"upgrade" from PyPDF3 to pypdf

I picked the wrong fork (pypdf3 instead of pypdf2).
PyPDF2 was a fork from the original pyPdf.
After several years, the fork was merged back into pypdf (now all lowercase).
pypdf3 is now unmaintained.

pypdf meanwhile has had a lot of interesting updates, which I should
look at.

.gitignore

@@ -22,6 +22,7 @@ __pycache__/*
 .idea
 .vscode
 tags
+src/pystitcher/_version.py
 
 # Package files
 *.egg

README.md

@@ -3,7 +3,7 @@
 pystitcher stitches your PDF files together, generating nice
 customizable bookmarks for you using a declarative input in the form of
 a markdown file. It is written in pure python and uses
-[PyPDF3](https://pypi.org/project/PyPDF3/) for reading and writing PDF
+[pypdf](https://pypi.org/project/pypdf/) for reading and writing PDF
 files.
 
 ## Installation

pyproject.toml

@@ -20,7 +20,7 @@ dependencies = [
     "html5lib>=1.1",
     "importlib-metadata; python_version<\"3.8\"",
     "Markdown>=3.6",
-    "PyPDF3>=1.0.6",
+    "pypdf>=4.3.1",
     "validators>=0.33.0",
 ]
 

src/pystitcher/stitcher.py

@@ -8,8 +8,8 @@
 import html5lib
 import markdown
 
-from PyPDF3 import PdfFileWriter, PdfFileReader
-from PyPDF3.generic import FloatObject
+from pypdf import PdfWriter, PdfReader
+from pypdf.generic import Fit
 from pystitcher import __version__
 from .bookmark import Bookmark
 
@@ -70,8 +70,8 @@ def _cacheURL(self, url):
     def _get_pdf_number_of_pages(self, filename):
         assert os.path.isfile(filename) and os.access(filename, os.R_OK), \
                 "File {} doesn't exist or isn't readable".format(filename)
-        pdf_reader = PdfFileReader(open(filename, "rb"))
-        return pdf_reader.numPages
+        pdf_reader = PdfReader(open(filename, "rb"))
+        return pdf_reader.get_num_pages()
 
     """
     Return an attribute with a default value of None
@@ -186,7 +186,7 @@ def _iterate_old_bookmarks(self, pdf, startPage, bookmarks, level = 1):
             for inner_bookmark in bookmarks:
                 self._iterate_old_bookmarks(pdf, startPage, inner_bookmark, level+1)
         else:
-            localPageNumber = pdf.getDestinationPageNumber(bookmarks)
+            localPageNumber = pdf.get_destination_page_number(bookmarks)
             globalPageNumber = startPage + localPageNumber - 1
             b = Bookmark(globalPageNumber, bookmarks.title, level, self.defaultFit)
             self.oldBookmarks.append(b)
@@ -198,9 +198,9 @@ def _iterate_old_bookmarks(self, pdf, startPage, bookmarks, level = 1):
     """
     def _insert_bookmarks(self, old_filename, outputFilename):
         stack = []
-        pdfInput = PdfFileReader(open(old_filename, 'rb'))
-        pdfOutput = PdfFileWriter()
-        pdfOutput.cloneDocumentFromReader(pdfInput)
+        pdfInput = PdfReader(open(old_filename, 'rb'))
+        pdfOutput = PdfWriter()
+        pdfOutput.clone_document_from_reader(pdfInput)
         for b in self.bookmarks:
             existingRef = None
             # Trim the stack till the top is useful (stack.level < b.level)
@@ -209,9 +209,9 @@ def _insert_bookmarks(self, old_filename, outputFilename):
             # If stack has something, use it
             if (len(stack) > 0):
                 existingRef = stack[len(stack) - 1][1]
-            bookmargArgs = [b.title, b.page-1, existingRef, None, False, False, b.fit] + b.cords
-            stack.append((b, pdfOutput.addBookmark(*bookmargArgs)))
-        pdfOutput.addMetadata(self._getMetadata())
+            bookmargArgs = [b.title, b.page-1, existingRef, None, False, False, Fit(b.fit)] + b.cords
+            stack.append((b, pdfOutput.add_outline_item(*bookmargArgs)))
+        pdfOutput.add_metadata(self._getMetadata())
         pdfOutput.write(open(outputFilename, 'wb'))
 
     """
@@ -220,15 +220,15 @@ def _insert_bookmarks(self, old_filename, outputFilename):
     as we're reading them
     """
     def _merge(self, output):
-        writer = PdfFileWriter()
+        writer = PdfWriter()
         for (inputFile,startPage,filters) in self.files:
             assert os.path.isfile(inputFile), ERROR_PATH.format(inputFile)
-            reader = PdfFileReader(open(inputFile, 'rb'))
+            reader = PdfReader(open(inputFile, 'rb'))
             # Recursively iterate through the old bookmarks
-            self._iterate_old_bookmarks(reader, startPage, reader.getOutlines())
+            self._iterate_old_bookmarks(reader, startPage, reader.outline)
             rotate, start, end = filters
             for page in range(start, end + 1):
-                writer.addPage(reader.getPage(page - 1).rotateClockwise(rotate))
+                writer.add_page(reader.get_page(page - 1).rotate(rotate))
 
         writer.write(output)
         output.close()

tests/test_integration.py

@@ -1,7 +1,7 @@
 import os
 import io
 
-import PyPDF3
+import pypdf
 from pystitcher.stitcher import Stitcher
 from pystitcher import __version__
 
@@ -55,16 +55,16 @@ def flatten_bookmarks(bookmarks, level=0):
 
 def get_all_bookmarks(pdf):
     """ Returns a list of all bookmarks with title, page number, and level in a PDF file"""
-    bookmarks = flatten_bookmarks(pdf.getOutlines())
-    return [(d[0]['/Title'], pdf.getDestinationPageNumber(d[0]), d[1]) for d in bookmarks]
+    bookmarks = flatten_bookmarks(pdf.outline)
+    return [(d[0]['/Title'], pdf.get_destination_page_number(d[0]), d[1]) for d in bookmarks]
 
 @pytest.mark.parametrize("name,pages,metadata,bookmarks", TEST_DATA)
 def test_book(name, pages, metadata, bookmarks):
     output_file = render(name)
-    pdf = PyPDF3.PdfFileReader(output_file)
-    assert pages == pdf.getNumPages()
+    pdf = pypdf.PdfReader(output_file)
+    assert pages == pdf.get_num_pages()
     assert bookmarks == get_all_bookmarks(pdf)
-    info = pdf.getDocumentInfo()
+    info = pdf.metadata
     identity = "pystitcher/%s" % __version__
     assert identity == info['/Producer']
     assert identity == info['/Creator']
@@ -74,23 +74,23 @@ def test_book(name, pages, metadata, bookmarks):
 def test_rotation():
     """ Validates the book-rotate.pdf with pages rotated."""
     output_file = render("rotate")
-    pdf = PyPDF3.PdfFileReader(output_file)
+    pdf = pypdf.PdfReader(output_file)
     # Note that inputs to getPage are 0-indexed
-    assert 90 == pdf.getPage(3)['/Rotate']
-    assert 90 == pdf.getPage(4)['/Rotate']
-    assert 90 == pdf.getPage(5)['/Rotate']
-    assert 180 == pdf.getPage(6)['/Rotate']
-    assert 180 == pdf.getPage(7)['/Rotate']
-    assert 180 == pdf.getPage(8)['/Rotate']
+    assert 90 == pdf.get_page(3)['/Rotate']
+    assert 90 == pdf.get_page(4)['/Rotate']
+    assert 90 == pdf.get_page(5)['/Rotate']
+    assert 180 == pdf.get_page(6)['/Rotate']
+    assert 180 == pdf.get_page(7)['/Rotate']
+    assert 180 == pdf.get_page(8)['/Rotate']
 
 def test_cleanup_disabled():
     f = io.StringIO()
     with redirect_stdout(f):
         output_file = render("min", False)
     temp_filename = f.getvalue()[29:-1]
     assert os.path.exists(temp_filename)
-    pdf = PyPDF3.PdfFileReader(temp_filename)
-    assert 3 == pdf.getNumPages()
-    assert [] == pdf.getOutlines()
+    pdf = pypdf.PdfReader(temp_filename)
+    assert 3 == pdf.get_num_pages()
+    assert [] == pdf.outline
     # Clean it up manually to avoid cluttering
     os.remove(temp_filename)