Add tesseract version check

src/ocrmypdf.py

@@ -28,6 +28,7 @@ import ruffus.cmdline as cmdline
 from .hocrtransform import HocrTransform
 from .pageinfo import pdf_get_all_pageinfo
 from .pdfa import generate_pdfa_def
+from .tesseract import TESS_VERSION
 
 
 warnings.simplefilter('ignore', pypdf.utils.PdfReadWarning)
@@ -38,6 +39,30 @@ 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_OTHER_ERROR=15
+
+# -------------
+# External dependencies
+
+MINIMUM_TESS_VERSION = '3.02.02'
+
+if TESS_VERSION < MINIMUM_TESS_VERSION:
+    print(
+        "Please install tesseract {0} or newer "
+        "(currently installed version is {1})".format(
+            MINIMUM_TESS_VERSION, TESS_VERSION),
+        file=sys.stderr)
+    sys.exit(EXIT_MISSING_DEPENDENCY)
+
+
+# -------------
+# Parser
+
 parser = cmdline.get_argparse(
     prog="OCRmyPDF",
     description="Generate searchable PDF file from an image-only PDF file.")

src/tesseract.py

@@ -0,0 +1,21 @@
+#!/usr/bin/env python3
+
+from cffi import FFI
+from ctypes.util import find_library
+import sys
+
+ffi = FFI()
+try:
+    libtess = ffi.dlopen(find_library('libtesseract'))
+except Exception:
+    print("Could not find Tesseract 3.02.02", file=sys.stderr)
+    sys.exit(1)
+
+ffi.cdef('''
+    const char* TessVersion();
+''')
+
+cstr_version = libtess.TessVersion()
+TESS_VERSION = ffi.string(cstr_version).decode('ascii')
+
+__all__ = ['TESS_VERSION']