Refactor exit codes; test for missing tessdata

Some versions of tesseract installed by homebrew end up without a
functional tessdata folder, and tesseract is not helpful in this
situation, so add a new test to make sure our output is at least
indicative of the problem.

In the process of properly handling return codes I discovered
test_override_metadata triggers a NPE inside JHOVE probably due to the
Unicode character checking.  This could be specific to my JRE (1.6.0_65,
Oracle) but it's probably JHOVE's fault.  A valid PDF/A (per Acrobat)
is still generated.

ocrmypdf/__init__.py

@@ -0,0 +1,11 @@
+from enum import IntEnum
+
+
+class ExitCode(IntEnum):
+    bad_args = 1
+    input_file = 2
+    missing_dependency = 3
+    invalid_output_pdfa = 4
+    file_access_error = 5
+    already_done_ocr = 6
+    other_error = 15

ocrmypdf/main.py

@@ -33,7 +33,7 @@ from .pageinfo import pdf_get_all_pageinfo
 from .pdfa import generate_pdfa_def
 from . import ghostscript
 from . import tesseract
-
+from . import ExitCode
 
 warnings.simplefilter('ignore', pypdf.utils.PdfReadWarning)
 
@@ -43,13 +43,6 @@ JHOVE_PATH = os.path.realpath(os.path.join(BASEDIR, 'jhove'))
 JHOVE_JAR = os.path.join(JHOVE_PATH, 'bin', 'JhoveApp.jar')
 JHOVE_CFG = os.path.join(JHOVE_PATH, 'conf', 'jhove.conf')
 
-EXIT_BAD_ARGS = 1
-EXIT_BAD_INPUT_FILE = 2
-EXIT_MISSING_DEPENDENCY = 3
-EXIT_INVALID_OUTPUT_PDFA = 4
-EXIT_FILE_ACCESS_ERROR = 5
-EXIT_ALREADY_DONE_OCR = 6
-EXIT_OTHER_ERROR = 15
 
 # -------------
 # External dependencies
@@ -66,7 +59,7 @@ if tesseract.version() < MINIMUM_TESS_VERSION:
         "Please install tesseract {0} or newer "
         "(currently installed version is {1})".format(
             MINIMUM_TESS_VERSION, tesseract.version()))
-    sys.exit(EXIT_MISSING_DEPENDENCY)
+    sys.exit(ExitCode.missing_dependency)
 
 
 # -------------
@@ -184,7 +177,7 @@ if not set(options.language).issubset(tesseract.languages()):
         "data for the following requested languages: ")
     for lang in (set(options.language) - tesseract.languages()):
         complain(lang, file=sys.stderr)
-    sys.exit(EXIT_BAD_ARGS)
+    sys.exit(ExitCode.bad_args)
 
 
 # ----------
@@ -197,7 +190,7 @@ if any((options.deskew, options.clean, options.clean_final)):
     except ImportError:
         complain(
             "Install the 'unpaper' program to use --deskew or --clean.")
-        sys.exit(EXIT_BAD_ARGS)
+        sys.exit(ExitCode.bad_args)
 else:
     unpaper = None
 
@@ -209,7 +202,7 @@ if options.debug_rendering and options.pdf_renderer == 'tesseract':
 if options.force_ocr and options.skip_text:
     complain(
         "Error: --force-ocr and --skip-text are mutually incompatible.")
-    sys.exit(EXIT_BAD_ARGS)
+    sys.exit(ExitCode.bad_args)
 
 if options.clean and not options.clean_final \
         and options.pdf_renderer == 'tesseract':
@@ -357,7 +350,7 @@ def is_ocr_required(pageinfo, log):
         if not options.force_ocr and not options.skip_text:
             log.error(s.format(page,
                                "aborting (use --force-ocr to force OCR)"))
-            sys.exit(EXIT_ALREADY_DONE_OCR)
+            sys.exit(ExitCode.already_done_ocr)
         elif options.force_ocr:
             log.info(s.format(page,
                               "rasterizing text and running OCR anyway"))
@@ -850,14 +843,14 @@ def run_pipeline():
 
     pdf_is_valid, pdf_is_pdfa = validate_pdfa(options.output_file, _log)
 
-    returncode = EXIT_OTHER_ERROR  # Assume error
+    returncode = ExitCode.other_error  # Assume error
 
     if not pdf_is_valid:
         _log.warning('Output file: The generated PDF/A file is INVALID')
-        returncode = EXIT_INVALID_OUTPUT_PDFA
+        returncode = ExitCode.invalid_output_pdfa
     elif pdf_is_valid and not pdf_is_pdfa:
         _log.warning('Output file: Generated file is VALID PDF but not PDF/A')
-        returncode = EXIT_INVALID_OUTPUT_PDFA
+        returncode = ExitCode.invalid_output_pdfa
     elif pdf_is_valid and pdf_is_pdfa:
         _log.info('Output file: The generated PDF/A file is VALID')
         returncode = 0

ocrmypdf/tesseract.py

@@ -6,6 +6,7 @@ import sys
 import os
 import re
 from functools import lru_cache
+from . import ExitCode
 
 
 @lru_cache(maxsize=1)
@@ -20,7 +21,7 @@ def version():
                 stderr=STDOUT)
     except CalledProcessError:
         print("Could not find Tesseract executable on system PATH.")
-        sys.exit(1)
+        sys.exit(ExitCode.missing_dependency)
 
     tesseract_version = re.match(r'tesseract\s(.+)', versions).group(1)
     return tesseract_version
@@ -32,9 +33,16 @@ def languages():
         'tesseract',
         '--list-langs'
     ]
-    langs = check_output(
-            args_tess, close_fds=True, universal_newlines=True,
-            stderr=STDOUT)
+    try:
+        langs = check_output(
+                args_tess, close_fds=True, universal_newlines=True,
+                stderr=STDOUT)
+    except CalledProcessError as e:
+        print("Tesseract failed to report available languages.")
+        print("Output from Tesseract:")
+        print("-" * 40)
+        print(e.output)
+        sys.exit(ExitCode.missing_dependency)
     return set(lang.strip() for lang in langs.splitlines()[1:])
 
 

tests/test_main.py

@@ -11,6 +11,7 @@ from unittest.mock import patch, create_autospec
 import pytest
 from ocrmypdf.pageinfo import pdf_get_all_pageinfo
 import PyPDF2 as pypdf
+from ocrmypdf import ExitCode
 
 
 if sys.version_info.major < 3:
@@ -243,4 +244,14 @@ def test_ghostscript_pdfa_fails(break_ghostscript_pdfa):
 
     p, out, err = run_ocrmypdf_env(
         'graph_ocred.pdf', 'not_a_pdfa.pdf', env, '-v', '1', '--skip-text')
-    assert p.returncode == 4, err  # not PDFA
+    assert p.returncode == ExitCode.invalid_output_pdfa, err  # not PDFA
+
+
+def test_tesseract_missing_tessdata():
+    env = os.environ
+    env['TESSDATA_PREFIX'] = '/tmp'
+
+    p, _, err = run_ocrmypdf_env(
+        'graph_ocred.pdf', 'not_a_pdfa.pdf', env, '-v', '1', '--skip-text')
+    assert p.returncode == ExitCode.missing_dependency, err
+