Refactor obtaining version from subprocesses

Issue #196 raised the need to deal with linker warnings on --version.

ocrmypdf/exec/__init__.py

@@ -4,9 +4,40 @@
 """Wrappers to manage subprocess calls"""
 
 import os
+import re
+import sys
+from subprocess import run, STDOUT, PIPE, CalledProcessError
+from ..exceptions import MissingDependencyError
 
 
 def get_program(name):
     "Check environment variables for overrides to this program"
     envvar = 'OCRMYPDF_' + name.upper()
     return os.environ.get(envvar, name)
+
+
+def get_version(program, *, 
+        version_arg='--version', regex=r'(\d+(\.\d+)*)'):
+    "Get the version of the specified program, "
+    args_prog = [
+        get_program(program),
+        version_arg
+    ]
+    try:
+        proc = run(
+            args_prog, close_fds=True, universal_newlines=True,
+            stdout=PIPE, stderr=STDOUT, check=True)
+        output = proc.stdout
+    except CalledProcessError as e:
+        raise MissingDependencyError(
+            "Could not find program '{}' on the PATH".format(program)) from e
+
+    try:
+        version = re.match(regex, output.strip()).group(1)
+    except AttributeError as e:
+        raise MissingDependencyError(
+            ("The program '{}' did not report its version. "
+            "Message was:\n{}").format(program, output)
+        )
+
+    return version

ocrmypdf/exec/ghostscript.py

@@ -8,28 +8,14 @@ from functools import lru_cache
 import re
 import sys
 from PIL import Image
-from . import get_program
+from . import get_program, get_version
 from ..exceptions import SubprocessOutputError, MissingDependencyError
 from ..helpers import fspath
 
 
 @lru_cache(maxsize=1)
 def version():
-    args_gs = [
-        get_program('gs'),
-        '--version'
-    ]
-    try:
-        proc = run(
-                args_gs, close_fds=True, universal_newlines=True,
-                stdout=PIPE, stderr=STDOUT, check=True)
-        ver = proc.stdout
-    except CalledProcessError as e:
-        print("Could not find Ghostscript executable on system PATH.",
-              file=sys.stderr)
-        raise MissingDependencyError from e
-
-    return ver.strip()
+    return get_version('gs')
 
 
 def _gs_error_reported(stream):

ocrmypdf/exec/qpdf.py

@@ -9,25 +9,12 @@ import re
 
 from ..exceptions import InputFileError, SubprocessOutputError, \
     MissingDependencyError, EncryptedPdfError
-from . import get_program
+from . import get_program, get_version
 
 
 @lru_cache(maxsize=1)
 def version():
-    args_qpdf = [
-        get_program('qpdf'),
-        '--version'
-    ]
-    try:
-        p = run(args_qpdf, universal_newlines=True, stderr=STDOUT,
-                stdout=PIPE)
-    except CalledProcessError as e:
-        print("Could not find qpdf executable on system PATH.",
-              file=sys.stderr)
-        raise MissingDependencyError() from e
-
-    qpdf_version = re.match(r'qpdf version (.+)', p.stdout).group(1)
-    return qpdf_version
+    return get_version('qpdf', regex=r'qpdf version (.+)')    
 
 
 def check(input_file, log=None):

ocrmypdf/exec/tesseract.py

@@ -14,7 +14,7 @@ from subprocess import PIPE, CalledProcessError, \
 
 from ..exceptions import MissingDependencyError, TesseractConfigError
 from ..helpers import page_number
-from . import get_program
+from . import get_program, get_version
 
 OrientationConfidence = namedtuple(
     'OrientationConfidence',
@@ -40,21 +40,7 @@ HOCR_TEMPLATE = """<?xml version="1.0" encoding="UTF-8"?>
 
 @lru_cache(maxsize=1)
 def version():
-    args_tess = [
-        get_program('tesseract'),
-        '--version'
-    ]
-    try:
-        versions = check_output(
-                args_tess, close_fds=True, universal_newlines=True,
-                stderr=STDOUT)
-    except CalledProcessError as e:
-        print("Could not find Tesseract executable on system PATH.",
-              file=sys.stderr)
-        raise MissingDependencyError from e
-
-    tesseract_version = re.match(r'tesseract\s(.+)', versions).group(1)
-    return tesseract_version
+    return get_version('tesseract', regex=r'tesseract\s(.+)')
 
 
 def v4():

ocrmypdf/exec/unpaper.py

@@ -9,19 +9,7 @@ import sys
 import os
 from functools import lru_cache
 from ..exceptions import MissingDependencyError
-from . import get_program
-
-
-@lru_cache(maxsize=1)
-def version():
-    args_unpaper = [
-        get_program('unpaper'),
-        '--version'
-    ]
-    ver = check_output(
-        args_unpaper, close_fds=True, universal_newlines=True,
-        stderr=STDOUT, timeout=5)
-    return ver.strip()
+from . import get_program, get_version
 
 
 try:
@@ -31,6 +19,11 @@ except ImportError:
     raise
 
 
+@lru_cache(maxsize=1)
+def version():
+    return get_version('unpaper')
+
+
 def run(input_file, output_file, dpi, log, mode_args):
     args_unpaper = [
         get_program('unpaper'),