menekis/OCRmyPDF
Public Access
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Remove tesseract_env, --tesseract-env

Browse Source
This commit is contained in:
James R. Barlow
2020-06-09 00:39:53 -07:00
parent 21c0e045cb
commit 3b6f6782f0
5 changed files with 22 additions and 74 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/ocrmypdf/api.py
+1 -1
View File
@@ -143,7 +143,7 @@ def create_options(
# These arguments with special handling for which we bypass
# argparse
if arg in {'tesseract_env', 'progress_bar', 'plugins'}:
if arg in {'progress_bar', 'plugins'}:
deferred.append((arg, val))
continue
src/ocrmypdf/builtin_plugins/tesseract_ocr.py
+6 -19
View File
@@ -81,7 +81,6 @@ def add_options(parser):
metavar='FILE',
help="Specify the location of the Tesseract user patterns file.",
)
tess.add_argument('--tesseract-env', type=str, help=argparse.SUPPRESS)
@hookimpl
@@ -98,15 +97,13 @@ def check_options(options):
options.pdf_renderer = 'sandwich'
if options.pdf_renderer == 'sandwich' and not tesseract.has_textonly_pdf(
options.tesseract_env, set(options.languages)
set(options.languages)
):
raise MissingDependencyError(
"You are using an alpha version of Tesseract 4.0 that does not support "
"the textonly_pdf parameter. We don't support versions this old."
)
if not tesseract.has_user_words(options.tesseract_env) and (
options.user_words or options.user_patterns
):
if not tesseract.has_user_words() and (options.user_words or options.user_patterns):
log.warning(
"Tesseract 4.0 ignores --user-words and --user-patterns, so these "
"arguments have no effect."
@@ -120,13 +117,6 @@ def check_options(options):
@hookimpl
def validate(pdfinfo, options):
if not options.tesseract_env:
return
# If we are running a Tesseract spoof, ensure it knows what the input file is
if os.environ.get('PYTEST_CURRENT_TEST'):
options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = os.fspath(options.input_file)
# Tesseract 4.x can be multithreaded, and we also run multiple workers. We want
# to manage how many threads it uses to avoid creating total threads than cores.
# Performance testing shows we're better off
@@ -135,11 +125,11 @@ def validate(pdfinfo, options):
# input file is small, then we allow Tesseract to use threads, subject to the
# constraint: (ocrmypdf workers) * (tesseract threads) <= max_workers.
# As of Tesseract 4.1, 3 threads is the most effective on a 4 core/8 thread system.
if not options.tesseract_env.get('OMP_THREAD_LIMIT', '').isnumeric():
if not os.environ.get('OMP_THREAD_LIMIT', '').isnumeric():
tess_threads = min(3, options.jobs // len(pdfinfo), len(pdfinfo))
options.tesseract_env['OMP_THREAD_LIMIT'] = str(tess_threads)
os.environ['OMP_THREAD_LIMIT'] = str(tess_threads)
else:
tess_threads = int(options.tesseract_env['OMP_THREAD_LIMIT'])
tess_threads = int(os.environ['OMP_THREAD_LIMIT'])
if tess_threads > 1:
log.info("Using Tesseract OpenMP thread limit %d", tess_threads)
@@ -160,7 +150,7 @@ class TesseractOcrEngine(OcrEngine):
@staticmethod
def languages(options):
return tesseract.get_languages(options.tesseract_env)
return tesseract.get_languages()
@staticmethod
def get_orientation(input_file, options):
@@ -168,7 +158,6 @@ class TesseractOcrEngine(OcrEngine):
input_file,
engine_mode=options.tesseract_oem,
timeout=options.tesseract_timeout,
tesseract_env=options.tesseract_env,
)
@staticmethod
@@ -184,7 +173,6 @@ class TesseractOcrEngine(OcrEngine):
pagesegmode=options.tesseract_pagesegmode,
user_words=options.user_words,
user_patterns=options.user_patterns,
tesseract_env=options.tesseract_env,
)
@staticmethod
@@ -200,7 +188,6 @@ class TesseractOcrEngine(OcrEngine):
pagesegmode=options.tesseract_pagesegmode,
user_words=options.user_words,
user_patterns=options.user_patterns,
tesseract_env=options.tesseract_env,
)
src/ocrmypdf/exec/tesseract.py
+12 -45
View File
@@ -65,11 +65,11 @@ class TesseractLoggerAdapter(logging.LoggerAdapter):
return '[tesseract] %s' % (msg), kwargs
def version(tesseract_env=None):
return get_version('tesseract', regex=r'tesseract\s(.+)', env=tesseract_env)
def version():
return get_version('tesseract', regex=r'tesseract\s(.+)')
def has_textonly_pdf(tesseract_env=None, langs=None):
def has_textonly_pdf(langs=None):
"""Does Tesseract have textonly_pdf capability?
Available in v4.00.00alpha since January 2017. Best to
@@ -79,12 +79,7 @@ def has_textonly_pdf(tesseract_env=None, langs=None):
params = ''
try:
proc = run(
args_tess,
check=True,
universal_newlines=True,
stdout=PIPE,
stderr=STDOUT,
env=tesseract_env,
args_tess, check=True, universal_newlines=True, stdout=PIPE, stderr=STDOUT
)
params = proc.stdout
except CalledProcessError as e:
@@ -97,16 +92,16 @@ def has_textonly_pdf(tesseract_env=None, langs=None):
return False
def has_user_words(tesseract_env=None):
def has_user_words():
"""Does Tesseract have --user-words capability?
Not available in 4.0, but available in 4.1. Also available in 3.x, but
we no longer support 3.x.
"""
return version(tesseract_env) >= '4.1'
return version() >= '4.1'
def get_languages(tesseract_env=None):
def get_languages():
def lang_error(output):
msg = (
"Tesseract failed to report available languages.\n"
@@ -119,12 +114,7 @@ def get_languages(tesseract_env=None):
args_tess = ['tesseract', '--list-langs']
try:
proc = run(
args_tess,
universal_newlines=True,
stdout=PIPE,
stderr=STDOUT,
check=True,
env=tesseract_env,
args_tess, universal_newlines=True, stdout=PIPE, stderr=STDOUT, check=True
)
output = proc.stdout
except CalledProcessError as e:
@@ -146,7 +136,7 @@ def tess_base_args(langs: List[str], engine_mode) -> List[str]:
return args
def get_orientation(input_file: Path, engine_mode, timeout: float, tesseract_env=None):
def get_orientation(input_file: Path, engine_mode, timeout: float):
args_tesseract = tess_base_args(['osd'], engine_mode) + [
'--psm',
'0',
@@ -155,14 +145,7 @@ def get_orientation(input_file: Path, engine_mode, timeout: float, tesseract_env
]
try:
p = run(
args_tesseract,
stdout=PIPE,
stderr=STDOUT,
timeout=timeout,
check=True,
env=tesseract_env,
)
p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
stdout = p.stdout
except TimeoutExpired:
return OrientationConfidence(angle=0, confidence=0.0)
@@ -257,7 +240,6 @@ def generate_hocr(
pagesegmode: int,
user_words,
user_patterns,
tesseract_env,
):
prefix = output_hocr.with_suffix('')
@@ -276,14 +258,7 @@ def generate_hocr(
# to the number of order parameters here
args_tesseract.extend([input_file, prefix, 'hocr', 'txt'] + tessconfig)
try:
p = run(
args_tesseract,
stdout=PIPE,
stderr=STDOUT,
timeout=timeout,
check=True,
env=tesseract_env,
)
p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
stdout = p.stdout
except TimeoutExpired:
# Generate a HOCR file with no recognized text if tesseract times out
@@ -325,7 +300,6 @@ def generate_pdf(
pagesegmode: int,
user_words,
user_patterns,
tesseract_env,
):
"""Use Tesseract to render a PDF.
@@ -358,14 +332,7 @@ def generate_pdf(
args_tesseract.extend([input_file, prefix, 'pdf', 'txt'] + tessconfig)
try:
p = run(
args_tesseract,
stdout=PIPE,
stderr=STDOUT,
timeout=timeout,
check=True,
env=tesseract_env,
)
p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
stdout = p.stdout
if os.path.exists(prefix + '.txt'):
shutil.move(prefix + '.txt', output_text)
tests/conftest.py
-2
View File
@@ -133,8 +133,6 @@ def run_ocrmypdf_api(input_file, output_file, *args):
str(arg) for arg in args if arg is not None
]
_parser, options, plugin_manager = get_parser_options_plugins(args=args)
if options.tesseract_env:
assert all(isinstance(v, (str, bytes)) for v in options.tesseract_env.values())
api.check_options(options, plugin_manager)
return api.run_pipeline(options, plugin_manager=None, api=False)
tests/test_tesseract.py
+3 -7
View File
@@ -70,13 +70,11 @@ def test_content_preservation(resources, outpdf):
assert len(page.images) > 1, "masks were rasterized"
def test_no_languages(tmp_path):
env = os.environ.copy()
def test_no_languages(tmp_path, monkeypatch):
(tmp_path / 'tessdata').mkdir()
env['TESSDATA_PREFIX'] = fspath(tmp_path)
monkeypatch.setenv('TESSDATA_PREFIX', fspath(tmp_path))
with pytest.raises(MissingDependencyError):
tesseract.get_languages(tesseract_env=env)
tesseract.get_languages()
def test_image_too_large_hocr(monkeypatch, resources, outdir):
@@ -95,7 +93,6 @@ def test_image_too_large_hocr(monkeypatch, resources, outdir):
pagesegmode=None,
user_words=None,
user_patterns=None,
tesseract_env=None,
)
assert "name='ocr-capabilities'" in Path(outdir / 'out.hocr').read_text()
@@ -116,7 +113,6 @@ def test_image_too_large_pdf(monkeypatch, resources, outdir):
pagesegmode=None,
user_words=None,
user_patterns=None,
tesseract_env=None,
)
assert Path(outdir / 'txt.txt').read_text() == '[skipped page]'
if os.name != 'nt': # different semantics