Merge remote-tracking branch 'origin/dev'

docs/configuration.md

@@ -704,6 +704,20 @@ but could result in missing text content.
         this value if you are certain your documents are not malicious and
         you need the text which was not OCRed
 
+#### [`PAPERLESS_OCR_COLOR_CONVERSION_STRATEGY=<RGB>`](#PAPERLESS_OCR_COLOR_CONVERSION_STRATEGY) {#PAPERLESS_OCR_COLOR_CONVERSION_STRATEGY}
+
+: Controls the Ghostscript color conversion strategy when creating the archive file. This setting
+will only be utilized if the output is a version of PDF/A.
+
+    Valid options are CMYK, Gray, LeaveColorUnchanged, RGB or UseDeviceIndependentColor.
+
+    You can find more on the settings [here](https://ghostscript.readthedocs.io/en/latest/VectorDevices.html#color-conversion-and-management) in the Ghostscript documentation.
+
+    !!! warning
+
+        Utilizing some of the options may result in errors when creating archive
+        files from PDFs.
+
 #### [`PAPERLESS_OCR_USER_ARGS=<json>`](#PAPERLESS_OCR_USER_ARGS) {#PAPERLESS_OCR_USER_ARGS}
 
 : OCRmyPDF offers many more options. Use this parameter to specify any

src-ui/src/app/components/common/share-links-dropdown/share-links-dropdown.component.spec.ts

@@ -192,4 +192,24 @@ describe('ShareLinksDropdownComponent', () => {
     component.share(link)
     // expect(navigatorSpy).toHaveBeenCalledWith({ url: component.getShareUrl(link) })
   })
+
+  it('should correctly generate share URLs', () => {
+    environment.apiBaseUrl = 'http://example.com/api/'
+    expect(component.getShareUrl({ slug: '123abc123' } as any)).toEqual(
+      'http://example.com/share/123abc123'
+    )
+    environment.apiBaseUrl = 'http://example.domainwithapiinit.com/api/'
+    expect(component.getShareUrl({ slug: '123abc123' } as any)).toEqual(
+      'http://example.domainwithapiinit.com/share/123abc123'
+    )
+    environment.apiBaseUrl = 'http://example.domainwithapiinit.com:1234/api/'
+    expect(component.getShareUrl({ slug: '123abc123' } as any)).toEqual(
+      'http://example.domainwithapiinit.com:1234/share/123abc123'
+    )
+    environment.apiBaseUrl =
+      'http://example.domainwithapiinit.com:1234/subpath/api/'
+    expect(component.getShareUrl({ slug: '123abc123' } as any)).toEqual(
+      'http://example.domainwithapiinit.com:1234/subpath/share/123abc123'
+    )
+  })
 })

src-ui/src/app/components/common/share-links-dropdown/share-links-dropdown.component.ts

@@ -80,7 +80,10 @@ export class ShareLinksDropdownComponent implements OnInit {
   }
 
   getShareUrl(link: PaperlessShareLink): string {
-    return `${environment.apiBaseUrl.replace('api', 'share')}${link.slug}`
+    const apiURL = new URL(environment.apiBaseUrl)
+    return `${apiURL.origin}${apiURL.pathname.replace(/\/api\/$/, '/share/')}${
+      link.slug
+    }`
   }
 
   getDaysRemaining(link: PaperlessShareLink): string {

src-ui/src/environments/environment.prod.ts

@@ -5,7 +5,7 @@ export const environment = {
   apiBaseUrl: document.baseURI + 'api/',
   apiVersion: '3',
   appTitle: 'Paperless-ngx',
-  version: '2.0.0',
+  version: '2.0.0-dev',
   webSocketHost: window.location.host,
   webSocketProtocol: window.location.protocol == 'https:' ? 'wss:' : 'ws:',
   webSocketBaseUrl: base_url.pathname + 'ws/',

src-ui/src/locale/messages.de_DE.xlf

@@ -3019,7 +3019,7 @@
           <context context-type="sourcefile">src/app/components/common/edit-dialog/custom-field-edit-dialog/custom-field-edit-dialog.component.ts</context>
           <context context-type="linenumber">39</context>
         </context-group>
-        <target state="translated">Neues benutzerdefinidertes Feld erstellen</target>
+        <target state="translated">Neues benutzerdefiniertes Feld erstellen</target>
       </trans-unit>
       <trans-unit id="8751213029607178010" datatype="html">
         <source>Edit custom field</source>

src/documents/migrations/1041_alter_consumptiontemplate_sources.py

@@ -0,0 +1,22 @@
+# Generated by Django 4.2.7 on 2023-11-30 14:29
+
+import multiselectfield.db.fields
+from django.db import migrations
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("documents", "1040_customfield_customfieldinstance_and_more"),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name="consumptiontemplate",
+            name="sources",
+            field=multiselectfield.db.fields.MultiSelectField(
+                choices=[(1, "Consume Folder"), (2, "Api Upload"), (3, "Mail Fetch")],
+                default="1,2,3",
+                max_length=5,
+            ),
+        ),
+    ]

src/documents/models.py

@@ -754,7 +754,7 @@ class ConsumptionTemplate(models.Model):
     order = models.IntegerField(_("order"), default=0)
 
     sources = MultiSelectField(
-        max_length=3,
+        max_length=5,
         choices=DocumentSourceChoices.choices,
         default=f"{DocumentSource.ConsumeFolder},{DocumentSource.ApiUpload},{DocumentSource.MailFetch}",
     )

src/locale/ru_RU/LC_MESSAGES/django.po

@@ -3,7 +3,7 @@ msgstr ""
 "Project-Id-Version: paperless-ngx\n"
 "Report-Msgid-Bugs-To: \n"
 "POT-Creation-Date: 2023-11-10 09:02-0800\n"
-"PO-Revision-Date: 2023-11-13 01:31\n"
+"PO-Revision-Date: 2023-11-29 08:50\n"
 "Last-Translator: \n"
 "Language-Team: Russian\n"
 "Language: ru_RU\n"
@@ -409,7 +409,7 @@ msgstr ""
 
 #: documents/models.py:454
 msgid "owner is"
-msgstr ""
+msgstr "владелец"
 
 #: documents/models.py:455
 msgid "has owner in"
@@ -417,7 +417,7 @@ msgstr ""
 
 #: documents/models.py:456
 msgid "does not have owner"
-msgstr ""
+msgstr "не имеет владельца"
 
 #: documents/models.py:457
 msgid "does not have owner in"
@@ -521,7 +521,7 @@ msgstr "пользователь"
 
 #: documents/models.py:679
 msgid "note"
-msgstr ""
+msgstr "заметка"
 
 #: documents/models.py:680
 msgid "notes"
@@ -529,15 +529,15 @@ msgstr "заметки"
 
 #: documents/models.py:688
 msgid "Archive"
-msgstr ""
+msgstr "Архив"
 
 #: documents/models.py:689
 msgid "Original"
-msgstr ""
+msgstr "Оригинал"
 
 #: documents/models.py:700
 msgid "expiration"
-msgstr ""
+msgstr "Срок действия"
 
 #: documents/models.py:707
 msgid "slug"
@@ -645,7 +645,7 @@ msgstr ""
 
 #: documents/models.py:887
 msgid "URL"
-msgstr ""
+msgstr "Ссылка"
 
 #: documents/models.py:888
 msgid "Date"
@@ -661,7 +661,7 @@ msgstr ""
 
 #: documents/models.py:891
 msgid "Float"
-msgstr ""
+msgstr "С плавающей точкой"
 
 #: documents/models.py:892
 msgid "Monetary"
@@ -669,15 +669,15 @@ msgstr ""
 
 #: documents/models.py:904
 msgid "data type"
-msgstr ""
+msgstr "Тип данных"
 
 #: documents/models.py:912
 msgid "custom field"
-msgstr ""
+msgstr "Пользовательское поле"
 
 #: documents/models.py:913
 msgid "custom fields"
-msgstr ""
+msgstr "Пользовательские поля"
 
 #: documents/models.py:973
 msgid "custom field instance"

src/paperless/settings.py

@@ -864,6 +864,11 @@ OCR_MAX_IMAGE_PIXELS: Optional[int] = None
 if os.environ.get("PAPERLESS_OCR_MAX_IMAGE_PIXELS") is not None:
     OCR_MAX_IMAGE_PIXELS: int = int(os.environ.get("PAPERLESS_OCR_MAX_IMAGE_PIXELS"))
 
+OCR_COLOR_CONVERSION_STRATEGY = os.getenv(
+    "PAPERLESS_OCR_COLOR_CONVERSION_STRATEGY",
+    "RGB",
+)
+
 OCR_USER_ARGS = os.getenv("PAPERLESS_OCR_USER_ARGS", "{}")
 
 # GNUPG needs a home directory for some reason

src/paperless_tesseract/parsers.py

@@ -186,6 +186,11 @@ class RasterisedDocumentParser(DocumentParser):
             "progress_bar": False,
         }
 
+        if "pdfa" in ocrmypdf_args["output_type"]:
+            ocrmypdf_args[
+                "color_conversion_strategy"
+            ] = settings.OCR_COLOR_CONVERSION_STRATEGY
+
         if settings.OCR_MODE == "force" or safe_fallback:
             ocrmypdf_args["force_ocr"] = True
         elif settings.OCR_MODE in ["skip", "skip_noarchive"]:
@@ -244,6 +249,10 @@ class RasterisedDocumentParser(DocumentParser):
                     f"no DPI information is present in this image and "
                     f"OCR_IMAGE_DPI is not set.",
                 )
+            if ocrmypdf_args["image_dpi"] < 70:  # pragma: no cover
+                self.log.warning(
+                    f"Image DPI of {ocrmypdf_args['image_dpi']} is low, OCR may fail",
+                )
 
         if settings.OCR_USER_ARGS and not safe_fallback:
             try: