paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-04-02 13:45:10 -05:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'jonaswinkler-new-features'

Browse Source
This commit is contained in:
Daniel Quinn 2018-09-23 12:42:02 +01:00
commit f4a09013d7
9 changed files with 410 additions and 59 deletions

2
.gitignore vendored

@ -81,3 +81,5 @@ docker-compose.env
scripts/import-for-development
scripts/nuke
# Static files collected by the collectstatic command
static/

10
paperless.conf.example

@ -59,6 +59,11 @@ PAPERLESS_EMAIL_SECRET=""
#### Security ####
###############################################################################
# Controls whether django's debug mode is enabled. Disable this on production
# systems. Debug mode is enabled by default.
PAPERLESS_DEBUG="false"
# Paperless can be instructed to attempt to encrypt your PDF files with GPG
# using the PAPERLESS_PASSPHRASE specified below. If however you're not
# concerned about encrypting these files (for example if you have disk
@ -203,3 +208,8 @@ PAPERLESS_EMAIL_SECRET=""
# positive integer, but if you don't define one in paperless.conf, a default of
# 100 will be used.
#PAPERLESS_LIST_PER_PAGE=100
# The number of years for which a correspondent will be included in the recent
# correspondents filter.
#PAPERLESS_RECENT_CORRESPONDENT_YEARS=1

22
requirements.txt Normal file → Executable file

@ -1,10 +1,10 @@
-i https://pypi.python.org/simple
apipkg==1.5; python_version != '3.1.*'
atomicwrites==1.2.1; python_version != '3.1.*'
apipkg==1.5; python_version != '3.3.*'
atomicwrites==1.2.1; python_version != '3.3.*'
attrs==18.2.0
certifi==2018.8.24
chardet==3.0.4
coverage==4.5.1; python_version != '3.1.*'
coverage==4.5.1; python_version < '4'
coveralls==1.5.0
dateparser==0.7.0
django-cors-headers==2.4.0
@ -14,9 +14,9 @@ django-filter==2.0.0
django==2.0.8
djangorestframework==3.8.2
docopt==0.6.2
execnet==1.5.0; python_version != '3.1.*'
execnet==1.5.0; python_version != '3.3.*'
factory-boy==2.11.1
faker==0.9.0
faker==0.9.0; python_version >= '2.7'
filemagic==1.6
fuzzywuzzy==0.15.0
gunicorn==19.9.0
@ -26,17 +26,17 @@ langdetect==1.0.7
more-itertools==4.3.0
pdftotext==2.1.0
pillow==5.2.0
pluggy==0.7.1; python_version != '3.1.*'
py==1.6.0; python_version != '3.1.*'
pluggy==0.7.1; python_version != '3.3.*'
py==1.6.0; python_version != '3.3.*'
pycodestyle==2.4.0
pyocr==0.5.3
pytest-cov==2.5.1
pytest-cov==2.6.0
pytest-django==3.4.2
pytest-env==0.6.2
pytest-forked==0.2
pytest-forked==0.2; python_version != '3.3.*'
pytest-sugar==0.9.1
pytest-xdist==1.23.0
pytest==3.7.4
pytest==3.8.0
python-dateutil==2.7.3
python-dotenv==0.9.1
python-gnupg==0.4.3
@ -48,4 +48,4 @@ six==1.11.0
termcolor==1.1.0
text-unidecode==1.2
tzlocal==1.5.1
urllib3==1.23; python_version != '3.0.*'
urllib3==1.23; python_version != '3.3.*'

146
src/documents/actions.py Normal file

@ -0,0 +1,146 @@
from django.contrib import messages
from django.contrib.admin import helpers
from django.contrib.admin.utils import model_ngettext
from django.core.exceptions import PermissionDenied
from django.template.response import TemplateResponse
from documents.models import Correspondent, Tag
def select_action(
modeladmin, request, queryset, title, action, modelclass,
success_message="", document_action=None, queryset_action=None):
opts = modeladmin.model._meta
app_label = opts.app_label
if not modeladmin.has_change_permission(request):
raise PermissionDenied
if request.POST.get('post'):
n = queryset.count()
selected_object = modelclass.objects.get(id=request.POST.get('obj_id'))
if n:
for document in queryset:
if document_action:
document_action(document, selected_object)
document_display = str(document)
modeladmin.log_change(request, document, document_display)
if queryset_action:
queryset_action(queryset, selected_object)
modeladmin.message_user(request, success_message % {
"selected_object": selected_object.name,
"count": n,
"items": model_ngettext(modeladmin.opts, n)
}, messages.SUCCESS)
# Return None to display the change list page again.
return None
context = dict(
modeladmin.admin_site.each_context(request),
title=title,
queryset=queryset,
opts=opts,
action_checkbox_name=helpers.ACTION_CHECKBOX_NAME,
media=modeladmin.media,
action=action,
objects=modelclass.objects.all(),
itemname=model_ngettext(modelclass, 1)
)
request.current_app = modeladmin.admin_site.name
return TemplateResponse(
request,
"admin/{}/{}/select_object.html".format(app_label, opts.model_name),
context
)
def simple_action(
modeladmin, request, queryset, success_message="",
document_action=None, queryset_action=None):
if not modeladmin.has_change_permission(request):
raise PermissionDenied
n = queryset.count()
if n:
for document in queryset:
if document_action:
document_action(document)
document_display = str(document)
modeladmin.log_change(request, document, document_display)
if queryset_action:
queryset_action(queryset)
modeladmin.message_user(request, success_message % {
"count": n, "items": model_ngettext(modeladmin.opts, n)
}, messages.SUCCESS)
# Return None to display the change list page again.
return None
def add_tag_to_selected(modeladmin, request, queryset):
return select_action(
modeladmin=modeladmin,
request=request,
queryset=queryset,
title="Add tag to multiple documents",
action="add_tag_to_selected",
modelclass=Tag,
success_message="Successfully added tag %(selected_object)s to "
"%(count)d %(items)s.",
document_action=lambda doc, tag: doc.tags.add(tag)
)
def remove_tag_from_selected(modeladmin, request, queryset):
return select_action(
modeladmin=modeladmin,
request=request,
queryset=queryset,
title="Remove tag from multiple documents",
action="remove_tag_from_selected",
modelclass=Tag,
success_message="Successfully removed tag %(selected_object)s from "
"%(count)d %(items)s.",
document_action=lambda doc, tag: doc.tags.remove(tag)
)
def set_correspondent_on_selected(modeladmin, request, queryset):
return select_action(
modeladmin=modeladmin,
request=request,
queryset=queryset,
title="Set correspondent on multiple documents",
action="set_correspondent_on_selected",
modelclass=Correspondent,
success_message="Successfully set correspondent %(selected_object)s "
"on %(count)d %(items)s.",
queryset_action=lambda qs, corr: qs.update(correspondent=corr)
)
def remove_correspondent_from_selected(modeladmin, request, queryset):
return simple_action(
modeladmin=modeladmin,
request=request,
queryset=queryset,
success_message="Successfully removed correspondent from %(count)d "
"%(items)s.",
queryset_action=lambda qs: qs.update(correspondent=None)
)
add_tag_to_selected.short_description = "Add tag to selected documents"
remove_tag_from_selected.short_description = \
"Remove tag from selected documents"
set_correspondent_on_selected.short_description = \
"Set correspondent on selected documents"
remove_correspondent_from_selected.short_description = \
"Remove correspondent from selected documents"

205
src/documents/admin.py

@ -1,42 +1,25 @@
from datetime import datetime
from datetime import datetime, timedelta
from django.conf import settings
from django.contrib import admin
from django.contrib.auth.models import User, Group
try:
from django.core.urlresolvers import reverse
except ImportError:
from django.urls import reverse
from django.contrib import admin, messages
from django.contrib.admin.templatetags.admin_urls import add_preserved_filters
from django.contrib.auth.models import Group, User
from django.db import models
from django.http import HttpResponseRedirect
from django.templatetags.static import static
from django.utils.safestring import mark_safe
from django.urls import reverse
from django.utils.html import format_html, format_html_join
from django.utils.http import urlquote
from django.utils.safestring import mark_safe
from .models import Correspondent, Tag, Document, Log
from documents.actions import (
add_tag_to_selected,
remove_correspondent_from_selected,
remove_tag_from_selected,
set_correspondent_on_selected
)
class MonthListFilter(admin.SimpleListFilter):
title = "Month"
# Parameter for the filter that will be used in the URL query.
parameter_name = "month"
def lookups(self, request, model_admin):
r = []
for document in Document.objects.all():
r.append((
document.created.strftime("%Y-%m"),
document.created.strftime("%B %Y")
))
return sorted(set(r), key=lambda x: x[0], reverse=True)
def queryset(self, request, queryset):
if not self.value():
return None
year, month = self.value().split("-")
return queryset.filter(created__year=year, created__month=month)
from .models import Correspondent, Document, Log, Tag
class FinancialYearFilter(admin.SimpleListFilter):
@ -104,18 +87,59 @@ class FinancialYearFilter(admin.SimpleListFilter):
created__lte=self._fy_end(end))
class RecentCorrespondentFilter(admin.RelatedFieldListFilter):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.title = "correspondent (recent)"
def field_choices(self, field, request, model_admin):
years = settings.PAPERLESS_RECENT_CORRESPONDENT_YEARS
days = 365 * years
lookups = []
if years and years > 0:
correspondents = Correspondent.objects.filter(
documents__created__gte=datetime.now() - timedelta(days=days)
).distinct()
for c in correspondents:
lookups.append((c.id, c.name))
return lookups
class CommonAdmin(admin.ModelAdmin):
list_per_page = settings.PAPERLESS_LIST_PER_PAGE
class CorrespondentAdmin(CommonAdmin):
list_display = ("name", "match", "matching_algorithm", "document_count")
list_display = (
"name",
"match",
"matching_algorithm",
"document_count",
"last_correspondence"
)
list_filter = ("matching_algorithm",)
list_editable = ("match", "matching_algorithm")
def get_queryset(self, request):
qs = super(CorrespondentAdmin, self).get_queryset(request)
qs = qs.annotate(
document_count=models.Count("documents"),
last_correspondence=models.Max("documents__created")
)
return qs
def document_count(self, obj):
return obj.documents.count()
return obj.document_count
document_count.admin_order_field = "document_count"
def last_correspondence(self, obj):
return obj.last_correspondence
last_correspondence.admin_order_field = "last_correspondence"
class TagAdmin(CommonAdmin):
@ -125,8 +149,14 @@ class TagAdmin(CommonAdmin):
list_filter = ("colour", "matching_algorithm")
list_editable = ("colour", "match", "matching_algorithm")
def get_queryset(self, request):
qs = super(TagAdmin, self).get_queryset(request)
qs = qs.annotate(document_count=models.Count("documents"))
return qs
def document_count(self, obj):
return obj.documents.count()
return obj.document_count
document_count.admin_order_field = "document_count"
class DocumentAdmin(CommonAdmin):
@ -140,12 +170,30 @@ class DocumentAdmin(CommonAdmin):
readonly_fields = ("added",)
list_display = ("title", "created", "added", "thumbnail", "correspondent",
"tags_")
list_filter = ("tags", "correspondent", FinancialYearFilter,
MonthListFilter)
list_filter = (
"tags",
("correspondent", RecentCorrespondentFilter),
"correspondent",
FinancialYearFilter
)
filter_horizontal = ("tags",)
ordering = ["-created", "correspondent"]
actions = [
add_tag_to_selected,
remove_tag_from_selected,
set_correspondent_on_selected,
remove_correspondent_from_selected
]
date_hierarchy = "created"
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.document_queue = []
def has_add_permission(self, request):
return False
@ -153,6 +201,79 @@ class DocumentAdmin(CommonAdmin):
return obj.created.date().strftime("%Y-%m-%d")
created_.short_description = "Created"
def changelist_view(self, request, extra_context=None):
response = super().changelist_view(
request,
extra_context=extra_context
)
if request.method == "GET":
cl = self.get_changelist_instance(request)
self.document_queue = [doc.id for doc in cl.queryset]
return response
def change_view(self, request, object_id=None, form_url='',
extra_context=None):
extra_context = extra_context or {}
if self.document_queue and object_id:
if int(object_id) in self.document_queue:
# There is a queue of documents
current_index = self.document_queue.index(int(object_id))
if current_index < len(self.document_queue) - 1:
# ... and there are still documents in the queue
extra_context["next_object"] = self.document_queue[
current_index + 1
]
return super(DocumentAdmin, self).change_view(
request,
object_id,
form_url,
extra_context=extra_context,
)
def response_change(self, request, obj):
# This is mostly copied from ModelAdmin.response_change()
opts = self.model._meta
preserved_filters = self.get_preserved_filters(request)
msg_dict = {
"name": opts.verbose_name,
"obj": format_html(
'<a href="{}">{}</a>',
urlquote(request.path),
obj
),
}
if "_saveandeditnext" in request.POST:
msg = format_html(
'The {name} "{obj}" was changed successfully. '
'Editing next object.',
**msg_dict
)
self.message_user(request, msg, messages.SUCCESS)
redirect_url = reverse(
"admin:{}_{}_change".format(opts.app_label, opts.model_name),
args=(request.POST["_next_object"],),
current_app=self.admin_site.name
)
redirect_url = add_preserved_filters(
{
"preserved_filters": preserved_filters,
"opts": opts
},
redirect_url
)
return HttpResponseRedirect(redirect_url)
return super().response_change(request, obj)
@mark_safe
def thumbnail(self, obj):
return self._html_tag(
"a",
@ -165,8 +286,8 @@ class DocumentAdmin(CommonAdmin):
),
href=obj.download_url
)
thumbnail.allow_tags = True
@mark_safe
def tags_(self, obj):
r = ""
for tag in obj.tags.all():
@ -183,10 +304,11 @@ class DocumentAdmin(CommonAdmin):
)
}
)
return mark_safe(r)
tags_.allow_tags = True
return r
@mark_safe
def document(self, obj):
# TODO: is this method even used anymore?
return self._html_tag(
"a",
self._html_tag(
@ -199,7 +321,6 @@ class DocumentAdmin(CommonAdmin):
),
href=obj.download_url
)
document.allow_tags = True
@staticmethod
def _html_tag(kind, inside=None, **kwargs):

18
src/documents/templates/admin/documents/document/change_form.html

@ -1,5 +1,21 @@
{% extends 'admin/change_form.html' %}
{% block content %}
{{ block.super }}
{% if next_object %}
<script type="text/javascript">//<![CDATA[
(function($){
$('<input type="submit" value="Save and edit next" name="_saveandeditnext" />')
.prependTo('div.submit-row');
$('<input type="hidden" value="{{next_object}}" name="_next_object" />')
.prependTo('div.submit-row');
})(django.jQuery);
//]]></script>
{% endif %}
{% endblock content %}
{% block footer %}
@ -10,4 +26,4 @@
django.jQuery(".field-created input").first().attr("type", "date")
</script>
{% endblock footer %}
{% endblock footer %}

50
src/documents/templates/admin/documents/document/select_object.html Normal file

@ -0,0 +1,50 @@
{% extends "admin/base_site.html" %}
{% load i18n l10n admin_urls static %}
{% load staticfiles %}
{% block extrahead %}
{{ block.super }}
{{ media }}
<script type="text/javascript" src="{% static 'admin/js/cancel.js' %}"></script>
{% endblock %}
{% block bodyclass %}{{ block.super }} app-{{ opts.app_label }} model-{{ opts.model_name }} delete-confirmation delete-selected-confirmation{% endblock %}
{% block breadcrumbs %}
<div class="breadcrumbs">
<a href="{% url 'admin:index' %}">{% trans 'Home' %}</a>
&rsaquo; <a href="{% url 'admin:app_list' app_label=opts.app_label %}">{{ opts.app_config.verbose_name }}</a>
&rsaquo; <a href="{% url opts|admin_urlname:'changelist' %}">{{ opts.verbose_name_plural|capfirst }}</a>
&rsaquo; {{ title }}
</div>
{% endblock %}
{% block content %}
<p>Please select the {{itemname}}.</p>
<form method="post">{% csrf_token %}
<div>
{% for obj in queryset %}
<input type="hidden" name="{{ action_checkbox_name }}" value="{{ obj.pk|unlocalize }}"/>
{% endfor %}
<p>
<select name="obj_id">
{% for obj in objects %}
<option value="{{ obj.id }}">{{ obj.name }}</option>
{% endfor %}
</select>
</p>
<input type="hidden" name="action" value="{{ action }}"/>
<input type="hidden" name="post" value="yes" />
<p>
<input type="submit" value="{% trans 'Confirm' %}" />
<a href="#" class="button cancel-link">{% trans "Go back" %}</a>
</p>
</div>
</form>
{% endblock %}

12
src/paperless/settings.py

@ -22,12 +22,12 @@ elif os.path.exists("/usr/local/etc/paperless.conf"):
load_dotenv("/usr/local/etc/paperless.conf")
def __get_boolean(key):
def __get_boolean(key, default="NO"):
"""
Return a boolean value based on whatever the user has supplied in the
environment based on whether the value "looks like" it's True or not.
"""
return bool(os.getenv(key, "NO").lower() in ("yes", "y", "1", "t", "true"))
return bool(os.getenv(key, default).lower() in ("yes", "y", "1", "t", "true"))
# Build paths inside the project like this: os.path.join(BASE_DIR, ...)
@ -47,7 +47,7 @@ SECRET_KEY = os.getenv(
# SECURITY WARNING: don't run with debug turned on in production!
DEBUG = True
DEBUG = __get_boolean("PAPERLESS_DEBUG", "YES")
LOGIN_URL = "admin:login"
@ -292,3 +292,9 @@ FY_END = os.getenv("PAPERLESS_FINANCIAL_YEAR_END")
# Specify the default date order (for autodetected dates)
DATE_ORDER = os.getenv("PAPERLESS_DATE_ORDER", "DMY")
# Specify for how many years a correspondent is considered recent. Recent
# correspondents will be shown in a separate "Recent correspondents" filter as
# well. Set to 0 to disable this filter.
PAPERLESS_RECENT_CORRESPONDENT_YEARS = int(os.getenv(
"PAPERLESS_RECENT_CORRESPONDENT_YEARS", 0))

4
src/paperless_tesseract/parsers.py

@ -172,8 +172,8 @@ class RasterisedDocumentParser(DocumentParser):
raw_text = self._assemble_ocr_sections(imgs, middle, raw_text)
return raw_text
raise OCRError(
"The guessed language is not available in this instance of "
"Tesseract."
"The guessed language ({}) is not available in this instance "
"of Tesseract.".format(guessed_language)
)
def _ocr(self, imgs, lang):