Support for writing metadata to PDF files (thanks to John Schember)

src/calibre/customize/builtins.py

@@ -251,6 +251,17 @@ class MOBIMetadataWriter(MetadataWriterPlugin):
     def set_metadata(self, stream, mi, type):
         from calibre.ebooks.metadata.mobi import set_metadata
         set_metadata(stream, mi)
+        
+class PDFMetadataWriter(MetadataWriterPlugin):
+
+    name        = 'Set PDF metadata'
+    file_types  = set(['pdf'])
+    description = _('Set metadata in %s files') % 'PDF'
+    author      = 'John Schember'
+    
+    def set_metadata(self, stream, mi, type):
+        from calibre.ebooks.metadata.pdf import set_metadata
+        set_metadata(stream, mi)
 
 
 plugins = [HTML2ZIP]

src/calibre/ebooks/metadata/pdf.py

@@ -2,10 +2,10 @@ __license__   = 'GPL v3'
 __copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
 '''Read meta information from PDF files'''
 
-import sys, os, re
+import sys, os, StringIO
 
 from calibre.ebooks.metadata import MetaInformation, authors_to_string, get_parser
-from pyPdf import PdfFileReader
+from pyPdf import PdfFileReader, PdfFileWriter
 
 def get_metadata(stream):
     """ Return metadata as a L{MetaInfo} object """
@@ -31,18 +31,27 @@ def get_metadata(stream):
 
 def set_metadata(stream, mi):
     stream.seek(0)
-    raw = stream.read()
-    if mi.title:
-        tit = mi.title.encode('utf-8') if isinstance(mi.title, unicode) else mi.title
-        raw = re.compile(r'<<.*?/Title\((.+?)\)', re.DOTALL).sub(lambda m: m.group().replace(m.group(1), tit), raw)
-    if mi.authors:
-        au = authors_to_string(mi.authors)
-        if isinstance(au, unicode):
-            au = au.encode('utf-8')
-        raw = re.compile(r'<<.*?/Author\((.+?)\)', re.DOTALL).sub(lambda m: m.group().replace(m.group(1), au), raw)
+    
+    # Use a StringIO object for the pdf because we will want to over
+    # write it later and if we are working on the stream directly it
+    # could cause some issues.
+    raw = StringIO.StringIO(stream.read())
+    orig_pdf = PdfFileReader(raw)
+    
+    title = mi.title if mi.title else orig_pdf.documentInfo.title
+    author = authors_to_string(mi.authors) if mi.authors else orig_pdf.documentInfo.author
+    
+    out_pdf = PdfFileWriter(title=title, author=author)
+    for page in orig_pdf.pages:
+        out_pdf.addPage(page)
+        
+    out_str = StringIO.StringIO()
+    out_pdf.write(out_str)
+    
     stream.seek(0)
     stream.truncate()
-    stream.write(raw)
+    out_str.seek(0)
+    stream.write(out_str.read())
     stream.seek(0)
 
 def option_parser():