Implement setting metadata in textual PDFs

src/calibre/ebooks/conversion/plugins/pdf_output.py

@@ -23,29 +23,6 @@ PAPER_SIZES = ('a0', 'a1', 'a2', 'a3', 'a4', 'a5', 'a6', 'b0', 'b1',
         'b2', 'b3', 'b4', 'b5', 'b6', 'legal', 'letter')
 
 
-class PDFMetadata(object):  # {{{
-
-    def __init__(self, mi=None):
-        from calibre import force_unicode
-        from calibre.ebooks.metadata import authors_to_string
-        self.title = _('Unknown')
-        self.author = _('Unknown')
-        self.tags = ''
-        self.mi = mi
-
-        if mi is not None:
-            if mi.title:
-                self.title = mi.title
-            if mi.authors:
-                self.author = authors_to_string(mi.authors)
-            if mi.tags:
-                self.tags = ', '.join(mi.tags)
-
-        self.title = force_unicode(self.title)
-        self.author = force_unicode(self.author)
-# }}}
-
-
 class PDFOutput(OutputFormatPlugin):
 
     name = 'PDF Output'
@@ -201,7 +178,7 @@ class PDFOutput(OutputFormatPlugin):
 
     def convert_images(self, images):
         from calibre.ebooks.pdf.image_writer import convert
-        convert(images, self.output_path, self.opts, PDFMetadata(self.metadata))
+        convert(images, self.output_path, self.opts, self.metadata)
 
     def get_cover_data(self):
         oeb = self.oeb
@@ -261,4 +238,4 @@ class PDFOutput(OutputFormatPlugin):
             oeb_output = plugin_for_output_format('oeb')
             oeb_output.convert(oeb_book, oeb_dir, self.input_plugin, self.opts, self.log)
             opfpath = glob.glob(os.path.join(oeb_dir, '*.opf'))[0]
-            convert(opfpath, self.opts, self.output_path, self.log)
+            convert(opfpath, self.opts, metadata=self.metadata, output_path=self.output_path, log=self.log)

src/calibre/ebooks/pdf/html_writer.py

@@ -6,17 +6,19 @@ from __future__ import absolute_import, division, print_function, unicode_litera
 
 import os
 
-from PyQt5.Qt import QApplication, QUrl, QTimer
+from PyQt5.Qt import QApplication, QTimer, QUrl
 from PyQt5.QtWebEngineWidgets import QWebEnginePage
 
 from calibre.constants import iswindows
 from calibre.ebooks.oeb.polish.container import Container as ContainerBase
 from calibre.ebooks.oeb.polish.split import merge_html
+from calibre.ebooks.pdf.image_writer import (
+    PDFMetadata, get_page_layout, update_metadata
+)
 from calibre.gui2 import setup_unix_signals
 from calibre.gui2.webengine import secure_webengine
-from calibre.ebooks.pdf.image_writer import get_page_layout
 from calibre.utils.logging import default_log
-
+from calibre.utils.podofo import get_podofo
 from polyglot.builtins import range
 
 OK, LOAD_FAILED, KILL_SIGNAL = range(0, 3)
@@ -96,7 +98,7 @@ class Renderer(QWebEnginePage):
         return self.pdf_data
 
 
-def convert(opf_path, opts, output_path=None, log=default_log):
+def convert(opf_path, opts, metadata=None, output_path=None, log=default_log):
     container = Container(opf_path, log)
     spine_names = [name for name, is_linear in container.spine_names]
     master = spine_names[0]
@@ -109,7 +111,14 @@ def convert(opf_path, opts, output_path=None, log=default_log):
     renderer = Renderer(opts)
     page_layout = get_page_layout(opts)
     pdf_data = renderer.convert_html_file(index_file, page_layout, settle_time=1)
+    podofo = get_podofo()
+    pdf_doc = podofo.PDFDoc()
+    pdf_doc.load(pdf_data)
 
+    if metadata is not None:
+        update_metadata(pdf_doc, PDFMetadata(metadata))
+
+    pdf_data = pdf_doc.write()
     if output_path is None:
         return pdf_data
     with open(output_path, 'wb') as f:

src/calibre/ebooks/pdf/image_writer.py

@@ -12,6 +12,30 @@ from calibre.ebooks.metadata.xmp import metadata_to_xmp_packet
 from calibre.utils.img import image_from_path
 from calibre.utils.podofo import get_podofo, set_metadata_implementation
 
+
+class PDFMetadata(object):  # {{{
+
+    def __init__(self, mi=None):
+        from calibre import force_unicode
+        from calibre.ebooks.metadata import authors_to_string
+        self.title = _('Unknown')
+        self.author = _('Unknown')
+        self.tags = ''
+        self.mi = mi
+
+        if mi is not None:
+            if mi.title:
+                self.title = mi.title
+            if mi.authors:
+                self.author = authors_to_string(mi.authors)
+            if mi.tags:
+                self.tags = ', '.join(mi.tags)
+
+        self.title = force_unicode(self.title)
+        self.author = force_unicode(self.author)
+# }}}
+
+
 # Page layout {{{
 
 
@@ -82,8 +106,17 @@ def draw_image_page(painter, img, preserve_aspect_ratio=True):
     painter.drawImage(page_rect, img)
 
 
-def convert(images, output_path, opts, pdf_metadata):
+def update_metadata(pdf_doc, pdf_metadata):
+    if pdf_metadata.mi:
+        xmp_packet = metadata_to_xmp_packet(pdf_metadata.mi)
+        set_metadata_implementation(
+            pdf_doc, pdf_metadata.title, pdf_metadata.mi.authors,
+            pdf_metadata.mi.book_producer, pdf_metadata.mi.tags, xmp_packet)
+
+
+def convert(images, output_path, opts, metadata):
     writer = QPdfWriter(output_path)
+    pdf_metadata = PDFMetadata(metadata)
     writer.setCreator(pdf_metadata.author)
     writer.setTitle(pdf_metadata.title)
     writer.setPageLayout(get_page_layout(opts, for_comic=True))
@@ -97,16 +130,13 @@ def convert(images, output_path, opts, pdf_metadata):
             draw_image_page(painter, img)
     finally:
         painter.end()
-    if pdf_metadata.mi:
+
-        podofo = get_podofo()
+    podofo = get_podofo()
-        pdf_doc = podofo.PDFDoc()
+    pdf_doc = podofo.PDFDoc()
-        with open(output_path, 'r+b') as f:
+    with open(output_path, 'r+b') as f:
-            raw = f.read()
+        raw = f.read()
-            pdf_doc.load(raw)
+        pdf_doc.load(raw)
-            xmp_packet = metadata_to_xmp_packet(pdf_metadata.mi)
+        update_metadata(pdf_doc, pdf_metadata)
-            set_metadata_implementation(
+        raw = pdf_doc.write()
-                pdf_doc, pdf_metadata.title, pdf_metadata.mi.authors,
+        f.seek(0), f.truncate()
-                pdf_metadata.mi.book_producer, pdf_metadata.tags, xmp_packet)
+        f.write(raw)
-            raw = pdf_doc.write()
-            f.seek(0), f.truncate()
-            f.write(raw)