From eae370b75207e1dc15f5f3f7ee7a402c7721d9a8 Mon Sep 17 00:00:00 2001 From: Radek Czajka Date: Mon, 20 Dec 2021 11:50:38 +0100 Subject: [PATCH 1/1] New mobi builder. --- src/librarian/builders/__init__.py | 2 + src/librarian/builders/epub.py | 21 ++++----- src/librarian/builders/mobi.py | 45 +++++++++++++++++++ src/librarian/elements/__init__.py | 1 - src/librarian/elements/base.py | 27 +++++++---- src/librarian/elements/styles/tytul_dziela.py | 4 +- src/librarian/hyphenator.py | 8 ++-- 7 files changed, 82 insertions(+), 26 deletions(-) create mode 100644 src/librarian/builders/mobi.py diff --git a/src/librarian/builders/__init__.py b/src/librarian/builders/__init__.py index e359cd6..d8acb82 100644 --- a/src/librarian/builders/__init__.py +++ b/src/librarian/builders/__init__.py @@ -4,6 +4,7 @@ from .html import HtmlBuilder, StandaloneHtmlBuilder, DaisyHtmlBuilder from .sanitize import Sanitizer from .daisy import DaisyBuilder from .epub import EpubBuilder +from .mobi import MobiBuilder from .pdf import PdfBuilder @@ -16,5 +17,6 @@ builders = OrderedDict([ ("sanitizer", Sanitizer), ("epub", EpubBuilder), + ("mobi", MobiBuilder), ("pdf", PdfBuilder), ]) diff --git a/src/librarian/builders/epub.py b/src/librarian/builders/epub.py index 4471e30..401136f 100644 --- a/src/librarian/builders/epub.py +++ b/src/librarian/builders/epub.py @@ -29,10 +29,11 @@ class Xhtml: class Builder: file_extension = None - def __init__(self, base_url=None, fundraising=None): + def __init__(self, base_url=None, fundraising=None, cover=None): self._base_url = base_url or 'file:///home/rczajka/for/fnp/librarian/temp~/maly/img/' self.fundraising = fundraising self.footnotes = etree.Element('div', id='footnotes') + self.make_cover = cover or make_cover self.cursors = { # None: None, @@ -78,6 +79,7 @@ class Builder: class EpubBuilder(Builder): file_extension = 'epub' + isbn_field = 'isbn_epub' def __init__(self, *args, **kwargs): self.chars = set() @@ -109,7 +111,6 @@ class EpubBuilder(Builder): self.set_metadata() - self.add_cover() self.add_title_page() @@ -212,10 +213,10 @@ class EpubBuilder(Builder): e = self.document.tree.find('//autor_utworu') if e is not None: - etree.SubElement(tp, 'h2', **{'class': 'author'}).text = e.raw_printable_text() + etree.SubElement(tp, 'h2', **{'class': 'author'}).text = e.raw_printable_text(self) e = self.document.tree.find('//nazwa_utworu') if e is not None: - etree.SubElement(tp, 'h1', **{'class': 'title'}).text = e.raw_printable_text() + etree.SubElement(tp, 'h1', **{'class': 'title'}).text = e.raw_printable_text(self) if not len(tp): for author in self.document.meta.authors: @@ -251,8 +252,8 @@ class EpubBuilder(Builder):

""")) - if self.document.meta.isbn_epub: - etree.SubElement(tp, 'p', **{"class": "info"}).text = self.document.meta.isbn_epub + if getattr(self.document.meta, self.isbn_field): + etree.SubElement(tp, 'p', **{"class": "info"}).text = getattr(self.document.meta, self.isbn_field) tp.append(etree.XML("""

WolneLektury.pl @@ -589,8 +590,8 @@ class EpubBuilder(Builder): else: p.text += m.cover_by - if m.isbn_epub: - newp().text = m.isbn_epub + if getattr(m, self.isbn_field): + newp().text = getattr(m, self.isbn_field) newp().text = '\u00a0' @@ -644,10 +645,10 @@ class EpubBuilder(Builder): def add_cover(self): # TODO: allow other covers - cover_maker = make_cover + cover_maker = self.make_cover cover_file = six.BytesIO() - cover = cover_maker(self.document.meta) + cover = cover_maker(self.document.meta, width=600) cover.save(cover_file) cover_name = 'cover.%s' % cover.ext() diff --git a/src/librarian/builders/mobi.py b/src/librarian/builders/mobi.py new file mode 100644 index 0000000..19b5036 --- /dev/null +++ b/src/librarian/builders/mobi.py @@ -0,0 +1,45 @@ +import os +import six +import subprocess +from tempfile import NamedTemporaryFile +from librarian import functions, get_resource, OutputFile +from librarian.hyphenator import Hyphenator +from .epub import EpubBuilder + + +class MobiBuilder(EpubBuilder): + file_extension = 'mobi' + isbn_field = 'isbn_mobi' + + def build(self, document, use_kindlegen=False, converter_path=None, **kwargs): + bibl_lng = document.meta.language + short_lng = functions.lang_code_3to2(bibl_lng) + try: + self.hyphenator = Hyphenator(get_resource('res/hyph-dictionaries/hyph_' + + short_lng + '.dic')) + except: + pass + + epub = super().build(document, **kwargs) + + devnull = open("/dev/null", 'w') + gen_kwargs = {"stdout": devnull, "stderr": devnull} + + output_file = NamedTemporaryFile(prefix='librarian', suffix='.mobi', + delete=False) + output_file.close() + + if use_kindlegen: + output_file_basename = os.path.basename(output_file.name) + subprocess.check_call([converter_path or 'kindlegen', + '-c2', epub.get_filename(), + '-o', output_file_basename], **gen_kwargs) + else: + subprocess.check_call([converter_path or 'ebook-convert', + epub.get_filename(), + output_file.name, '--no-inline-toc', + '--mobi-file-type=both', + '--mobi-ignore-margins', + ], **gen_kwargs) + return OutputFile.from_filename(output_file.name) + diff --git a/src/librarian/elements/__init__.py b/src/librarian/elements/__init__.py index 6ec0ba7..c934299 100644 --- a/src/librarian/elements/__init__.py +++ b/src/librarian/elements/__init__.py @@ -115,7 +115,6 @@ WL_ELEMENTS = { "animacja": figures.Animacja, "ilustr": figures.Ilustr, - "ref": etree.ElementBase, # Section "wywiad_pyt": blocks.WywiadPyt, "wywiad_odp": blocks.WywiadOdp, diff --git a/src/librarian/elements/base.py b/src/librarian/elements/base.py index 3e0f898..46ae29f 100644 --- a/src/librarian/elements/base.py +++ b/src/librarian/elements/base.py @@ -3,7 +3,6 @@ import re from lxml import etree from librarian import dcparser, RDFNS -from librarian.html import raw_printable_text from librarian.util import get_translation @@ -76,25 +75,35 @@ class WLElement(etree.ElementBase): if parent is not None: parent.signal(signal) - def raw_printable_text(self): + def raw_printable_text(self, builder): + from librarian.html import raw_printable_text + # TODO: podtagi, wyroznienia, etc t = '' - t += self.normalize_text(self.text) + t += self.normalize_text(self.text, builder) for c in self: if not isinstance(c, WLElement): continue if c.tag not in ('pe', 'pa', 'pt', 'pr', 'motyw'): - t += c.raw_printable_text() - t += self.normalize_text(c.tail) + t += c.raw_printable_text(builder) + t += self.normalize_text(c.tail, builder) return t - def normalize_text(self, text): + def normalize_text(self, text, builder): text = text or '' for e, s in self.text_substitutions: text = text.replace(e, s) # FIXME: TEmporary turnoff # text = re.sub(r'\s+', ' ', text) ### TODO: Added now for epub + + if getattr(builder, 'hyphenator', None) is not None: + newt = '' + wlist = re.compile(r'\w+|[^\w]', re.UNICODE).findall(text) + for w in wlist: + newt += builder.hyphenator.inserted(w, u'\u00AD') + text = newt + text = re.sub(r'(?<=\s\w)\s+', u'\u00A0', text) return text @@ -102,7 +111,7 @@ class WLElement(etree.ElementBase): def _build_inner(self, builder, build_method): child_count = len(self) if self.CAN_HAVE_TEXT and self.text: - text = self.normalize_text(self.text) + text = self.normalize_text(self.text, builder) if self.STRIP: text = text.lstrip() if not child_count: @@ -112,7 +121,7 @@ class WLElement(etree.ElementBase): if isinstance(child, WLElement): getattr(child, build_method)(builder) if self.CAN_HAVE_TEXT and child.tail: - text = self.normalize_text(child.tail) + text = self.normalize_text(child.tail, builder) if self.STRIP and i == child_count - 1: text = text.rstrip() builder.push_text(text) @@ -187,7 +196,7 @@ class WLElement(etree.ElementBase): builder.add_toc_entry( fragment, - self.raw_printable_text(), + self.raw_printable_text(builder), self.SECTION_PRECEDENCE ) diff --git a/src/librarian/elements/styles/tytul_dziela.py b/src/librarian/elements/styles/tytul_dziela.py index 906d98c..77852d5 100644 --- a/src/librarian/elements/styles/tytul_dziela.py +++ b/src/librarian/elements/styles/tytul_dziela.py @@ -6,8 +6,8 @@ class TytulDziela(WLElement): EPUB_TAG = HTML_TAG = 'em' EPUB_CLASS = HTML_CLASS = 'book-title' - def normalize_text(self, text): - txt = super(TytulDziela, self).normalize_text(text) + def normalize_text(self, text, builder): + txt = super(TytulDziela, self).normalize_text(text, builder) if self.attrib.get('typ') == '1': txt = '„{txt}”'.format(txt=txt) return txt diff --git a/src/librarian/hyphenator.py b/src/librarian/hyphenator.py index aa5b4c3..fdd50d4 100644 --- a/src/librarian/hyphenator.py +++ b/src/librarian/hyphenator.py @@ -81,13 +81,13 @@ class Hyph_dict(object): """ def __init__(self, filename): self.patterns = {} - f = open(filename) + f = open(filename, 'rb') charset = f.readline().strip() - if charset.startswith('charset '): + if charset.startswith(b'charset '): charset = charset[8:].strip() for pat in f: - pat = pat.decode(charset).strip() + pat = pat.decode(charset.decode('latin1')).strip() if not pat or pat[0] == '%': continue # replace ^^hh with the real character pat = parse_hex(hexrepl, pat) @@ -211,7 +211,7 @@ class Hyphenator(object): the string 'let-ter-gre-pen'. The hyphen string to use can be given as the second parameter, that defaults to '-'. """ - if isinstance(word, str): + if isinstance(word, bytes): word = word.decode('latin1') l = list(word) for p in reversed(self.positions(word)): -- 2.20.1