matthias/mayan-edms
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Refactor OCR app. Removes document parsing. Moves OCR processing to

Browse Source 
model manager. Add submit and finish events.

Signed-off-by: Roberto Rosario <roberto.rosario.gonzalez@gmail.com>
This commit is contained in:
Roberto Rosario
2017-08-23 02:04:57 -04:00
parent 2052caada4
commit 317d07a355
20 changed files with 309 additions and 497 deletions
Download Patch File Download Diff File Expand all files Collapse all files

45
mayan/apps/ocr/models.py
View File

@@ -6,6 +6,8 @@ from django.utils.translation import ugettext_lazy as _
from documents.models import DocumentPage, DocumentType, DocumentVersion
from .managers import DocumentPageOCRContentManager
class DocumentTypeSettings(models.Model):
"""
@@ -25,6 +27,24 @@ class DocumentTypeSettings(models.Model):
verbose_name_plural = _('Document types settings')
@python_2_unicode_compatible
class DocumentPageOCRContent(models.Model):
document_page = models.OneToOneField(
DocumentPage, on_delete=models.CASCADE, related_name='ocr_content',
verbose_name=_('Document page')
)
content = models.TextField(blank=True, verbose_name=_('Content'))
objects = DocumentPageOCRContentManager()
def __str__(self):
return force_text(self.document_page)
class Meta:
verbose_name = _('Document page OCR content')
verbose_name_plural = _('Document pages OCR contents')
@python_2_unicode_compatible
class DocumentVersionOCRError(models.Model):
document_version = models.ForeignKey(
@@ -32,7 +52,7 @@ class DocumentVersionOCRError(models.Model):
verbose_name=_('Document version')
)
datetime_submitted = models.DateTimeField(
auto_now=True, db_index=True, verbose_name=_('Date time submitted')
auto_now_add=True, db_index=True, verbose_name=_('Date time submitted')
)
result = models.TextField(blank=True, null=True, verbose_name=_('Result'))
@@ -41,24 +61,5 @@ class DocumentVersionOCRError(models.Model):
class Meta:
ordering = ('datetime_submitted',)
verbose_name = _('Document Version OCR Error')
verbose_name_plural = _('Document Version OCR Errors')
@python_2_unicode_compatible
class DocumentPageContent(models.Model):
"""
Model that describes a document page content
"""
document_page = models.OneToOneField(
DocumentPage, on_delete=models.CASCADE, related_name='ocr_content',
verbose_name=_('Document page')
)
content = models.TextField(blank=True, verbose_name=_('Content'))
def __str__(self):
return force_text(self.document_page)
class Meta:
verbose_name = _('Document page content')
verbose_name_plural = _('Document pages contents')
verbose_name = _('Document version OCR error')
verbose_name_plural = _('Document version OCR errors')