paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-10-30 03:56:23 -05:00
Code Issues Packages Projects Releases Wiki Activity

Enhancement: add barcode frontend config (#9742)

Browse Source
This commit is contained in:
shamoon
2025-05-11 12:44:06 -07:00
committed by GitHub
parent bcb0ae1ee5
commit 6a5be992c0
9 changed files with 437 additions and 49 deletions
Download Patch File Download Diff File Expand all files Collapse all files

73
src/documents/barcodes.py
View File

@@ -15,13 +15,16 @@ from pikepdf import Pdf
from documents.converters import convert_from_tiff_to_pdf
from documents.data_models import ConsumableDocument
from documents.data_models import DocumentMetadataOverrides
from documents.models import Tag
from documents.plugins.base import ConsumeTaskPlugin
from documents.plugins.base import StopConsumeTaskError
from documents.plugins.helpers import ProgressManager
from documents.plugins.helpers import ProgressStatusOptions
from documents.utils import copy_basic_file_stats
from documents.utils import copy_file_with_basic_stats
from documents.utils import maybe_override_pixel_limit
from paperless.config import BarcodeConfig
if TYPE_CHECKING:
from collections.abc import Callable
@@ -39,6 +42,7 @@ class Barcode:
page: int
value: str
settings: BarcodeConfig
@property
def is_separator(self) -> bool:
@@ -46,7 +50,7 @@ class Barcode:
Returns True if the barcode value equals the configured separation value,
False otherwise
"""
return self.value == settings.CONSUMER_BARCODE_STRING
return self.value == self.settings.barcode_string
@property
def is_asn(self) -> bool:
@@ -54,7 +58,7 @@ class Barcode:
Returns True if the barcode value matches the configured ASN prefix,
False otherwise
"""
return self.value.startswith(settings.CONSUMER_ASN_BARCODE_PREFIX)
return self.value.startswith(self.settings.barcode_asn_prefix)
class BarcodePlugin(ConsumeTaskPlugin):
@@ -67,17 +71,41 @@ class BarcodePlugin(ConsumeTaskPlugin):
- ASN from barcode detection is enabled or
- Barcode support is enabled and the mime type is supported
"""
if settings.CONSUMER_BARCODE_TIFF_SUPPORT:
if self.settings.barcode_enable_tiff_support:
supported_mimes: set[str] = {"application/pdf", "image/tiff"}
else:
supported_mimes = {"application/pdf"}
return (
settings.CONSUMER_ENABLE_ASN_BARCODE
or settings.CONSUMER_ENABLE_BARCODES
or settings.CONSUMER_ENABLE_TAG_BARCODE
self.settings.barcode_enable_asn
or self.settings.barcodes_enabled
or self.settings.barcode_enable_tag
) and self.input_doc.mime_type in supported_mimes
def get_settings(self) -> BarcodeConfig:
"""
Returns the settings for this plugin (Django settings or app config)
"""
return BarcodeConfig()
def __init__(
self,
input_doc: ConsumableDocument,
metadata: DocumentMetadataOverrides,
status_mgr: ProgressManager,
base_tmp_dir: Path,
task_id: str,
) -> None:
super().__init__(
input_doc,
metadata,
status_mgr,
base_tmp_dir,
task_id,
)
# need these for able_to_run
self.settings = self.get_settings()
def setup(self) -> None:
self.temp_dir = tempfile.TemporaryDirectory(
dir=self.base_tmp_dir,
@@ -99,7 +127,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
# try reading tags from barcodes
if (
settings.CONSUMER_ENABLE_TAG_BARCODE
self.settings.barcode_enable_tag
and (tags := self.tags) is not None
and len(tags) > 0
):
@@ -110,7 +138,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
logger.info(f"Found tags in barcode: {tags}")
# Lastly attempt to split documents
if settings.CONSUMER_ENABLE_BARCODES and (
if self.settings.barcodes_enabled and (
separator_pages := self.get_separation_pages()
):
# We have pages to split against
@@ -155,10 +183,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
# Update/overwrite an ASN if possible
# After splitting, as otherwise each split document gets the same ASN
if (
settings.CONSUMER_ENABLE_ASN_BARCODE
and (located_asn := self.asn) is not None
):
if self.settings.barcode_enable_asn and (located_asn := self.asn) is not None:
logger.info(f"Found ASN in barcode: {located_asn}")
self.metadata.asn = located_asn
@@ -245,8 +270,8 @@ class BarcodePlugin(ConsumeTaskPlugin):
# Get limit from configuration
barcode_max_pages: int = (
num_of_pages
if settings.CONSUMER_BARCODE_MAX_PAGES == 0
else settings.CONSUMER_BARCODE_MAX_PAGES
if self.settings.barcode_max_pages == 0
else self.settings.barcode_max_pages
)
if barcode_max_pages < num_of_pages: # pragma: no cover
@@ -261,7 +286,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
# Convert page to image
page = convert_from_path(
self.pdf_file,
dpi=settings.CONSUMER_BARCODE_DPI,
dpi=self.settings.barcode_dpi,
output_folder=self.temp_dir.name,
first_page=current_page_number + 1,
last_page=current_page_number + 1,
@@ -272,7 +297,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
logger.debug(f"Image is at {page_filepath}")
# Upscale image if configured
factor = settings.CONSUMER_BARCODE_UPSCALE
factor = self.settings.barcode_upscale
if factor > 1.0:
logger.debug(
f"Upscaling image by {factor} for better barcode detection",
@@ -285,7 +310,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
# Detect barcodes
for barcode_value in reader(page):
self.barcodes.append(
Barcode(current_page_number, barcode_value),
Barcode(current_page_number, barcode_value, self.settings),
)
# Delete temporary image file
@@ -308,7 +333,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
def asn(self) -> int | None:
"""
Search the parsed barcodes for any ASNs.
The first barcode that starts with CONSUMER_ASN_BARCODE_PREFIX
The first barcode that starts with barcode_asn_prefix
is considered the ASN to be used.
Returns the detected ASN (or None)
"""
@@ -317,7 +342,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
# Ensure the barcodes have been read
self.detect()
# get the first barcode that starts with CONSUMER_ASN_BARCODE_PREFIX
# get the first barcode that starts with barcode_asn_prefix
asn_text: str | None = next(
(x.value for x in self.barcodes if x.is_asn),
None,
@@ -326,7 +351,7 @@ class BarcodePlugin(ConsumeTaskPlugin):
if asn_text:
logger.debug(f"Found ASN Barcode: {asn_text}")
# remove the prefix and remove whitespace
asn_text = asn_text[len(settings.CONSUMER_ASN_BARCODE_PREFIX) :].strip()
asn_text = asn_text[len(self.settings.barcode_asn_prefix) :].strip()
# remove non-numeric parts of the remaining string
asn_text = re.sub(r"\D", "", asn_text)
@@ -356,9 +381,9 @@ class BarcodePlugin(ConsumeTaskPlugin):
for raw in tag_texts.split(","):
try:
tag_str: str | None = None
for regex in settings.CONSUMER_TAG_BARCODE_MAPPING:
for regex in self.settings.barcode_tag_mapping:
if re.match(regex, raw, flags=re.IGNORECASE):
sub = settings.CONSUMER_TAG_BARCODE_MAPPING[regex]
sub = self.settings.barcode_tag_mapping[regex]
tag_str = (
re.sub(regex, sub, raw, flags=re.IGNORECASE)
if sub
@@ -394,13 +419,13 @@ class BarcodePlugin(ConsumeTaskPlugin):
"""
# filter all barcodes for the separator string
# get the page numbers of the separating barcodes
retain = settings.CONSUMER_BARCODE_RETAIN_SPLIT_PAGES
retain = self.settings.barcode_retain_split_pages
separator_pages = {
bc.page: retain
for bc in self.barcodes
if bc.is_separator and (not retain or (retain and bc.page > 0))
} # as below, dont include the first page if retain is enabled
if not settings.CONSUMER_ENABLE_ASN_BARCODE:
if not self.settings.barcode_enable_asn:
return separator_pages
# add the page numbers of the ASN barcodes

57
src/documents/tests/test_api_app_config.py
View File

@@ -32,28 +32,39 @@ class TestApiAppConfig(DirectoriesMixin, APITestCase):
self.assertEqual(response.status_code, status.HTTP_200_OK)
self.assertEqual(
json.dumps(response.data[0]),
json.dumps(
{
"id": 1,
"user_args": None,
"output_type": None,
"pages": None,
"language": None,
"mode": None,
"skip_archive_file": None,
"image_dpi": None,
"unpaper_clean": None,
"deskew": None,
"rotate_pages": None,
"rotate_pages_threshold": None,
"max_image_pixels": None,
"color_conversion_strategy": None,
"app_title": None,
"app_logo": None,
},
),
self.maxDiff = None
self.assertDictEqual(
response.data[0],
{
"id": 1,
"output_type": None,
"pages": None,
"language": None,
"mode": None,
"skip_archive_file": None,
"image_dpi": None,
"unpaper_clean": None,
"deskew": None,
"rotate_pages": None,
"rotate_pages_threshold": None,
"max_image_pixels": None,
"color_conversion_strategy": None,
"user_args": None,
"app_title": None,
"app_logo": None,
"barcodes_enabled": None,
"barcode_enable_tiff_support": None,
"barcode_string": None,
"barcode_retain_split_pages": None,
"barcode_enable_asn": None,
"barcode_asn_prefix": None,
"barcode_upscale": None,
"barcode_dpi": None,
"barcode_max_pages": None,
"barcode_enable_tag": None,
"barcode_tag_mapping": None,
},
)
def test_api_get_ui_settings_with_config(self):
@@ -118,6 +129,7 @@ class TestApiAppConfig(DirectoriesMixin, APITestCase):
{
"user_args": "",
"language": "",
"barcode_tag_mapping": "",
},
),
content_type="application/json",
@@ -126,6 +138,7 @@ class TestApiAppConfig(DirectoriesMixin, APITestCase):
config = ApplicationConfiguration.objects.first()
self.assertEqual(config.user_args, None)
self.assertEqual(config.language, None)
self.assertEqual(config.barcode_tag_mapping, None)
def test_api_replace_app_logo(self):
"""

22
src/documents/tests/test_barcodes.py
View File

@@ -22,6 +22,7 @@ from documents.tests.utils import DocumentConsumeDelayMixin
from documents.tests.utils import DummyProgressManager
from documents.tests.utils import FileSystemAssertsMixin
from documents.tests.utils import SampleDirMixin
from paperless.models import ApplicationConfiguration
try:
import zxingcpp # noqa: F401
@@ -547,6 +548,27 @@ class TestBarcode(
},
)
def test_barcode_config(self):
"""
GIVEN:
- Barcode app config is set (settings are not)
WHEN:
- Document with barcode is processed
THEN:
- The barcode config is used
"""
app_config = ApplicationConfiguration.objects.first()
app_config.barcodes_enabled = True
app_config.barcode_string = "CUSTOM BARCODE"
app_config.save()
test_file = self.BARCODE_SAMPLE_DIR / "barcode-39-custom.pdf"
with self.get_reader(test_file) as reader:
reader.detect()
separator_page_numbers = reader.get_separation_pages()
self.assertEqual(reader.pdf_file, test_file)
self.assertDictEqual(separator_page_numbers, {0: False})
@override_settings(CONSUMER_BARCODE_SCANNER="PYZBAR")
class TestBarcodeNewConsume(

57
src/paperless/config.py
View File

@@ -96,10 +96,65 @@ class OcrConfig(OutputTypeConfig):
user_args = json.loads(settings.OCR_USER_ARGS)
except json.JSONDecodeError:
user_args = {}
self.user_args = user_args
@dataclasses.dataclass
class BarcodeConfig(BaseConfig):
"""
Barcodes settings
"""
barcodes_enabled: bool = dataclasses.field(init=False)
barcode_enable_tiff_support: bool = dataclasses.field(init=False)
barcode_string: str = dataclasses.field(init=False)
barcode_retain_split_pages: bool = dataclasses.field(init=False)
barcode_enable_asn: bool = dataclasses.field(init=False)
barcode_asn_prefix: str = dataclasses.field(init=False)
barcode_upscale: float = dataclasses.field(init=False)
barcode_dpi: int = dataclasses.field(init=False)
barcode_max_pages: int = dataclasses.field(init=False)
barcode_enable_tag: bool = dataclasses.field(init=False)
barcode_tag_mapping: dict[str, str] = dataclasses.field(init=False)
def __post_init__(self) -> None:
app_config = self._get_config_instance()
self.barcodes_enabled = (
app_config.barcodes_enabled or settings.CONSUMER_ENABLE_BARCODES
)
self.barcode_enable_tiff_support = (
app_config.barcode_enable_tiff_support
or settings.CONSUMER_BARCODE_TIFF_SUPPORT
)
self.barcode_string = (
app_config.barcode_string or settings.CONSUMER_BARCODE_STRING
)
self.barcode_retain_split_pages = (
app_config.barcode_retain_split_pages
or settings.CONSUMER_BARCODE_RETAIN_SPLIT_PAGES
)
self.barcode_enable_asn = (
app_config.barcode_enable_asn or settings.CONSUMER_ENABLE_ASN_BARCODE
)
self.barcode_asn_prefix = (
app_config.barcode_asn_prefix or settings.CONSUMER_ASN_BARCODE_PREFIX
)
self.barcode_upscale = (
app_config.barcode_upscale or settings.CONSUMER_BARCODE_UPSCALE
)
self.barcode_dpi = app_config.barcode_dpi or settings.CONSUMER_BARCODE_DPI
self.barcode_max_pages = (
app_config.barcode_max_pages or settings.CONSUMER_BARCODE_MAX_PAGES
)
self.barcode_enable_tag = (
app_config.barcode_enable_tag or settings.CONSUMER_ENABLE_TAG_BARCODE
)
self.barcode_tag_mapping = (
app_config.barcode_tag_mapping or settings.CONSUMER_TAG_BARCODE_MAPPING
)
@dataclasses.dataclass
class GeneralConfig(BaseConfig):
"""

100
src/paperless/migrations/0004_applicationconfiguration_barcode_asn_prefix_and_more.py Normal file
View File

@@ -0,0 +1,100 @@
# Generated by Django 5.1.7 on 2025-04-02 19:21
import django.core.validators
from django.db import migrations
from django.db import models
class Migration(migrations.Migration):
dependencies = [
("paperless", "0003_alter_applicationconfiguration_max_image_pixels"),
]
operations = [
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_asn_prefix",
field=models.CharField(
blank=True,
max_length=32,
null=True,
verbose_name="Sets the ASN barcode prefix",
),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_dpi",
field=models.PositiveIntegerField(
null=True,
validators=[django.core.validators.MinValueValidator(1)],
verbose_name="Sets the barcode DPI",
),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_enable_asn",
field=models.BooleanField(null=True, verbose_name="Enables ASN barcode"),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_enable_tag",
field=models.BooleanField(null=True, verbose_name="Enables tag barcode"),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_enable_tiff_support",
field=models.BooleanField(
null=True,
verbose_name="Enables barcode TIFF support",
),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_max_pages",
field=models.PositiveIntegerField(
null=True,
validators=[django.core.validators.MinValueValidator(1)],
verbose_name="Sets the maximum pages for barcode",
),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_retain_split_pages",
field=models.BooleanField(null=True, verbose_name="Retains split pages"),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_string",
field=models.CharField(
blank=True,
max_length=32,
null=True,
verbose_name="Sets the barcode string",
),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_tag_mapping",
field=models.JSONField(
null=True,
verbose_name="Sets the tag barcode mapping",
),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcode_upscale",
field=models.FloatField(
null=True,
validators=[django.core.validators.MinValueValidator(1.0)],
verbose_name="Sets the barcode upscale factor",
),
),
migrations.AddField(
model_name="applicationconfiguration",
name="barcodes_enabled",
field=models.BooleanField(
null=True,
verbose_name="Enables barcode scanning",
),
),
]

81
src/paperless/models.py
View File

@@ -167,6 +167,10 @@ class ApplicationConfiguration(AbstractSingletonModel):
null=True,
)
"""
Settings for the Paperless application
"""
app_title = models.CharField(
verbose_name=_("Application title"),
null=True,
@@ -184,6 +188,83 @@ class ApplicationConfiguration(AbstractSingletonModel):
upload_to="logo/",
)
"""
Settings for the barcode scanner
"""
# PAPERLESS_CONSUMER_ENABLE_BARCODES
barcodes_enabled = models.BooleanField(
verbose_name=_("Enables barcode scanning"),
null=True,
)
# PAPERLESS_CONSUMER_BARCODE_TIFF_SUPPORT
barcode_enable_tiff_support = models.BooleanField(
verbose_name=_("Enables barcode TIFF support"),
null=True,
)
# PAPERLESS_CONSUMER_BARCODE_STRING
barcode_string = models.CharField(
verbose_name=_("Sets the barcode string"),
null=True,
blank=True,
max_length=32,
)
# PAPERLESS_CONSUMER_BARCODE_RETAIN_SPLIT_PAGES
barcode_retain_split_pages = models.BooleanField(
verbose_name=_("Retains split pages"),
null=True,
)
# PAPERLESS_CONSUMER_ENABLE_ASN_BARCODE
barcode_enable_asn = models.BooleanField(
verbose_name=_("Enables ASN barcode"),
null=True,
)
# PAPERLESS_CONSUMER_ASN_BARCODE_PREFIX
barcode_asn_prefix = models.CharField(
verbose_name=_("Sets the ASN barcode prefix"),
null=True,
blank=True,
max_length=32,
)
# PAPERLESS_CONSUMER_BARCODE_UPSCALE
barcode_upscale = models.FloatField(
verbose_name=_("Sets the barcode upscale factor"),
null=True,
validators=[MinValueValidator(1.0)],
)
# PAPERLESS_CONSUMER_BARCODE_DPI
barcode_dpi = models.PositiveIntegerField(
verbose_name=_("Sets the barcode DPI"),
null=True,
validators=[MinValueValidator(1)],
)
# PAPERLESS_CONSUMER_BARCODE_MAX_PAGES
barcode_max_pages = models.PositiveIntegerField(
verbose_name=_("Sets the maximum pages for barcode"),
null=True,
validators=[MinValueValidator(1)],
)
# PAPERLESS_CONSUMER_ENABLE_TAG_BARCODE
barcode_enable_tag = models.BooleanField(
verbose_name=_("Enables tag barcode"),
null=True,
)
# PAPERLESS_CONSUMER_TAG_BARCODE_MAPPING
barcode_tag_mapping = models.JSONField(
verbose_name=_("Sets the tag barcode mapping"),
null=True,
)
class Meta:
verbose_name = _("paperless application settings")

3
src/paperless/serialisers.py
View File

@@ -185,11 +185,14 @@ class ProfileSerializer(serializers.ModelSerializer):
class ApplicationConfigurationSerializer(serializers.ModelSerializer):
user_args = serializers.JSONField(binary=True, allow_null=True)
barcode_tag_mapping = serializers.JSONField(binary=True, allow_null=True)
def run_validation(self, data):
# Empty strings treated as None to avoid unexpected behavior
if "user_args" in data and data["user_args"] == "":
data["user_args"] = None
if "barcode_tag_mapping" in data and data["barcode_tag_mapping"] == "":
data["barcode_tag_mapping"] = None
if "language" in data and data["language"] == "":
data["language"] = None
return super().run_validation(data)