From c75bc0568e357f8e852449f51a879c1a2274ccf8 Mon Sep 17 00:00:00 2001
From: "W. Trevor King" <wking@drexel.edu>
Date: Thu, 9 Feb 2012 10:46:01 -0500
Subject: [PATCH] Add `--pdfmarks` option to pdf-merge.py.

This makes it easy get a pdfmark file from one PDF and and apply it to
another (among other things).  The workflow suggested by Larry was:

On Thu, Feb 09, 2012 at 03:25:09PM +0800, Larry Cai wrote:
> 1. pdftk source.pdf  dump_data_utf8 > in.info # save rawdata
> 2. pdftk A=book-cover.pdf B=source.pdf cat A1-2 B3-end output replaced.pdf
> # replace two pages for the book
> 3. pdf-convert < in.info > pdfmarks
> 4. gs -dBATCH -dNOPAUSE -sDEVICE=pdfwrite -sOutputFile=output.pdf
> replaced.pdf  pdfmarks # put bookmark back

With the new option, you can accomplish the same goal with

  $ rm -f pdfmarks
  $ pdf-convert.py --pdfmarks pdfmarks source.pdf
  $ pdftk A=book-cover.pdf B=source.pdf cat A1-2 B3-end output replaced.pdf
  $ pdf-convert.py --pdfmarks pdfmarks --output output.pdf replaced.pdf

With the first call to pdf-convert.py creating pdfmarks (using
`pdftk`), and the second call applying it (using `gs`).
---
 .../pdf-merge.py                              | 22 ++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/posts/PDF_bookmarks_with_Ghostscript/pdf-merge.py b/posts/PDF_bookmarks_with_Ghostscript/pdf-merge.py
index cb6bec0..6266fe4 100755
--- a/posts/PDF_bookmarks_with_Ghostscript/pdf-merge.py
+++ b/posts/PDF_bookmarks_with_Ghostscript/pdf-merge.py
@@ -18,12 +18,14 @@
 
 """Merge PDFs perserving bookmarks.
 
-Thanks to Larry Cai for suggesting that Unicode be supported.
+Thanks to Larry Cai for suggesting that Unicode be supported and for
+discussion about the `--pdfmarks` option.
 """
 
 import codecs as _codecs
 import locale as _locale
 import os as _os
+import os.path as _os_path
 import re as _re
 import subprocess as _subprocess
 import sys as _sys
@@ -373,6 +375,14 @@ if __name__ == '__main__':
                         help='path to the pdftk executable')
     parser.add_argument('--gs', dest='gs', default=GS,
                         help='path to the gs (Ghostscript) executable')
+    parser.add_argument('--pdfmarks', dest='pdfmarks',
+                        help=('path to pdfmarks file.  If not given, a '
+                              'temporary file is used.  If given and the file '
+                              'is missing, execution will stop after the file '
+                              'is created (before the Ghostscript run).  If '
+                              'given and the file exists, no attempt will be '
+                              'make to use pdftk to generate the mark file (I '
+                              'assume your input file is what you want).'))
     parser.add_argument('--argv-encoding', dest='argv_encoding',
                         help=('Optionally override the locale encoding for '
                               'your command line arguments.'))
@@ -418,7 +428,13 @@ if __name__ == '__main__':
         keywords = [unicode(k, argv_encoding) for k in args.keywords]
     else:
         keywords = None
-    pdfmarks = generate_pdfmarks(
-        inputs, title=title, author=author, keywords=keywords)
+    if args.pdfmarks and _os_path.isfile(args.pdfmarks):
+        pdfmarks = open(args.pdfmarks, 'r').read()
+    else:
+        pdfmarks = generate_pdfmarks(
+            inputs, title=title, author=author, keywords=keywords)
+        if args.pdfmarks:
+            open(args.pdfmarks, 'w').write(pdfmarks)
+            _sys.exit(0)
     merge_pdfs(inputs=inputs, pdfmarks=pdfmarks, output=args.output,
                pause_for_manual_tweaking=args.pause_for_manual_tweaking)
-- 
2.26.2