v9.0.3 notes; Remove test_tesseract_config_notfound from suite

Previously we included the
   filename, which required Postscript to run with file access enabled. For
   security, Ghostscript 9.28 enables ``-dSAFER`` and as such, no longer
   permits access to any file by default. This fix is necessary for
   compatibility with Ghostscript 9.28.

We use ASCII85 for a slightly more compact representation.

docs/release_notes.rst

@@ -13,6 +13,18 @@ Note that it is licensed under GPLv3, so scripts that
 ``import ocrmypdf`` and are released publicly should probably also be
 licensed under GPLv3.
 
+v9.0.3
+======
+
+-  Embed an encoded version of the sRGB ICC profile in the intermediate
+   Postscript file (used for PDF/A conversion). Previously we included the
+   filename, which required Postscript to run with file access enabled. For
+   security, Ghostscript 9.28 enables ``-dSAFER`` and as such, no longer
+   permits access to any file by default. This fix is necessary for
+   compatibility with Ghostscript 9.28.
+-  Exclude a test that sometimes times out and fails in continuous integration
+   from the standard test suite.
+
 v9.0.2
 ======
 
@@ -23,7 +35,7 @@ v9.0.2
 -  Fixed an issue that caused inversion of black and white in monochrome images.
    We are not certain but the problem seems to be linked to Leptonica 1.76.0 and
    older.
--  Fixed some cases where the test suite failed or produced unexpected if
+-  Fixed some cases where the test suite failed if
    English or German Tesseract language packs were not installed.
 -  Fixed a runtime error if the Tesseract English language is not installed.
 -  Improved explicit closing of Pillow images after use.

src/ocrmypdf/exec/ghostscript.py

@@ -259,6 +259,7 @@ def generate_pdfa(
                 "-dQUIET",
                 "-dBATCH",
                 "-dNOPAUSE",
+                "-dSAFER",
                 "-dCompatibilityLevel=" + str(pdf_version),
                 "-sDEVICE=pdfwrite",
                 "-dAutoRotatePages=/None",

src/ocrmypdf/pdfa.py

@@ -31,6 +31,7 @@ Ghostscript's handling of pdfmark.
 
 """
 
+import base64
 import os
 from binascii import hexlify
 from pathlib import Path
@@ -48,12 +49,10 @@ SRGB_ICC_PROFILE = pkg_resources.resource_filename('ocrmypdf', ICC_PROFILE_RELPA
 # files, from the Ghostscript documentation. Lines beginning with % are
 # comments. Python substitution variables have a '$' prefix.
 pdfa_def_template = u"""%!
-% Define entries in the document Info dictionary :
+% Define an ICC profile :
 /ICCProfile $icc_profile
 def
 
-% Define an ICC profile :
-
 [/_objdef {icc_PDFA} /type /stream /OBJ pdfmark
 [{icc_PDFA}
 <<
@@ -67,7 +66,7 @@ def
     (ERROR, unable to determine ProcessColorModel) == flush
   } ifelse
 >> /PUT pdfmark
-[{icc_PDFA} ICCProfile (r) file /PUT pdfmark
+[{icc_PDFA} ICCProfile /PUT pdfmark
 
 % Define the output intent dictionary :
 
@@ -104,16 +103,10 @@ def generate_pdfa_ps(target_filename, icc='sRGB'):
     else:
         raise NotImplementedError("Only supporting sRGB")
 
-    # pdfmark must contain the full path to the ICC profile, and pdfmark must be
+    # Read the ICC profile, encode as ASCII85 and convert to a string which we
-    # also encoded in ASCII. ocrmypdf can be installed anywhere, including to
+    # will insert in the .ps file
-    # paths that have a non-ASCII character in the filename. Ghostscript
+    bytes_icc_profile = Path(icc_profile).read_bytes()
-    # accepts hex-encoded strings and converts them to byte strings, so
+    icc_profile = base64.a85encode(bytes_icc_profile, adobe=True).decode('ascii')
-    # we encode the path with fsencode() and use the hex representation.
-    # UTF-16 not accepted here. (Even though ASCII encodable is the usual case,
-    # do this always to avoid making it a rare conditional.)
-    bytes_icc_profile = os.fsencode(icc_profile)
-    hex_icc_profile = hexlify(bytes_icc_profile)
-    icc_profile = '<' + hex_icc_profile.decode('ascii') + '>'
 
     t = Template(pdfa_def_template)
     ps = t.substitute(icc_profile=icc_profile, icc_identifier=icc)

tests/test_main.py

@@ -613,6 +613,7 @@ language_model_penalty_non_freq_dict_word 0
     )
 
 
+@pytest.mark.slow  # This test sometimes times out in CI
 @pytest.mark.parametrize('renderer', RENDERERS)
 def test_tesseract_config_notfound(renderer, resources, outdir):
     cfg_file = outdir / 'nofile.cfg'