From 1cb2af06431fa09cf32b74d6d7aab9bfbd722293 Mon Sep 17 00:00:00 2001
From: milde <milde@929543f6-e4f2-0310-98a6-ba3bd3dd1d04>
Date: Sun, 19 Nov 2023 09:44:17 +0000
Subject: [PATCH] More smartquotes optimizations.

Simplify regexps (we don't need to consider HTML character entities
when transforming a Docutils document tree).

git-svn-id: http://svn.code.sf.net/p/docutils/code/trunk@9480 929543f6-e4f2-0310-98a6-ba3bd3dd1d04
---
 docutils/HISTORY.txt                              |  5 ++++
 docutils/docutils/utils/smartquotes.py            | 36 +++++++++++------------
 docutils/test/test_transforms/test_smartquotes.py |  4 +--
 3 files changed, 24 insertions(+), 21 deletions(-)

diff --git a/docutils/HISTORY.txt b/docutils/HISTORY.txt
index 450373e7d..073a4406c 100644
--- a/docutils/HISTORY.txt
+++ b/docutils/HISTORY.txt
@@ -71,6 +71,11 @@ Changes since 0.20.1
   - Update to version `1.4 <https://pypi.org/project/roman/4.1/>`__.
     Fixes feature-requests:#95 (license is now ZPL 2.1).
 
+* docutils/utils/smartquotes.py
+
+  - Pre-compile regexps once, not with every call of `educateQuotes()`
+    (patch #206 by Chris Sewell). Simplify regexps.
+
 * docutils/writers/latex2e/__init__.py
 
   - Fix placement of hyperlink target (label) for tables (bug #440).
diff --git a/docutils/docutils/utils/smartquotes.py b/docutils/docutils/utils/smartquotes.py
index 7b2bc39b0..dc8beb17d 100755
--- a/docutils/docutils/utils/smartquotes.py
+++ b/docutils/docutils/utils/smartquotes.py
@@ -1,6 +1,6 @@
 #!/usr/bin/python3
 # :Id: $Id$
-# :Copyright: © 2010 Günter Milde,
+# :Copyright: © 2010-2023 Günter Milde,
 #             original `SmartyPants`_: © 2003 John Gruber
 #             smartypants.py:          © 2004, 2007 Chad Miller
 # :Maintainer: docutils-develop@lists.sourceforge.net
@@ -149,19 +149,13 @@ ASCII straight quotes, plain dashes, and plain dots, but your published
 posts (and final HTML output) will appear with smart quotes, em-dashes,
 and proper ellipses.
 
-SmartyPants does not modify characters within ``<pre>``, ``<code>``, ``<kbd>``,
-``<math>`` or ``<script>`` tag blocks. Typically, these tags are used to
-display text where smart quotes and other "smart punctuation" would not be
-appropriate, such as source code or example markup.
-
-
 Backslash Escapes
 =================
 
 If you need to use literal straight quotes (or plain hyphens and periods),
 `smartquotes` accepts the following backslash escape sequences to force
-ASCII-punctuation. Mind, that you need two backslashes as Docutils expands it,
-too.
+ASCII-punctuation. Mind, that you need two backslashes in "docstrings", as
+Python expands them, too.
 
 ========  =========
 Escape    Character
@@ -259,6 +253,10 @@ the source::
 Version History
 ===============
 
+1.10    2023-11-18
+        - Pre-compile regexps once, not with every call of `educateQuotes()`
+          (patch #206 by Chris Sewell). Simplify regexps.
+
 1.9     2022-03-04
         - Code cleanup. Require Python 3.
 
@@ -387,10 +385,10 @@ example, ``"1"`` is equivalent to ``"qBde"``.
 class smartchars:
     """Smart quotes and dashes"""
 
-    endash = '–'      # "&#8211;" EN DASH
-    emdash = '—'      # "&#8212;" EM DASH
-    ellipsis = '…'    # "&#8230;" HORIZONTAL ELLIPSIS
-    apostrophe = '’'  # "&#8217;" RIGHT SINGLE QUOTATION MARK
+    endash = '–'      # EN DASH
+    emdash = '—'      # EM DASH
+    ellipsis = '…'    # HORIZONTAL ELLIPSIS
+    apostrophe = '’'  # RIGHT SINGLE QUOTATION MARK
 
     # quote characters (language-specific, set in __init__())
     # https://en.wikipedia.org/wiki/Non-English_usage_of_quotation_marks
@@ -509,9 +507,9 @@ class RegularExpressions:
     # character classes:
     _CH_CLASSES = {'open': '[([{]',    # opening braces
                    'close': r'[^\s]',  # everything except whitespace
-                   'punct': r"""[-!" #\$\%'()*+,.\/:;<=>?\@\[\\\]\^_`{|}~]""",
-                   'dash': r'[-–—]|&[mn]dash;|&\#8211;|&\#8212;|&\#x201[34];',
-                   'sep': '[\\s\u200B\u200C]|&nbsp;',  # Whitespace, ZWSP, ZWNJ
+                   'punct': r"""[-!"  #\$\%'()*+,.\/:;<=>?\@\[\\\]\^_`{|}~]""",
+                   'dash': r'[-–—]',
+                   'sep': '[\\s\u200B\u200C]',  # Whitespace, ZWSP, ZWNJ
                    }
     START_SINGLE = re.compile(r"^'(?=%s\\B)" % _CH_CLASSES['punct'])
     START_DOUBLE = re.compile(r'^"(?=%s\\B)' % _CH_CLASSES['punct'])
@@ -679,7 +677,7 @@ def educateQuotes(text, language='en'):
     Returns:    The `text`, with "educated" curly quote characters.
 
     Example input:  "Isn't this fun?"
-    Example output: “Isn’t this fun?“;
+    Example output: “Isn’t this fun?“
     """
 
     smart = smartchars(language)
@@ -737,7 +735,7 @@ def educateBackticks(text, language='en'):
     Returns:    The `text`, with ``backticks'' -style double quotes
                 translated into HTML curly quote entities.
     Example input:  ``Isn't this fun?''
-    Example output: “Isn't this fun?“;
+    Example output: “Isn't this fun?“
     """
     smart = smartchars(language)
 
@@ -813,7 +811,7 @@ def educateEllipses(text):
                 an ellipsis character.
 
     Example input:  Huh...?
-    Example output: Huh&#8230;?
+    Example output: Huh…?
     """
 
     text = text.replace(r'...', smartchars.ellipsis)
diff --git a/docutils/test/test_transforms/test_smartquotes.py b/docutils/test/test_transforms/test_smartquotes.py
index 6153310bf..062bcd3c9 100644
--- a/docutils/test/test_transforms/test_smartquotes.py
+++ b/docutils/test/test_transforms/test_smartquotes.py
@@ -32,6 +32,8 @@ from docutils.utils import new_document
 
 
 class TransformTestCase(unittest.TestCase):
+    maxDiff = None
+
     def test_transforms(self):
         parser = Parser()
         settings = get_default_settings(Parser)
@@ -183,7 +185,6 @@ ZWNJ\u200C"a" and\u200C'a',
 escaped space\\ "a" and\\ 'a',
 
 hyphen -"a", -'a'
-&mdash;"a",&mdash;'a'
 en dash –"a"–'a',
 em dash —"a"—'a'.
 
@@ -210,7 +211,6 @@ But not if followed by (optional punctuation and) whitespace:
         escaped space“a” and‘a’,
     <paragraph>
         hyphen -“a”, -‘a’
-        &mdash;“a”,&mdash;‘a’
         en dash –“a”–‘a’,
         em dash —“a”—‘a’.
     <paragraph>
-- 
2.11.4.GIT