1-match(DEFAULT)',
                    '[   ]Remove-non-matching-lines',
                    '[   ]Sort-non-matching-lines-to-bottom',
                ])
                )
            if user_options is None: return  # user cancel
            if len(user_options.strip()) == 0:
                self.mb('Cannot specify empty options!  Try again.')
                continue
            if self.option_check(user_options, 'String'): self.key_transform = str
            elif self.option_check(user_options, 'Decimal'): self.key_transform = float
            elif self.option_check(user_options, 'Ignorecase-cultural'): self.key_transform = ignorecase_culture_invariant
            else: self.key_transform = int
            self.do_descending_sort = self.option_check(user_options, 'Descending')
            self.each_line_must_have_regex_group1_key_match = self.option_check(user_options, 'Sort-only-if-all-lines-have-group-1-match')
            self.remove_lines_that_dont_match_regex = self.option_check(user_options, 'Remove-non-matching-lines')
            self.sort_lines_that_dont_match_to_bottom = self.option_check(user_options, 'Sort-non-matching-lines-to-bottom')
            if not (self.remove_lines_that_dont_match_regex or self.sort_lines_that_dont_match_to_bottom):
                self.each_line_must_have_regex_group1_key_match = True
            if self.each_line_must_have_regex_group1_key_match + self.remove_lines_that_dont_match_regex + self.sort_lines_that_dont_match_to_bottom > 1:
                self.mb('Can only specify one way of handling unmatched lines! Try again!')
                continue
            break

        # enforce single-line matching and only get the FIRST match on each line:
        user_regex_with_group_1 = '(?-s)' + user_regex_with_group_1 + '.*'

        search_start_pos = editor.positionFromLine(search_start_line)
        search_end_pos = editor.getLineEndPosition(search_end_line)

        # special "seeding", in case the very first line of the selection isn't hit:
        self.last_matched_line_num = search_start_line - 1

        editor.research(user_regex_with_group_1, lambda m: self.match_found(m), 0, search_start_pos, search_end_pos)

        if self.zero_matches_for_regex:
            self.mb('No lines matched the regex; aborting!')
            return

        if self.user_did_not_specify_group1__ERROR:
            self.mb('No sort key specified (via using capture group 1) in the regex; aborting!')
            return

        if self.key_did_not_match_specified_type__ERROR:
            self.mb('Key data did not match specified type ({t}) on line {L}; aborting!'.format(L=self.user_line_num_of_error,
                t={ int : 'Integer', float : 'Decimal', str : 'String', ignorecase_culture_invariant: 'String' }[self.key_transform]))
            return

        if self.each_line_must_have_regex_group1_key_match and self.not_all_lines_matched__ERROR:
            self.mb('Not every line in the source data matched the specified regex (first non-matching line was {L}); aborting!'.format(L=self.user_line_num_of_error))
            return

        if len(self.key_and_line_content_tup_list) < 2:
            self.mb('Nothing (reasonable) found to sort; aborting!')
            return

        sorted_line_list = [tup[1] for tup in sorted(self.key_and_line_content_tup_list, reverse=self.do_descending_sort)]
        
        if self.matched_line_numbers:
            first_unmatched_line = search_start_line + len(self.matched_line_numbers) + 1
            self.mb('Line numbers {0}-{1} in sorted text were NOT matched.'.format(first_unmatched_line, search_end_line + 1),
                title='not all lines matched')
            for line_num in range(search_start_line, search_end_line + 1):
                if line_num not in self.matched_line_numbers:
                    unmatched_line = editor.getLine(line_num)
                    sorted_line_list.append(unmatched_line.rstrip('\r\n'))

        eol = ['\r\n', '\r', '\n'][editor.getEOLMode()]

        doc_len_before = editor.getLength()

        editor.setSelection(search_end_pos, search_start_pos)
        editor.replaceSel(eol.join(sorted_line_list))

        doc_len_delta = editor.getLength() - doc_len_before

        # leave sorted text selected:
        editor.setSelection(search_end_pos + doc_len_delta, search_start_pos)

    def match_found(self, m):

        self.zero_matches_for_regex = False

        # we're already corrupt; don't bother processing further matches after the one where we found the problem
        if self.user_did_not_specify_group1__ERROR: return
        if self.key_did_not_match_specified_type__ERROR: return
        if self.not_all_lines_matched__ERROR: return

        (start_pos_of_match, end_pos_of_match) = m.span(0)

        line_num_of_match = editor.lineFromPosition(start_pos_of_match)

        if line_num_of_match != self.last_matched_line_num + 1:
            if self.each_line_must_have_regex_group1_key_match:
                self.not_all_lines_matched__ERROR = True
                system_line_num_of_err = self.last_matched_line_num + 1
                self.user_line_num_of_error = system_line_num_of_err + 1
                return
        self.last_matched_line_num = line_num_of_match
        if self.sort_lines_that_dont_match_to_bottom:
            self.matched_line_numbers.add(line_num_of_match)

        try:
            g1_str = m.group(1)
        except IndexError:
            self.user_did_not_specify_group1__ERROR = True
            return

        try:
            key_with_correct_type = self.key_transform(g1_str)
        except ValueError:
            self.key_did_not_match_specified_type__ERROR = True
            self.user_line_num_of_error = line_num_of_match + 1
            return

        line_content = editor.getLine(line_num_of_match).rstrip('\n\r')

        tup = (key_with_correct_type, line_content)
        self.key_and_line_content_tup_list.append(tup)

    def option_check(self, input_text, option_text):
        m = re.search(r'\[([^]]+)\] ?{opt}'.format(opt=option_text), input_text)
        retval = True if m and m.group(1) != ' ' * len(m.group(1)) else False
        return retval

    def search_regex_is_invalid_error_msg(self, test_regex):
        try:
            # test test_regex for validity
            editor.research(test_regex, lambda _: None)
        except RuntimeError as r:
            return str(r)
        return ''

    def mb(self, msg, flags=0, title=''):  # a message-box function
        return notepad.messageBox(msg, title if title else self.this_script_name, flags)

    def mb_ok_cancel(self, msg, title=''):  # returns True(OK) or False(Cancel)
        okay = notepad.messageBox(msg, title if title else self.this_script_name, MESSAGEBOXFLAGS.OKCANCEL) == MESSAGEBOXFLAGS.RESULTOK
        return okay

    def prompt(self, prompt_text, default_text=''):
        if '\n' not in prompt_text: prompt_text = '\r\n' + prompt_text
        prompt_text += ':'
        return notepad.prompt(prompt_text, self.this_script_name, default_text)

#-------------------------------------------------------------------------------

if __name__ == '__main__': SLWRG1ASK()

Maybe some day I’ll come up with a way to sort by arbitrary rearrangements of multiple groups like \2\L\1\E, but this is fine for now.

EDIT: I suppose that sorting by \2\L\1\E could be accomplished by doing the find/replace with \2\L\1\E🤷🤷$0 as replacement, then sort however you want, then find/replace ^.*🤷🤷 with nothing.

–
Moderator EDIT (2024-Jan-14): The author of the script has found a fairly serious bug with the code published here for those that use Mac-style or Linux-style line-endings in their files. The logic for Mac and Linux was reversed, and thus if the script was used on one type of file, the line-endings for the opposite type of file could end up in the file after the script is run. This is insidious, because unless one works with visible line-endings turned on, this is likely not noticed. Some detail on the problem is HERE. The script above has been corrected per that instruction.

Alan Kilborn

One thing I didn’t get around to doing in the script was making it easier on a user to specify a “decimal” key.

It probably would have been simpler for the user to specify a regex such as key(DECIMAL) rather than, e.g., key([-+]?[0-9]*\.?[0-9]+).

The script code could scan the user supplied regex for “DECIMAL” and replace with the somewhat hard-to-remember regex.

See HERE for the origins of that decimal-number regex.

Mark Olson

@Alan-Kilborn
Yep, good idea. I just added

NUMBER_RE = r'([-+]?[0-9]*\.?[0-9]+(?:[eE][-+]?\d+)?)'
INT_RE = r'([+-]?[0-9]+)'

and user_regex_with_group_1.replace('(INT)', INT_RE).replace('(DECIMAL)', NUMBER_RE) after the regex is assigned but before it’s used.

It’s easy to get the basic structure of integer and decimal regexes, but there are way too many edge cases that are too easily forgotten.

Mark Olson

alright this is my final version based on Alan Kilborn’s suggestions. I will restrain myself from posting any other versions until I have any substantial new improvements

# -*- coding: utf-8 -*-
from __future__ import print_function

#########################################
#
#  SortLinesWithRegexGroup1AsSortKey (SLWRG1ASK)
#
#########################################

# references:
#  https://community.notepad-plus-plus.org/topic/24325

#-------------------------------------------------------------------------------

from Npp import *
import inspect
import os
import re
import threading
import unicodedata
import SendKeys as sk

#-------------------------------------------------------------------------------

NUMBER_RE = r'([-+]?[0-9]*\.?[0-9]+(?:[eE][-+]?\d+)?)'
INT_RE = r'([+-]?[0-9]+)'

def ignorecase_culture_invariant(x):
    return unicodedata.normalize('NFD', x.lower())

class SLWRG1ASK(object):

    def __init__(self):

        self.this_script_name = inspect.getframeinfo(inspect.currentframe()).filename.split(os.sep)[-1].rsplit('.', 1)[0]

        self.key_and_line_content_tup_list = []
        self.matched_line_numbers = set()
        self.key_did_not_match_specified_type__ERROR = False
        self.user_did_not_specify_group1__ERROR = False
        self.not_all_lines_matched__ERROR = False
        self.last_matched_line_num = None
        self.user_line_num_of_error = None
        self.zero_matches_for_regex = True
        self.key_transform = None
        self.do_descending_sort = False
        self.each_line_must_have_regex_group1_key_match = True
        self.remove_lines_that_dont_match_regex = False
        self.sort_lines_that_dont_match_to_bottom = False

        rect_sel_mode = editor.getSelectionMode() in [ SELECTIONMODE.RECTANGLE, SELECTIONMODE.THIN ]
        if rect_sel_mode or editor.getSelections() > 1:
            self.mb('Cannot operate on a column selection or multiple selections; aborting!')
            return

        if len(editor.getSelText()) == 0:
            if not self.mb_ok_cancel('No selection active; confirm desire to sort entire file by pressing OK.'):
                return

        (search_start_line, search_end_line) = editor.getUserLineSelection()
        if editor.positionFromLine(search_end_line) == editor.getSelectionEnd(): search_end_line -= 1
        num_lines_in_user_sel = search_end_line - search_start_line + 1
        if num_lines_in_user_sel < 2: return  # if selection only on 1 line, nothing to do; already sorted!

        user_regex_with_group_1 = ''
        used_DECIMAL_RE = False
        while True:
            user_regex_with_group_1 = self.prompt('Enter regex with group 1 being the sort key.\r\nIf you wish, you can use (INT) to capture an integer or (DECIMAL) to capture a decimal', user_regex_with_group_1)
            if user_regex_with_group_1 is None: return  # user cancel
            if len(user_regex_with_group_1.strip()) == 0:
                self.mb('Cannot specify an empty regex!  Try again.')
                continue
            if '(' not in user_regex_with_group_1:
                self.mb('Need to specify capture group 1 (as the sort key) in the regex!  Try again.')
                continue
            regex_err_msg = self.search_regex_is_invalid_error_msg(user_regex_with_group_1)
            if len(regex_err_msg) > 0:
                self.mb('Bad regular expression!\r\n\r\n{e}\r\n\r\nTry again.'.format(e=regex_err_msg))
                continue
            used_DECIMAL_RE = '(DECIMAL)' in user_regex_with_group_1
            user_regex_with_group_1 = user_regex_with_group_1.replace('(INT)', INT_RE).replace('(DECIMAL)', NUMBER_RE)
            break

        while True:
            threading.Timer(0.25, lambda : sk.SendKeys("{RIGHT}")).start()  # clear auto-selection of prompt text when box appears
            user_options = self.prompt(
                '    Regex specified: {r}\r\n'
                'Choose options by placing 1 x in each group (no need to x to get DEFAULT choice)'.format(r=user_regex_with_group_1),
                '\r\n'.join([
                    'Group 1 transformation:  [   ]Integer(DEFAULT)      [  {0}]Decimal      [   ]String      [   ]Ignorecase-cultural'.format(' x'[used_DECIMAL_RE]),
                    'Sort order:  [   ]Ascending(DEFAULT)    [   ]Descending',
                    'Other:',
                    '[   ]Sort-only-if-all-lines-have-group-1-match(DEFAULT)',
                    '[   ]Remove-non-matching-lines',
                    '[   ]Sort-non-matching-lines-to-bottom',
                ])
                )
            if user_options is None: return  # user cancel
            if len(user_options.strip()) == 0:
                self.mb('Cannot specify empty options!  Try again.')
                continue
            if self.option_check(user_options, 'String'): self.key_transform = str
            elif self.option_check(user_options, 'Decimal'): self.key_transform = float
            elif self.option_check(user_options, 'Ignorecase-cultural'): self.key_transform = ignorecase_culture_invariant
            else: self.key_transform = int
            self.do_descending_sort = self.option_check(user_options, 'Descending')
            self.each_line_must_have_regex_group1_key_match = self.option_check(user_options, 'Sort-only-if-all-lines-have-group-1-match')
            self.remove_lines_that_dont_match_regex = self.option_check(user_options, 'Remove-non-matching-lines')
            self.sort_lines_that_dont_match_to_bottom = self.option_check(user_options, 'Sort-non-matching-lines-to-bottom')
            if not (self.remove_lines_that_dont_match_regex or self.sort_lines_that_dont_match_to_bottom):
                self.each_line_must_have_regex_group1_key_match = True
            if self.each_line_must_have_regex_group1_key_match + self.remove_lines_that_dont_match_regex + self.sort_lines_that_dont_match_to_bottom > 1:
                self.mb('Can only specify one way of handling unmatched lines! Try again!')
                continue
            break

        # enforce single-line matching and only get the FIRST match on each line:
        user_regex_with_group_1 = '(?-s)' + user_regex_with_group_1 + '.*'

        search_start_pos = editor.positionFromLine(search_start_line)
        search_end_pos = editor.getLineEndPosition(search_end_line)

        # special "seeding", in case the very first line of the selection isn't hit:
        self.last_matched_line_num = search_start_line - 1

        editor.research(user_regex_with_group_1, lambda m: self.match_found(m), 0, search_start_pos, search_end_pos)

        if self.zero_matches_for_regex:
            self.mb('No lines matched the regex; aborting!')
            return

        if self.user_did_not_specify_group1__ERROR:
            self.mb('No sort key specified (via using capture group 1) in the regex; aborting!')
            return

        if self.key_did_not_match_specified_type__ERROR:
            self.mb('Key data did not match specified type ({t}) on line {L}; aborting!'.format(L=self.user_line_num_of_error,
                t={ int : 'Integer', float : 'Decimal', str : 'String', ignorecase_culture_invariant: 'String' }[self.key_transform]))
            return

        if self.each_line_must_have_regex_group1_key_match and self.not_all_lines_matched__ERROR:
            self.mb('Not every line in the source data matched the specified regex (first non-matching line was {L}); aborting!'.format(L=self.user_line_num_of_error))
            return

        if len(self.key_and_line_content_tup_list) < 2:
            self.mb('Nothing (reasonable) found to sort; aborting!')
            return

        sorted_line_list = [tup[1] for tup in sorted(self.key_and_line_content_tup_list, reverse=self.do_descending_sort)]
        
        if self.matched_line_numbers:
            first_unmatched_line = search_start_line + len(self.matched_line_numbers) + 1
            self.mb('Line numbers {0}-{1} in sorted text were NOT matched.'.format(first_unmatched_line, search_end_line + 1),
                title='not all lines matched')
            for line_num in range(search_start_line, search_end_line + 1):
                if line_num not in self.matched_line_numbers:
                    unmatched_line = editor.getLine(line_num)
                    sorted_line_list.append(unmatched_line.rstrip('\r\n'))

        eol = ['\r\n', '\r', '\n'][editor.getEOLMode()]

        doc_len_before = editor.getLength()

        editor.setSelection(search_end_pos, search_start_pos)
        editor.replaceSel(eol.join(sorted_line_list))

        doc_len_delta = editor.getLength() - doc_len_before

        # leave sorted text selected:
        editor.setSelection(search_end_pos + doc_len_delta, search_start_pos)

    def match_found(self, m):

        self.zero_matches_for_regex = False

        # we're already corrupt; don't bother processing further matches after the one where we found the problem
        if self.user_did_not_specify_group1__ERROR: return
        if self.key_did_not_match_specified_type__ERROR: return
        if self.not_all_lines_matched__ERROR: return

        (start_pos_of_match, end_pos_of_match) = m.span(0)

        line_num_of_match = editor.lineFromPosition(start_pos_of_match)

        if line_num_of_match != self.last_matched_line_num + 1:
            if self.each_line_must_have_regex_group1_key_match:
                self.not_all_lines_matched__ERROR = True
                system_line_num_of_err = self.last_matched_line_num + 1
                self.user_line_num_of_error = system_line_num_of_err + 1
                return
        self.last_matched_line_num = line_num_of_match
        if self.sort_lines_that_dont_match_to_bottom:
            self.matched_line_numbers.add(line_num_of_match)

        try:
            g1_str = m.group(1)
        except IndexError:
            self.user_did_not_specify_group1__ERROR = True
            return

        try:
            key_with_correct_type = self.key_transform(g1_str)
        except ValueError:
            self.key_did_not_match_specified_type__ERROR = True
            self.user_line_num_of_error = line_num_of_match + 1
            return

        line_content = editor.getLine(line_num_of_match).rstrip('\n\r')

        tup = (key_with_correct_type, line_content)
        self.key_and_line_content_tup_list.append(tup)

    def option_check(self, input_text, option_text):
        m = re.search(r'\[([^]]+)\] ?{opt}'.format(opt=option_text), input_text)
        retval = True if m and m.group(1) != ' ' * len(m.group(1)) else False
        return retval

    def search_regex_is_invalid_error_msg(self, test_regex):
        try:
            # test test_regex for validity
            editor.research(test_regex, lambda _: None)
        except RuntimeError as r:
            return str(r)
        return ''

    def mb(self, msg, flags=0, title=''):  # a message-box function
        return notepad.messageBox(msg, title if title else self.this_script_name, flags)

    def mb_ok_cancel(self, msg, title=''):  # returns True(OK) or False(Cancel)
        okay = notepad.messageBox(msg, title if title else self.this_script_name, MESSAGEBOXFLAGS.OKCANCEL) == MESSAGEBOXFLAGS.RESULTOK
        return okay

    def prompt(self, prompt_text, default_text=''):
        if '\n' not in prompt_text: prompt_text = '\r\n' + prompt_text
        prompt_text += ':'
        return notepad.prompt(prompt_text, self.this_script_name, default_text)

#-------------------------------------------------------------------------------

if __name__ == '__main__': SLWRG1ASK()

–
Moderator EDIT (2024-Jan-14): The author of the script has found a fairly serious bug with the code published here for those that use Mac-style or Linux-style line-endings in their files. The logic for Mac and Linux was reversed, and thus if the script was used on one type of file, the line-endings for the opposite type of file could end up in the file after the script is run. This is insidious, because unless one works with visible line-endings turned on, this is likely not noticed. Some detail on the problem is HERE. The script above has been corrected per that instruction.

Mark Olson

@Mark-Olson
I discovered that my previous implementation of normalized case-insensitive text wasn’t quite right.

I can’t edit my last post, but the function ignorecase_culture_invariant should be replaced with this:

def ignorecase_culture_invariant(x):
    return unicodedata.normalize('NFD', x.upper())

The previous implementation (using x.lower()) doesn’t work for German ß, and there are probably some other places where it fails.

Mark Olson

Well, I said I would wait to post again until I had a significant improvement, and I believe I do.

I have now updated the script so it can sort multi-line blocks, for example, sorting a bunch of multi-line XML elements by some attribute.

Normally the only way to sort multi-line blocks is to do the following:

use a regex that replaces all the newlines in each block with some delimiter that is not present in the file while preserving newlines outside the block.
Sort the lines of the document using the above script or one of the built-in sorts
replace the delimiter with newlines.

This process is annoying, so I decided to automate it.

Behold!

# -*- coding: utf-8 -*-
from __future__ import print_function

#########################################
#
#  SortLinesWithRegexGroup1AsSortKey (SLWRG1ASK)
#
#########################################

# references:
#  https://community.notepad-plus-plus.org/topic/24325

#-------------------------------------------------------------------------------

from Npp import *
import inspect
import os
import re
import threading
import unicodedata
import SendKeys as sk

#-------------------------------------------------------------------------------

NUMBER_RE = r'([-+]?[0-9]*\.?[0-9]+(?:[eE][-+]?\d+)?)'
INT_RE = r'([+-]?[0-9]+)'

def ignorecase_culture_invariant(x):
    return unicodedata.normalize('NFD', x.upper())

class SLWRG1ASK(object):

    def __init__(self):

        self.this_script_name = inspect.getframeinfo(inspect.currentframe()).filename.split(os.sep)[-1].rsplit('.', 1)[0]

        self.key_and_line_content_tup_list = []
        self.matched_line_numbers = set()
        self.key_did_not_match_specified_type__ERROR = False
        self.user_did_not_specify_group1__ERROR = False
        self.not_all_lines_matched__ERROR = False
        self.last_matched_line_num = None
        self.user_line_num_of_error = None
        self.zero_matches_for_regex = True
        self.key_transform = None
        self.do_descending_sort = False
        self.each_line_must_have_regex_group1_key_match = True
        self.remove_lines_that_dont_match_regex = False
        self.sort_lines_that_dont_match_to_bottom = False
        self.multiline = False
        self.multiline_block_delim = ''
        self.eol = [ '\r\n', '\r', '\n' ][ editor.getEOLMode() ]

        rect_sel_mode = editor.getSelectionMode() in [ SELECTIONMODE.RECTANGLE, SELECTIONMODE.THIN ]
        if rect_sel_mode or editor.getSelections() > 1:
            self.mb('Cannot operate on a column selection or multiple selections; aborting!')
            return

        if len(editor.getSelText()) == 0:
            if not self.mb_ok_cancel('No selection active; confirm desire to sort entire file by pressing OK.'):
                return

        (search_start_line, search_end_line) = editor.getUserLineSelection()
        if editor.positionFromLine(search_end_line) == editor.getSelectionEnd(): search_end_line -= 1
        num_lines_in_user_sel = search_end_line - search_start_line + 1
        if num_lines_in_user_sel < 2: return  # if selection only on 1 line, nothing to do; already sorted!

        user_regex_with_group_1 = ''
        used_DECIMAL_RE = False
        while True:
            user_regex_with_group_1 = self.prompt('Enter regex with group 1 being the sort key.\r\nIf you wish, you can use (INT) to capture an integer or (DECIMAL) to capture a decimal', user_regex_with_group_1)
            if user_regex_with_group_1 is None: return  # user cancel
            if len(user_regex_with_group_1.strip()) == 0:
                self.mb('Cannot specify an empty regex!  Try again.')
                continue
            if '(' not in user_regex_with_group_1:
                self.mb('Need to specify capture group 1 (as the sort key) in the regex!  Try again.')
                continue
            regex_err_msg = self.search_regex_is_invalid_error_msg(user_regex_with_group_1)
            if len(regex_err_msg) > 0:
                self.mb('Bad regular expression!\r\n\r\n{e}\r\n\r\nTry again.'.format(e=regex_err_msg))
                continue
            used_DECIMAL_RE = '(DECIMAL)' in user_regex_with_group_1
            user_regex_with_group_1 = user_regex_with_group_1.replace('(INT)', INT_RE).replace('(DECIMAL)', NUMBER_RE)
            break

        while True:
            threading.Timer(0.25, lambda : sk.SendKeys("{RIGHT}")).start()  # clear auto-selection of prompt text when box appears
            user_options = self.prompt(
                '    Regex specified: {r}\r\n'
                'Choose options by placing 1 x in each group (no need to x to get DEFAULT choice)'.format(r=user_regex_with_group_1),
                '\r\n'.join([
                    'Group 1 transformation:  [   ]Integer(DEFAULT)      [  {0}]Decimal      [   ]String      [   ]Ignorecase-cultural'.format(' x'[used_DECIMAL_RE]),
                    'Sort order:  [   ]Ascending(DEFAULT)    [   ]Descending',
                    'Other:',
                    '[   ]Sort-only-if-all-lines-have-group-1-match(DEFAULT)',
                    '[   ]Remove-non-matching-lines',
                    '[   ]Sort-non-matching-lines-to-bottom',
                    '[   ]Multi-line-blocks',
                ])
                )
            if user_options is None: return  # user cancel
            if len(user_options.strip()) == 0:
                self.mb('Cannot specify empty options!  Try again.')
                continue
            if self.option_check(user_options, 'String'): self.key_transform = str
            elif self.option_check(user_options, 'Decimal'): self.key_transform = float
            elif self.option_check(user_options, 'Ignorecase-cultural'): self.key_transform = ignorecase_culture_invariant
            else: self.key_transform = int
            self.do_descending_sort = self.option_check(user_options, 'Descending')
            self.each_line_must_have_regex_group1_key_match = self.option_check(user_options, 'Sort-only-if-all-lines-have-group-1-match')
            self.remove_lines_that_dont_match_regex = self.option_check(user_options, 'Remove-non-matching-lines')
            self.sort_lines_that_dont_match_to_bottom = self.option_check(user_options, 'Sort-non-matching-lines-to-bottom')
            self.multiline = self.option_check(user_options, 'Multi-line')
            if not (self.remove_lines_that_dont_match_regex or self.sort_lines_that_dont_match_to_bottom):
                self.each_line_must_have_regex_group1_key_match = True
            if self.each_line_must_have_regex_group1_key_match + self.remove_lines_that_dont_match_regex + self.sort_lines_that_dont_match_to_bottom > 1:
                self.mb('Can only specify one way of handling unmatched lines! Try again!')
                continue
            break

        search_start_pos = editor.positionFromLine(search_start_line)
        search_end_pos = editor.getLineEndPosition(search_end_line)
        
        if self.multiline:
            # find a character (or multi-char sequence) that is not in the document
            delim = 1
            delim_len = 1
            allText = editor.getRangePointer(search_start_pos, search_end_pos)
            while True:
                # consider only NUL, SOH, STX, ETX, EOT, ENQ, ACK, BEL
                delim = delim % 8
                if delim == 0:
                    delim += 1
                delim_str = chr(delim) * delim_len
                if delim_str not in allText:
                    self.multiline_block_delim = delim_str
                    break
                else:
                    delim += 1
                    delim_len += 1
        else:
            # enforce single-line matching and only get the FIRST match on each line:
            user_regex_with_group_1 = '(?-s)' + user_regex_with_group_1 + '.*'


        # special "seeding", in case the very first line of the selection isn't hit:
        self.last_matched_line_num = search_start_line - 1

        editor.research(user_regex_with_group_1, lambda m: self.match_found(m), 0, search_start_pos, search_end_pos)

        if self.zero_matches_for_regex:
            self.mb('No lines matched the regex; aborting!')
            return

        if self.user_did_not_specify_group1__ERROR:
            self.mb('No sort key specified (via using capture group 1) in the regex; aborting!')
            return

        if self.key_did_not_match_specified_type__ERROR:
            self.mb('Key data did not match specified type ({t}) on line {L}; aborting!'.format(L=self.user_line_num_of_error,
                t={ int : 'Integer', float : 'Decimal', str : 'String', ignorecase_culture_invariant: 'String' }[self.key_transform]))
            return

        if self.each_line_must_have_regex_group1_key_match and self.not_all_lines_matched__ERROR:
            self.mb('Not every line in the source data matched the specified regex (first non-matching line was {L}) OR there were multiple matches in at least one line; aborting!'.format(L=self.user_line_num_of_error))
            return

        if len(self.key_and_line_content_tup_list) < 2:
            self.mb('Nothing (reasonable) found to sort; aborting!')
            return
        

        sorted_line_list = [tup[1] for tup in sorted(self.key_and_line_content_tup_list, reverse=self.do_descending_sort)]
        
        if self.multiline:
            sorted_line_list = [x.replace(self.multiline_block_delim, self.eol) for x in sorted_line_list]
        elif self.matched_line_numbers:
            first_unmatched_line = search_start_line + len(self.matched_line_numbers) + 1
            self.mb('Line numbers {0}-{1} in sorted text were NOT matched.'.format(first_unmatched_line, search_end_line + 1),
                title='not all lines matched')
            for line_num in range(search_start_line, search_end_line + 1):
                if line_num not in self.matched_line_numbers:
                    unmatched_line = editor.getLine(line_num)
                    sorted_line_list.append(unmatched_line.rstrip('\r\n'))

        doc_len_before = editor.getLength()

        editor.setSelection(search_end_pos, search_start_pos)
        editor.replaceSel(self.eol.join(sorted_line_list))

        doc_len_delta = editor.getLength() - doc_len_before

        # leave sorted text selected:
        editor.setSelection(search_end_pos + doc_len_delta, search_start_pos)

    def match_found(self, m):

        self.zero_matches_for_regex = False

        # we're already corrupt; don't bother processing further matches after the one where we found the problem
        if self.user_did_not_specify_group1__ERROR: return
        if self.key_did_not_match_specified_type__ERROR: return
        if self.not_all_lines_matched__ERROR: return

        (start_pos_of_match, end_pos_of_match) = m.span(0)
        line_num_of_match = editor.lineFromPosition(start_pos_of_match)
        
        try:
            g1_str = m.group(1)
        except IndexError:
            self.user_did_not_specify_group1__ERROR = True
            return

        try:
            key_with_correct_type = self.key_transform(g1_str)
        except ValueError:
            self.key_did_not_match_specified_type__ERROR = True
            self.user_line_num_of_error = line_num_of_match + 1
            return

        if self.multiline:
            line_content = m.group(0).rstrip('\n\r').replace(self.eol, self.multiline_block_delim)
        else:
            if line_num_of_match != self.last_matched_line_num + 1:
                if self.each_line_must_have_regex_group1_key_match:
                    self.not_all_lines_matched__ERROR = True
                    system_line_num_of_err = self.last_matched_line_num + 1
                    self.user_line_num_of_error = system_line_num_of_err + 1
                    return
            self.last_matched_line_num = line_num_of_match
            if self.sort_lines_that_dont_match_to_bottom:
                self.matched_line_numbers.add(line_num_of_match)
            line_content = editor.getLine(line_num_of_match).rstrip('\n\r')

        tup = (key_with_correct_type, line_content)
        self.key_and_line_content_tup_list.append(tup)

    def option_check(self, input_text, option_text):
        m = re.search(r'\\[([^]]+)\\] ?{opt}'.format(opt=option_text), input_text)
        retval = True if m and m.group(1) != ' ' * len(m.group(1)) else False
        return retval

    def search_regex_is_invalid_error_msg(self, test_regex):
        try:
            # test test_regex for validity
            editor.research(test_regex, lambda _: None)
        except RuntimeError as r:
            return str(r)
        return ''

    def mb(self, msg, flags=0, title=''):  # a message-box function
        return notepad.messageBox(msg, title if title else self.this_script_name, flags)

    def mb_ok_cancel(self, msg, title=''):  # returns True(OK) or False(Cancel)
        okay = notepad.messageBox(msg, title if title else self.this_script_name, MESSAGEBOXFLAGS.OKCANCEL) == MESSAGEBOXFLAGS.RESULTOK
        return okay

    def prompt(self, prompt_text, default_text=''):
        if '\n' not in prompt_text: prompt_text = '\r\n' + prompt_text
        prompt_text += ':'
        return notepad.prompt(prompt_text, self.this_script_name, default_text)

#-------------------------------------------------------------------------------

if __name__ == '__main__': SLWRG1ASK()

EXAMPLE: sorting the functions in a Python file

def foo(a) -> ruon:
    fdnfoen
    eorjerjeon
    if neoren:
        eorneor
    for orneor in reiroewh:
        orneroej
        def xjorneo(zser):
            ornero("defoorne")
        
def zargothrax(rinonrururr: int, bozo: float, 
defrnore: rourno) -> ziiron:
    oerjjeorno
def bar(b,
c,
d,
e):
    oerneorn
    erjelrkererei
    o33n4o3jo3443i3i
    043304 343kj3kj
    def noer(x):
        fnoerno
def uonreorn (x, b, z): ornepronp

def barru(rnx):
    pass
    
def ournoerno
  (fnorne
 ):
    pass

can have its functions sorted using the following regex: (?-si)^def\h+(\w+)\s*$[\w:\s,]+?$\s*(?:->[^:]+)?:(?:.+$|\h*\R(?:\x20+.*\R?)+) and the following options selected:

Group 1 transformation:  [   ]Integer(DEFAULT)      [   ]Decimal      [   ]String      [   x]Ignorecase-cultural
Sort order:  [   ]Ascending(DEFAULT)    [   ]Descending
Other:
[   ]Sort-only-if-all-lines-have-group-1-match(DEFAULT)
[   ]Remove-non-matching-lines
[   ]Sort-non-matching-lines-to-bottom
[  x ]Multi-line-blocks

Paul Wormer

@Mark-Olson said in How i can sort number from largest to smallest on notepad++:

(?-si)^def\h+(\w+)\s*([\w:\s,]+?)\s*(?:->[^:]+)?:(?:.+$|\h*\R(?:\x20+.*\R?)+)

I see that you use \h, \s and \x20 for a space in one and the same regex. Is there a deeper reason for that? Or is it didactic in that you want to point out that there are more possibilities for a space?

Mark Olson

@Paul-Wormer said in How i can sort number from largest to smallest on notepad++:

I see that you use \h, \s and \x20 for a space in one and the same regex. Is there a deeper reason for that?

Yes, there are reasons for all of those decisions:

\h between def and the name of the function because they must be on the same line but (I believe) there can be arbitrary spacing between them
\s in the region containing the arguments because the arguments can be separated by arbitrary whitespace
\x20 because I’m assuming that the function is indented by spaces.

Paul Wormer

@Mark-Olson OK, that is clear. Thank you for the explanation.

Alan Kilborn

If you’ve used a script in this thread, you might want to double check your copy of it for a bug I’ve discovered.
Look to previous postings in this topic thread where the script has been changed – find the text moderator edit (2024-Jan-14).
There’s a link there that describes the bug in more detail, and shows what needs to be changed in an old copy (or you can simply grab a copy of the current version).