From 2ce345e5dd9dba1f54f9c74b63b553877d4d003c Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Sun, 6 Oct 2019 13:04:07 +0700
Subject: [PATCH 01/18] FB2 Output: Speed up image data splitting

---
 src/calibre/ebooks/fb2/fb2ml.py | 12 +++---------
 1 file changed, 3 insertions(+), 9 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index ba1f775645..10ea49dc64 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -19,7 +19,7 @@ from calibre.constants import __appname__, __version__
 from calibre.utils.localization import lang_as_iso639_1
 from calibre.utils.img import save_cover_data_to
 from calibre.ebooks.oeb.base import urlnormalize
-from polyglot.builtins import unicode_type, string_or_bytes
+from polyglot.builtins import unicode_type, string_or_bytes, range
 from polyglot.binary import as_base64_unicode
 
 
@@ -315,14 +315,8 @@ class FB2MLizer(object):
                         raw_data = as_base64_unicode(item.data)
                         content_type = item.media_type
                     # Don't put the encoded image on a single line.
-                    data = ''
-                    col = 1
-                    for char in raw_data:
-                        if col == 72:
-                            data += '\n'
-                            col = 1
-                        col += 1
-                        data += char
+                    step = 72
+                    data = '\n'.join(raw_data[i:i+step] for i in range(0, len(raw_data), step))
                     images.append('<binary id="%s" content-type="%s">%s\n</binary>' % (self.image_hrefs[item.href], content_type, data))
                 except Exception as e:
                     self.log.error('Error: Could not include file %s because '

From 6071352d678fa1ad3dcd73f2c1c932a198a7a4c4 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Sun, 6 Oct 2019 13:28:45 +0700
Subject: [PATCH 02/18] FB2 Output: Fix id for PNG images

---
 src/calibre/ebooks/fb2/fb2ml.py | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 10ea49dc64..3a191cd54e 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -253,9 +253,8 @@ class FB2MLizer(object):
 
         if cover_href:
             # Only write the image tag if it is in the manifest.
-            if cover_href in self.oeb_book.manifest.hrefs.keys():
-                if cover_href not in self.image_hrefs.keys():
-                    self.image_hrefs[cover_href] = '_%s.jpg' % len(self.image_hrefs.keys())
+            if cover_href in self.oeb_book.manifest.hrefs and cover_href not in self.image_hrefs:
+                self.image_hrefs[cover_href] = 'img_%s' % len(self.image_hrefs)
             return '<coverpage><image xlink:href="#%s" /></coverpage>' % self.image_hrefs[cover_href]
 
         return ''
@@ -462,7 +461,7 @@ class FB2MLizer(object):
                 ihref = urlnormalize(page.abshref(elem_tree.attrib['src']))
                 if ihref in self.oeb_book.manifest.hrefs:
                     if ihref not in self.image_hrefs:
-                        self.image_hrefs[ihref] = '_%s.jpg' % len(self.image_hrefs)
+                        self.image_hrefs[ihref] = 'img_%s' % len(self.image_hrefs)
                     p_txt, p_tag = self.ensure_p()
                     fb2_out += p_txt
                     tags += p_tag

From 88159288a87d5e408932edae232378c077739753 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Sun, 6 Oct 2019 13:07:17 +0700
Subject: [PATCH 03/18] FB2 Output: Remove extra line breaks and prettify XML
 markup

---
 src/calibre/ebooks/fb2/fb2ml.py | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 3a191cd54e..91d05f0140 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -78,7 +78,7 @@ class FB2MLizer(object):
         # Clean up pargraph endings.
         text = re.sub(r'(?miu)\s*</p>', '</p>', text)
         # Put paragraphs following a paragraph on a separate line.
-        text = re.sub(r'(?miu)</p>\s*<p>', '</p>\n\n<p>', text)
+        text = re.sub(r'(?miu)</p>\s*<p>', '</p>\n<p>', text)
 
         # Remove empty title elements.
         text = re.sub(r'(?miu)<title>\s*</title>', '', text)
@@ -88,11 +88,11 @@ class FB2MLizer(object):
         text = re.sub(r'(?miu)<section>\s*</section>', '', text)
         # Clean up sections start and ends.
         text = re.sub(r'(?miu)\s*</section>', '\n</section>', text)
-        text = re.sub(r'(?miu)</section>\s*', '</section>\n\n', text)
+        text = re.sub(r'(?miu)</section>\s*', '</section>\n', text)
         text = re.sub(r'(?miu)\s*<section>', '\n<section>', text)
         text = re.sub(r'(?miu)<section>\s*', '<section>\n', text)
         # Put sectnions followed by sections on a separate line.
-        text = re.sub(r'(?miu)</section>\s*<section>', '</section>\n\n<section>', text)
+        text = re.sub(r'(?miu)</section>\s*<section>', '</section>\n<section>', text)
 
         if self.opts.insert_blank_line:
             text = re.sub(r'(?miu)</p>', '</p><empty-line />', text)
@@ -223,7 +223,7 @@ class FB2MLizer(object):
                 </description>\n''') % metadata
 
     def fb2_footer(self):
-        return '\n</FictionBook>'
+        return '</FictionBook>'
 
     def get_cover(self):
         from calibre.ebooks.oeb.base import OEB_RASTER_IMAGES
@@ -291,7 +291,8 @@ class FB2MLizer(object):
             text.append('</section>')
             self.section_level -= 1
 
-        return ''.join(text) + '</body>'
+        text.append('</body>')
+        return ''.join(text) + '\n'
 
     def fb2mlize_images(self):
         '''
@@ -316,11 +317,11 @@ class FB2MLizer(object):
                     # Don't put the encoded image on a single line.
                     step = 72
                     data = '\n'.join(raw_data[i:i+step] for i in range(0, len(raw_data), step))
-                    images.append('<binary id="%s" content-type="%s">%s\n</binary>' % (self.image_hrefs[item.href], content_type, data))
+                    images.append('<binary id="%s" content-type="%s">%s</binary>' % (self.image_hrefs[item.href], content_type, data))
                 except Exception as e:
                     self.log.error('Error: Could not include file %s because '
                         '%s.' % (item.href, e))
-        return ''.join(images)
+        return '\n'.join(images) + '\n'
 
     def create_flat_toc(self, nodes, level):
         for item in nodes:

From 0f5b72b4372c799d127221efb579787b3ada8d5d Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Sun, 6 Oct 2019 13:35:27 +0700
Subject: [PATCH 04/18] FB2 Output: Use short xlink namespace

---
 src/calibre/ebooks/fb2/fb2ml.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 91d05f0140..9d88d5554f 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -196,7 +196,7 @@ class FB2MLizer(object):
             metadata['comments'] = '<annotation>{}</annotation>'.format(prepare_string_for_xml(html2text(comments.value.strip())))
 
         return textwrap.dedent('''
-            <FictionBook xmlns="http://www.gribuser.ru/xml/fictionbook/2.0" xmlns:xlink="http://www.w3.org/1999/xlink">
+            <FictionBook xmlns="http://www.gribuser.ru/xml/fictionbook/2.0" xmlns:l="http://www.w3.org/1999/xlink">
                 <description>
                     <title-info>
                         <genre>%(genre)s</genre>
@@ -255,7 +255,7 @@ class FB2MLizer(object):
             # Only write the image tag if it is in the manifest.
             if cover_href in self.oeb_book.manifest.hrefs and cover_href not in self.image_hrefs:
                 self.image_hrefs[cover_href] = 'img_%s' % len(self.image_hrefs)
-            return '<coverpage><image xlink:href="#%s" /></coverpage>' % self.image_hrefs[cover_href]
+            return '<coverpage><image l:href="#%s" /></coverpage>' % self.image_hrefs[cover_href]
 
         return ''
 
@@ -466,7 +466,7 @@ class FB2MLizer(object):
                     p_txt, p_tag = self.ensure_p()
                     fb2_out += p_txt
                     tags += p_tag
-                    fb2_out.append('<image xlink:href="#%s" />' % self.image_hrefs[ihref])
+                    fb2_out.append('<image l:href="#%s" />' % self.image_hrefs[ihref])
                 else:
                     self.log.warn(u'Ignoring image not in manifest: %s'%ihref)
         if tag in ('br', 'hr') or ems >= 1:

From 906b8773c5005ffb1699f086a1d4fc346df3ffa4 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Sun, 6 Oct 2019 13:36:04 +0700
Subject: [PATCH 05/18] FB2 Output: Remove extra space in short tags

---
 src/calibre/ebooks/fb2/fb2ml.py | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 9d88d5554f..55349cd426 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -72,7 +72,7 @@ class FB2MLizer(object):
 
     def clean_text(self, text):
         # Condense empty paragraphs into a line break.
-        text = re.sub(r'(?miu)(<p>\s*</p>\s*){3,}', '<empty-line />', text)
+        text = re.sub(r'(?miu)(<p>\s*</p>\s*){3,}', '<empty-line/>', text)
         # Remove empty paragraphs.
         text = re.sub(r'(?miu)<p>\s*</p>', '', text)
         # Clean up pargraph endings.
@@ -95,7 +95,7 @@ class FB2MLizer(object):
         text = re.sub(r'(?miu)</section>\s*<section>', '</section>\n<section>', text)
 
         if self.opts.insert_blank_line:
-            text = re.sub(r'(?miu)</p>', '</p><empty-line />', text)
+            text = re.sub(r'(?miu)</p>', '</p><empty-line/>', text)
 
         return text
 
@@ -152,7 +152,7 @@ class FB2MLizer(object):
             index = '1'
             if self.oeb_book.metadata.series_index:
                 index = self.oeb_book.metadata.series_index[0]
-            metadata['sequence'] = '<sequence name="%s" number="%s" />' % (prepare_string_for_xml('%s' % self.oeb_book.metadata.series[0]), index)
+            metadata['sequence'] = '<sequence name="%s" number="%s"/>' % (prepare_string_for_xml('%s' % self.oeb_book.metadata.series[0]), index)
 
         year = publisher = isbn = ''
         identifiers = self.oeb_book.metadata['identifier']
@@ -255,7 +255,7 @@ class FB2MLizer(object):
             # Only write the image tag if it is in the manifest.
             if cover_href in self.oeb_book.manifest.hrefs and cover_href not in self.image_hrefs:
                 self.image_hrefs[cover_href] = 'img_%s' % len(self.image_hrefs)
-            return '<coverpage><image l:href="#%s" /></coverpage>' % self.image_hrefs[cover_href]
+            return '<coverpage><image l:href="#%s"/></coverpage>' % self.image_hrefs[cover_href]
 
         return ''
 
@@ -466,7 +466,7 @@ class FB2MLizer(object):
                     p_txt, p_tag = self.ensure_p()
                     fb2_out += p_txt
                     tags += p_tag
-                    fb2_out.append('<image l:href="#%s" />' % self.image_hrefs[ihref])
+                    fb2_out.append('<image l:href="#%s"/>' % self.image_hrefs[ihref])
                 else:
                     self.log.warn(u'Ignoring image not in manifest: %s'%ihref)
         if tag in ('br', 'hr') or ems >= 1:
@@ -483,12 +483,12 @@ class FB2MLizer(object):
                     closed_tags.append(t)
                     if t == 'p':
                         break
-                fb2_out.append('<empty-line />' * multiplier)
+                fb2_out.append('<empty-line/>' * multiplier)
                 closed_tags.reverse()
                 for t in closed_tags:
                     fb2_out.append('<%s>' % t)
             else:
-                fb2_out.append('<empty-line />' * multiplier)
+                fb2_out.append('<empty-line/>' * multiplier)
         if tag in ('div', 'li', 'p'):
             p_text, added_p = self.close_open_p(tag_stack+tags)
             fb2_out += p_text

From 62757eadffa79722aa1ddf78be95cfe1f0cec923 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Mon, 7 Oct 2019 12:22:07 +0700
Subject: [PATCH 06/18] FB2 Output: Remove empty lines from header

---
 src/calibre/ebooks/fb2/fb2ml.py | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 55349cd426..13fbba976d 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -19,7 +19,7 @@ from calibre.constants import __appname__, __version__
 from calibre.utils.localization import lang_as_iso639_1
 from calibre.utils.img import save_cover_data_to
 from calibre.ebooks.oeb.base import urlnormalize
-from polyglot.builtins import unicode_type, string_or_bytes, range
+from polyglot.builtins import unicode_type, string_or_bytes, range, filter
 from polyglot.binary import as_base64_unicode
 
 
@@ -195,12 +195,12 @@ class FB2MLizer(object):
             from calibre.utils.html2text import html2text
             metadata['comments'] = '<annotation>{}</annotation>'.format(prepare_string_for_xml(html2text(comments.value.strip())))
 
-        return textwrap.dedent('''
+        header = textwrap.dedent('''\
             <FictionBook xmlns="http://www.gribuser.ru/xml/fictionbook/2.0" xmlns:l="http://www.w3.org/1999/xlink">
                 <description>
                     <title-info>
                         <genre>%(genre)s</genre>
-                            %(author)s
+                        %(author)s
                         <book-title>%(title)s</book-title>
                         %(cover)s
                         <lang>%(lang)s</lang>
@@ -220,7 +220,10 @@ class FB2MLizer(object):
                         %(year)s
                         %(isbn)s
                     </publish-info>
-                </description>\n''') % metadata
+                </description>''') % metadata
+
+        # Remove empty lines.
+        return '\n'.join(filter(unicode_type.strip, header.splitlines())) + '\n'
 
     def fb2_footer(self):
         return '</FictionBook>'

From 58352097ad14daaecef33be46a8240f5261e325a Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Mon, 7 Oct 2019 12:27:00 +0700
Subject: [PATCH 07/18] FB2 Output: Proper join of XML parts

---
 src/calibre/ebooks/fb2/fb2ml.py | 24 +++++++++++++-----------
 1 file changed, 13 insertions(+), 11 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 13fbba976d..04d8ade536 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -59,16 +59,18 @@ class FB2MLizer(object):
         return self.fb2mlize_spine()
 
     def fb2mlize_spine(self):
-        output = [self.fb2_header()]
-        output.append(self.get_text())
-        output.append(self.fb2mlize_images())
-        output.append(self.fb2_footer())
-        output = self.clean_text(''.join(output))
+        output = (
+            self.fb2_header(),
+            self.get_text(),
+            self.fb2mlize_images(),
+            self.fb2_footer(),
+        )
+        output = self.clean_text('\n'.join(output))
 
         if self.opts.pretty_print:
-            return '<?xml version="1.0" encoding="UTF-8"?>\n%s' % etree.tostring(etree.fromstring(output), encoding='unicode', pretty_print=True)
-        else:
-            return '<?xml version="1.0" encoding="UTF-8"?>' + output
+            output = etree.tostring(etree.fromstring(output), encoding='unicode', pretty_print=True)
+
+        return '<?xml version="1.0" encoding="UTF-8"?>\n' + output
 
     def clean_text(self, text):
         # Condense empty paragraphs into a line break.
@@ -223,7 +225,7 @@ class FB2MLizer(object):
                 </description>''') % metadata
 
         # Remove empty lines.
-        return '\n'.join(filter(unicode_type.strip, header.splitlines())) + '\n'
+        return '\n'.join(filter(unicode_type.strip, header.splitlines()))
 
     def fb2_footer(self):
         return '</FictionBook>'
@@ -295,7 +297,7 @@ class FB2MLizer(object):
             self.section_level -= 1
 
         text.append('</body>')
-        return ''.join(text) + '\n'
+        return ''.join(text)
 
     def fb2mlize_images(self):
         '''
@@ -324,7 +326,7 @@ class FB2MLizer(object):
                 except Exception as e:
                     self.log.error('Error: Could not include file %s because '
                         '%s.' % (item.href, e))
-        return '\n'.join(images) + '\n'
+        return '\n'.join(images)
 
     def create_flat_toc(self, nodes, level):
         for item in nodes:

From d02966423c2201558494a3a9bd9698cf24741d11 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Mon, 7 Oct 2019 12:28:18 +0700
Subject: [PATCH 08/18] FB2 Output: Remove empty tags

---
 src/calibre/ebooks/fb2/fb2ml.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 04d8ade536..4421623869 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -73,6 +73,8 @@ class FB2MLizer(object):
         return '<?xml version="1.0" encoding="UTF-8"?>\n' + output
 
     def clean_text(self, text):
+        # Remove empty tags.
+        text = re.sub(r'(?miu)<(strong|emphasis|strikethrough|sub|sup)>\s*</\1>', '', text)
         # Condense empty paragraphs into a line break.
         text = re.sub(r'(?miu)(<p>\s*</p>\s*){3,}', '<empty-line/>', text)
         # Remove empty paragraphs.

From 0c1ca17192fa408d7d731d07b9b44158268aaccc Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Mon, 7 Oct 2019 12:28:49 +0700
Subject: [PATCH 09/18] FB2 Output: Annotation should be wrapped in p

---
 src/calibre/ebooks/fb2/fb2ml.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 4421623869..30791c8ed3 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -197,7 +197,7 @@ class FB2MLizer(object):
             metadata['comments'] = ''
         else:
             from calibre.utils.html2text import html2text
-            metadata['comments'] = '<annotation>{}</annotation>'.format(prepare_string_for_xml(html2text(comments.value.strip())))
+            metadata['comments'] = '<annotation><p>{}</p></annotation>'.format(prepare_string_for_xml(html2text(comments.value).strip()))
 
         header = textwrap.dedent('''\
             <FictionBook xmlns="http://www.gribuser.ru/xml/fictionbook/2.0" xmlns:l="http://www.w3.org/1999/xlink">

From 8b1194730923c308f5fe073fbd4a25fd773d8291 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Mon, 7 Oct 2019 17:48:17 +0700
Subject: [PATCH 10/18] FB2 Output: Keep contents of empty tags if they placed
 between words

---
 src/calibre/ebooks/fb2/fb2ml.py | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 30791c8ed3..2a408c3f94 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -73,8 +73,9 @@ class FB2MLizer(object):
         return '<?xml version="1.0" encoding="UTF-8"?>\n' + output
 
     def clean_text(self, text):
-        # Remove empty tags.
-        text = re.sub(r'(?miu)<(strong|emphasis|strikethrough|sub|sup)>\s*</\1>', '', text)
+        # Remove pointless tags, but keep their contents.
+        text = re.sub(r'(?miu)<(strong|emphasis|strikethrough|sub|sup)>(\s*)</\1>', r'\2', text)
+
         # Condense empty paragraphs into a line break.
         text = re.sub(r'(?miu)(<p>\s*</p>\s*){3,}', '<empty-line/>', text)
         # Remove empty paragraphs.

From abe408bc3943f46fa083cf1f5a86fb71edb15388 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Tue, 8 Oct 2019 11:11:13 +0700
Subject: [PATCH 11/18] FB2 Output: Dedent the header to position it on the
 same level as the body

---
 src/calibre/ebooks/fb2/fb2ml.py | 49 +++++++++++++++++----------------
 1 file changed, 25 insertions(+), 24 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 2a408c3f94..a9d604da65 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -200,32 +200,33 @@ class FB2MLizer(object):
             from calibre.utils.html2text import html2text
             metadata['comments'] = '<annotation><p>{}</p></annotation>'.format(prepare_string_for_xml(html2text(comments.value).strip()))
 
+        # Keep the indentation level of the description the same as the body.
         header = textwrap.dedent('''\
             <FictionBook xmlns="http://www.gribuser.ru/xml/fictionbook/2.0" xmlns:l="http://www.w3.org/1999/xlink">
-                <description>
-                    <title-info>
-                        <genre>%(genre)s</genre>
-                        %(author)s
-                        <book-title>%(title)s</book-title>
-                        %(cover)s
-                        <lang>%(lang)s</lang>
-                        %(keywords)s
-                        %(sequence)s
-                        %(comments)s
-                    </title-info>
-                    <document-info>
-                        %(author)s
-                        <program-used>%(appname)s %(version)s</program-used>
-                        <date>%(date)s</date>
-                        <id>%(id)s</id>
-                        <version>1.0</version>
-                    </document-info>
-                    <publish-info>
-                        %(publisher)s
-                        %(year)s
-                        %(isbn)s
-                    </publish-info>
-                </description>''') % metadata
+            <description>
+                <title-info>
+                    <genre>%(genre)s</genre>
+                    %(author)s
+                    <book-title>%(title)s</book-title>
+                    %(cover)s
+                    <lang>%(lang)s</lang>
+                    %(keywords)s
+                    %(sequence)s
+                    %(comments)s
+                </title-info>
+                <document-info>
+                    %(author)s
+                    <program-used>%(appname)s %(version)s</program-used>
+                    <date>%(date)s</date>
+                    <id>%(id)s</id>
+                    <version>1.0</version>
+                </document-info>
+                <publish-info>
+                    %(publisher)s
+                    %(year)s
+                    %(isbn)s
+                </publish-info>
+            </description>''') % metadata
 
         # Remove empty lines.
         return '\n'.join(filter(unicode_type.strip, header.splitlines()))

From 3b8d769128922a939ca7b6c72578d1eaad779b95 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Tue, 8 Oct 2019 12:16:06 +0700
Subject: [PATCH 12/18] FB2 Output: Simplify regular expressions

---
 src/calibre/ebooks/fb2/fb2ml.py | 27 ++++++++++++++-------------
 1 file changed, 14 insertions(+), 13 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index a9d604da65..a5745e1bc2 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -76,27 +76,28 @@ class FB2MLizer(object):
         # Remove pointless tags, but keep their contents.
         text = re.sub(r'(?miu)<(strong|emphasis|strikethrough|sub|sup)>(\s*)</\1>', r'\2', text)
 
+        # Clean up paragraphs endings.
+        text = re.sub(r'(?miu)\s+</p>', '</p>', text)
         # Condense empty paragraphs into a line break.
-        text = re.sub(r'(?miu)(<p>\s*</p>\s*){3,}', '<empty-line/>', text)
+        text = re.sub(r'(?miu)(?:<p></p>\s*){3,}', '<empty-line/>', text)
         # Remove empty paragraphs.
-        text = re.sub(r'(?miu)<p>\s*</p>', '', text)
-        # Clean up pargraph endings.
-        text = re.sub(r'(?miu)\s*</p>', '</p>', text)
-        # Put paragraphs following a paragraph on a separate line.
+        text = re.sub(r'(?miu)<p></p>', '', text)
+        # Put the paragraph following a paragraph on a separate line.
         text = re.sub(r'(?miu)</p>\s*<p>', '</p>\n<p>', text)
 
-        # Remove empty title elements.
-        text = re.sub(r'(?miu)<title>\s*</title>', '', text)
+        # Clean up title endings.
         text = re.sub(r'(?miu)\s+</title>', '</title>', text)
+        # Remove empty title elements.
+        text = re.sub(r'(?miu)<title></title>', '', text)
+        # Put the paragraph following a title on a separate line.
+        text = re.sub(r'(?miu)</title>\s*<p>', '</title>\n<p>', text)
 
         # Remove empty sections.
         text = re.sub(r'(?miu)<section>\s*</section>', '', text)
-        # Clean up sections start and ends.
-        text = re.sub(r'(?miu)\s*</section>', '\n</section>', text)
-        text = re.sub(r'(?miu)</section>\s*', '</section>\n', text)
-        text = re.sub(r'(?miu)\s*<section>', '\n<section>', text)
-        text = re.sub(r'(?miu)<section>\s*', '<section>\n', text)
-        # Put sectnions followed by sections on a separate line.
+        # Clean up sections starts and ends.
+        text = re.sub(r'(?miu)\s*<section>\s*', '\n<section>\n', text)
+        text = re.sub(r'(?miu)\s*</section>\s*', '\n</section>\n', text)
+        # Put the section following a section on a separate line.
         text = re.sub(r'(?miu)</section>\s*<section>', '</section>\n<section>', text)
 
         if self.opts.insert_blank_line:

From cb45766216929a28fc21e128f963006f37a1ffb9 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Tue, 8 Oct 2019 12:16:35 +0700
Subject: [PATCH 13/18] FB2 Output: Put line breaks between paragraphs on a
 separate line

---
 src/calibre/ebooks/fb2/fb2ml.py | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index a5745e1bc2..433db22fa1 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -103,6 +103,10 @@ class FB2MLizer(object):
         if self.opts.insert_blank_line:
             text = re.sub(r'(?miu)</p>', '</p><empty-line/>', text)
 
+        # Put line breaks between paragraphs on a separate line.
+        text = re.sub(r'(?miu)</(p|title)>\s*<empty-line/>', r'</\1>\n<empty-line/>', text)
+        text = re.sub(r'(?miu)<empty-line/>\s*<p>', '<empty-line/>\n<p>', text)
+
         return text
 
     def fb2_header(self):

From 162d86429deb95df251e2cf186e8199b1ed2013c Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Tue, 8 Oct 2019 12:19:02 +0700
Subject: [PATCH 14/18] FB2 Output: Ignore case flag is not required because
 the writer outputs only lowercase tag names

---
 src/calibre/ebooks/fb2/fb2ml.py | 30 +++++++++++++++---------------
 1 file changed, 15 insertions(+), 15 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 433db22fa1..5697f2262a 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -74,38 +74,38 @@ class FB2MLizer(object):
 
     def clean_text(self, text):
         # Remove pointless tags, but keep their contents.
-        text = re.sub(r'(?miu)<(strong|emphasis|strikethrough|sub|sup)>(\s*)</\1>', r'\2', text)
+        text = re.sub(r'(?mu)<(strong|emphasis|strikethrough|sub|sup)>(\s*)</\1>', r'\2', text)
 
         # Clean up paragraphs endings.
-        text = re.sub(r'(?miu)\s+</p>', '</p>', text)
+        text = re.sub(r'(?mu)\s+</p>', '</p>', text)
         # Condense empty paragraphs into a line break.
-        text = re.sub(r'(?miu)(?:<p></p>\s*){3,}', '<empty-line/>', text)
+        text = re.sub(r'(?mu)(?:<p></p>\s*){3,}', '<empty-line/>', text)
         # Remove empty paragraphs.
-        text = re.sub(r'(?miu)<p></p>', '', text)
+        text = re.sub(r'(?mu)<p></p>', '', text)
         # Put the paragraph following a paragraph on a separate line.
-        text = re.sub(r'(?miu)</p>\s*<p>', '</p>\n<p>', text)
+        text = re.sub(r'(?mu)</p>\s*<p>', '</p>\n<p>', text)
 
         # Clean up title endings.
-        text = re.sub(r'(?miu)\s+</title>', '</title>', text)
+        text = re.sub(r'(?mu)\s+</title>', '</title>', text)
         # Remove empty title elements.
-        text = re.sub(r'(?miu)<title></title>', '', text)
+        text = re.sub(r'(?mu)<title></title>', '', text)
         # Put the paragraph following a title on a separate line.
-        text = re.sub(r'(?miu)</title>\s*<p>', '</title>\n<p>', text)
+        text = re.sub(r'(?mu)</title>\s*<p>', '</title>\n<p>', text)
 
         # Remove empty sections.
-        text = re.sub(r'(?miu)<section>\s*</section>', '', text)
+        text = re.sub(r'(?mu)<section>\s*</section>', '', text)
         # Clean up sections starts and ends.
-        text = re.sub(r'(?miu)\s*<section>\s*', '\n<section>\n', text)
-        text = re.sub(r'(?miu)\s*</section>\s*', '\n</section>\n', text)
+        text = re.sub(r'(?mu)\s*<section>\s*', '\n<section>\n', text)
+        text = re.sub(r'(?mu)\s*</section>\s*', '\n</section>\n', text)
         # Put the section following a section on a separate line.
-        text = re.sub(r'(?miu)</section>\s*<section>', '</section>\n<section>', text)
+        text = re.sub(r'(?mu)</section>\s*<section>', '</section>\n<section>', text)
 
         if self.opts.insert_blank_line:
-            text = re.sub(r'(?miu)</p>', '</p><empty-line/>', text)
+            text = re.sub(r'(?mu)</p>', '</p><empty-line/>', text)
 
         # Put line breaks between paragraphs on a separate line.
-        text = re.sub(r'(?miu)</(p|title)>\s*<empty-line/>', r'</\1>\n<empty-line/>', text)
-        text = re.sub(r'(?miu)<empty-line/>\s*<p>', '<empty-line/>\n<p>', text)
+        text = re.sub(r'(?mu)</(p|title)>\s*<empty-line/>', r'</\1>\n<empty-line/>', text)
+        text = re.sub(r'(?mu)<empty-line/>\s*<p>', '<empty-line/>\n<p>', text)
 
         return text
 

From d7a4e10c6022d13a17f87ae304d4b6ff84e02c17 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Tue, 8 Oct 2019 15:57:13 +0700
Subject: [PATCH 15/18] FB2 Output: Revert some changes in regular expressions
 due to different behavior

---
 src/calibre/ebooks/fb2/fb2ml.py | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 5697f2262a..12dd539e53 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -81,22 +81,24 @@ class FB2MLizer(object):
         # Condense empty paragraphs into a line break.
         text = re.sub(r'(?mu)(?:<p></p>\s*){3,}', '<empty-line/>', text)
         # Remove empty paragraphs.
-        text = re.sub(r'(?mu)<p></p>', '', text)
+        text = re.sub(r'(?mu)<p></p>\s*', '', text)
         # Put the paragraph following a paragraph on a separate line.
         text = re.sub(r'(?mu)</p>\s*<p>', '</p>\n<p>', text)
 
         # Clean up title endings.
         text = re.sub(r'(?mu)\s+</title>', '</title>', text)
         # Remove empty title elements.
-        text = re.sub(r'(?mu)<title></title>', '', text)
+        text = re.sub(r'(?mu)<title></title>\s*', '', text)
         # Put the paragraph following a title on a separate line.
         text = re.sub(r'(?mu)</title>\s*<p>', '</title>\n<p>', text)
 
         # Remove empty sections.
         text = re.sub(r'(?mu)<section>\s*</section>', '', text)
         # Clean up sections starts and ends.
-        text = re.sub(r'(?mu)\s*<section>\s*', '\n<section>\n', text)
-        text = re.sub(r'(?mu)\s*</section>\s*', '\n</section>\n', text)
+        text = re.sub(r'(?mu)\s*<section>', '\n<section>', text)
+        text = re.sub(r'(?mu)<section>\s*', '<section>\n', text)
+        text = re.sub(r'(?mu)\s*</section>', '\n</section>', text)
+        text = re.sub(r'(?mu)</section>\s*', '</section>\n', text)
         # Put the section following a section on a separate line.
         text = re.sub(r'(?mu)</section>\s*<section>', '</section>\n<section>', text)
 

From e7175e802b4a48e280bb495fd98acedaa3e5fc8d Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Tue, 8 Oct 2019 16:10:33 +0700
Subject: [PATCH 16/18] FB2 Output: Remove section regexp what do nothing and
 rearrange regexps

---
 src/calibre/ebooks/fb2/fb2ml.py | 16 +++++++---------
 1 file changed, 7 insertions(+), 9 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 12dd539e53..197f063baa 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -85,6 +85,9 @@ class FB2MLizer(object):
         # Put the paragraph following a paragraph on a separate line.
         text = re.sub(r'(?mu)</p>\s*<p>', '</p>\n<p>', text)
 
+        if self.opts.insert_blank_line:
+            text = re.sub(r'(?mu)</p>', '</p><empty-line/>', text)
+
         # Clean up title endings.
         text = re.sub(r'(?mu)\s+</title>', '</title>', text)
         # Remove empty title elements.
@@ -92,6 +95,10 @@ class FB2MLizer(object):
         # Put the paragraph following a title on a separate line.
         text = re.sub(r'(?mu)</title>\s*<p>', '</title>\n<p>', text)
 
+        # Put line breaks between paragraphs on a separate line.
+        text = re.sub(r'(?mu)</(p|title)>\s*<empty-line/>', r'</\1>\n<empty-line/>', text)
+        text = re.sub(r'(?mu)<empty-line/>\s*<p>', '<empty-line/>\n<p>', text)
+
         # Remove empty sections.
         text = re.sub(r'(?mu)<section>\s*</section>', '', text)
         # Clean up sections starts and ends.
@@ -99,15 +106,6 @@ class FB2MLizer(object):
         text = re.sub(r'(?mu)<section>\s*', '<section>\n', text)
         text = re.sub(r'(?mu)\s*</section>', '\n</section>', text)
         text = re.sub(r'(?mu)</section>\s*', '</section>\n', text)
-        # Put the section following a section on a separate line.
-        text = re.sub(r'(?mu)</section>\s*<section>', '</section>\n<section>', text)
-
-        if self.opts.insert_blank_line:
-            text = re.sub(r'(?mu)</p>', '</p><empty-line/>', text)
-
-        # Put line breaks between paragraphs on a separate line.
-        text = re.sub(r'(?mu)</(p|title)>\s*<empty-line/>', r'</\1>\n<empty-line/>', text)
-        text = re.sub(r'(?mu)<empty-line/>\s*<p>', '<empty-line/>\n<p>', text)
 
         return text
 

From c42d2db421e7532dd0b3c5711b8c3ba805ef52b3 Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Wed, 9 Oct 2019 12:13:25 +0700
Subject: [PATCH 17/18] FB2 Output: Handle external links

---
 src/calibre/ebooks/fb2/fb2ml.py | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 197f063baa..7198d0f321 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -21,12 +21,13 @@ from calibre.utils.img import save_cover_data_to
 from calibre.ebooks.oeb.base import urlnormalize
 from polyglot.builtins import unicode_type, string_or_bytes, range, filter
 from polyglot.binary import as_base64_unicode
+from polyglot.urllib import urlparse
 
 
 class FB2MLizer(object):
     '''
     Todo: * Include more FB2 specific tags in the conversion.
-          * Handle a tags.
+          * Handle notes and anchor links.
     '''
 
     def __init__(self, log):
@@ -508,6 +509,14 @@ class FB2MLizer(object):
             fb2_out += p_text
             if added_p:
                 tags.append('p')
+        if tag == 'a' and elem_tree.attrib.get('href', None):
+            # Handle only external links for now
+            if urlparse(elem_tree.attrib['href']).netloc:
+                p_txt, p_tag = self.ensure_p()
+                fb2_out += p_txt
+                tags += p_tag
+                fb2_out.append('<a l:href="%s">' % urlnormalize(elem_tree.attrib['href']))
+                tags.append('a')
         if tag == 'b' or style['font-weight'] in ('bold', 'bolder'):
             s_out, s_tags = self.handle_simple_tag('strong', tag_stack+tags)
             fb2_out += s_out

From 3e680ac307023d962a41b84f62fc7c2f72a3068a Mon Sep 17 00:00:00 2001
From: Andrey Efremov <duxus@yandex.ru>
Date: Wed, 9 Oct 2019 12:18:31 +0700
Subject: [PATCH 18/18] FB2 Output: Merge ifs

---
 src/calibre/ebooks/fb2/fb2ml.py | 25 ++++++++++++-------------
 1 file changed, 12 insertions(+), 13 deletions(-)

diff --git a/src/calibre/ebooks/fb2/fb2ml.py b/src/calibre/ebooks/fb2/fb2ml.py
index 7198d0f321..71573bac8b 100644
--- a/src/calibre/ebooks/fb2/fb2ml.py
+++ b/src/calibre/ebooks/fb2/fb2ml.py
@@ -471,19 +471,18 @@ class FB2MLizer(object):
         # Process the XHTML tag and styles. Converted to an FB2 tag.
         # Use individual if statement not if else. There can be
         # only one XHTML tag but it can have multiple styles.
-        if tag == 'img':
-            if elem_tree.attrib.get('src', None):
-                # Only write the image tag if it is in the manifest.
-                ihref = urlnormalize(page.abshref(elem_tree.attrib['src']))
-                if ihref in self.oeb_book.manifest.hrefs:
-                    if ihref not in self.image_hrefs:
-                        self.image_hrefs[ihref] = 'img_%s' % len(self.image_hrefs)
-                    p_txt, p_tag = self.ensure_p()
-                    fb2_out += p_txt
-                    tags += p_tag
-                    fb2_out.append('<image l:href="#%s"/>' % self.image_hrefs[ihref])
-                else:
-                    self.log.warn(u'Ignoring image not in manifest: %s'%ihref)
+        if tag == 'img' and elem_tree.attrib.get('src', None):
+            # Only write the image tag if it is in the manifest.
+            ihref = urlnormalize(page.abshref(elem_tree.attrib['src']))
+            if ihref in self.oeb_book.manifest.hrefs:
+                if ihref not in self.image_hrefs:
+                    self.image_hrefs[ihref] = 'img_%s' % len(self.image_hrefs)
+                p_txt, p_tag = self.ensure_p()
+                fb2_out += p_txt
+                tags += p_tag
+                fb2_out.append('<image l:href="#%s"/>' % self.image_hrefs[ihref])
+            else:
+                self.log.warn(u'Ignoring image not in manifest: %s' % ihref)
         if tag in ('br', 'hr') or ems >= 1:
             if ems < 1:
                 multiplier = 1