From ec5584d7b56b7a9c3f8411b48331a98c98010f8c Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Mon, 17 Feb 2025 11:28:44 +0530
Subject: [PATCH] Cache XPath() invocations and make insert preserving indent
 re-useable

---
 src/calibre/ebooks/oeb/base.py             |  2 ++
 src/calibre/ebooks/oeb/polish/container.py | 24 ++------------------
 src/calibre/ebooks/oeb/polish/utils.py     | 26 ++++++++++++++++++++++
 3 files changed, 30 insertions(+), 22 deletions(-)

diff --git a/src/calibre/ebooks/oeb/base.py b/src/calibre/ebooks/oeb/base.py
index e0e3d5a326..1192c870d0 100644
--- a/src/calibre/ebooks/oeb/base.py
+++ b/src/calibre/ebooks/oeb/base.py
@@ -12,6 +12,7 @@ import os
 import re
 import sys
 from collections import defaultdict
+from functools import lru_cache
 from itertools import count
 from operator import attrgetter
 
@@ -393,6 +394,7 @@ def isqname(name):
     return name and QNAME_RE.match(name) is not None
 
 
+@lru_cache(128)
 def XPath(expr):
     return etree.XPath(expr, namespaces=XPNSMAP)
 
diff --git a/src/calibre/ebooks/oeb/polish/container.py b/src/calibre/ebooks/oeb/polish/container.py
index 9e41a1f1a5..1a5b117f8f 100644
--- a/src/calibre/ebooks/oeb/polish/container.py
+++ b/src/calibre/ebooks/oeb/polish/container.py
@@ -47,7 +47,7 @@ from calibre.ebooks.oeb.base import (
 from calibre.ebooks.oeb.parse_utils import NotHTML, parse_html
 from calibre.ebooks.oeb.polish.errors import DRMError, InvalidBook
 from calibre.ebooks.oeb.polish.parsing import parse as parse_html_tweak
-from calibre.ebooks.oeb.polish.utils import OEB_FONTS, CommentFinder, PositionFinder, adjust_mime_for_epub, guess_type, parse_css
+from calibre.ebooks.oeb.polish.utils import OEB_FONTS, CommentFinder, PositionFinder, adjust_mime_for_epub, guess_type, insert_self_closing, parse_css
 from calibre.ptempfile import PersistentTemporaryDirectory, PersistentTemporaryFile
 from calibre.utils.filenames import hardlink_file, nlinks_file, retry_on_fail
 from calibre.utils.ipc.simple_worker import WorkerError, fork_job
@@ -942,27 +942,7 @@ class Container(ContainerBase):  # {{{
     def insert_into_xml(self, parent, item, index=None):
         '''Insert item into parent (or append if index is None), fixing
         indentation. Only works with self closing items.'''
-        if index is None:
-            parent.append(item)
-        else:
-            parent.insert(index, item)
-        idx = parent.index(item)
-        if idx == 0:
-            item.tail = parent.text
-            # If this is the only child of this parent element, we need a
-            # little extra work as we have gone from a self-closing <foo />
-            # element to <foo><item /></foo>
-            if len(parent) == 1:
-                sibling = parent.getprevious()
-                if sibling is None:
-                    # Give up!
-                    return
-                parent.text = sibling.text
-                item.tail = sibling.tail
-        else:
-            item.tail = parent[idx-1].tail
-            if idx == len(parent)-1:
-                parent[idx-1].tail = parent.text
+        insert_self_closing(parent, item, index)
 
     def opf_get_or_create(self, name):
         ''' Convenience method to either return the first XML element with the
diff --git a/src/calibre/ebooks/oeb/polish/utils.py b/src/calibre/ebooks/oeb/polish/utils.py
index 2313d8c28c..6cccf11d53 100644
--- a/src/calibre/ebooks/oeb/polish/utils.py
+++ b/src/calibre/ebooks/oeb/polish/utils.py
@@ -291,3 +291,29 @@ def extract(elem):
                 p[idx-1].tail = (p[idx-1].tail or '') + elem.tail
             else:
                 p.text = (p.text or '') + elem.tail
+
+
+def insert_self_closing(parent, item, index=None):
+    '''Insert item into parent (or append if index is None), fixing
+    indentation. Only works with self closing items.'''
+    if index is None:
+        parent.append(item)
+    else:
+        parent.insert(index, item)
+    idx = parent.index(item)
+    if idx == 0:
+        item.tail = parent.text
+        # If this is the only child of this parent element, we need a
+        # little extra work as we have gone from a self-closing <foo />
+        # element to <foo><item /></foo>
+        if len(parent) == 1:
+            sibling = parent.getprevious()
+            if sibling is None:
+                # Give up!
+                return
+            parent.text = sibling.text
+            item.tail = sibling.tail
+    else:
+        item.tail = parent[idx-1].tail
+        if idx == len(parent)-1:
+            parent[idx-1].tail = parent.text