Inicia a pesquisa textual

.gitignore

@@ -92,3 +92,4 @@ collected_static
 bower
 bower_components
 media
+whoosh/

requirements/requirements.txt

@@ -13,19 +13,23 @@ django-extensions==1.6.7
 django-extra-views==0.8.0
 django-filter==0.15.3
 django-floppyforms==1.6.2
+django-haystack==2.5.1
 django-model-utils==2.5
 django-sass-processor==0.4.6
 djangorestframework
 drfdocs
 easy-thumbnails==2.3
 git+git://github.com/interlegis/trml2pdf.git
+haystack==0.36
 libsass==0.11.1
 psycopg2==2.6.2
 python-decouple==3.0
 pytz==2016.4
 pyyaml==3.11
 rtyaml==0.0.3
+textract==1.5.0
 unipath==1.1
 python-magic==0.4.12
 gunicorn==19.6.0
 django-reversion==2.0.8
+whoosh==2.7.4

sapl/base/search_indexes.py

@@ -0,0 +1,57 @@
+import textract
+
+from haystack import indexes
+from sapl.materia.models import DocumentoAcessorio, MateriaLegislativa
+from django.template import Context, loader
+
+
+class DocumentoAcessorioIndex(indexes.SearchIndex, indexes.Indexable):
+    text = indexes.CharField(document=True, use_template=True)
+
+    filename = 'arquivo'
+    model = DocumentoAcessorio
+    template_name = 'materia/documentoacessorio_text.txt'
+
+    def get_model(self):
+        return self.model
+
+    def index_queryset(self, using=None):
+        """Used when the entire index for model is updated."""
+        return self.get_model().objects.all()
+
+    def prepare(self, obj):
+        if not self.filename or not self.model or not self.template_name:
+            raise Exception
+
+        data = super(DocumentoAcessorioIndex, self).prepare(obj)
+
+        arquivo = getattr(obj, self.filename)
+
+        if arquivo:
+            try:
+                arquivo.open()
+            except OSError:
+                return self.prepared_data
+
+            extracted_data = textract.process(
+                arquivo.path).decode(
+                'utf-8').replace('\n', ' ')
+
+            # Now we'll finally perform the template processing to render the
+            # text field with *all* of our metadata visible for templating:
+            t = loader.select_template((
+                'search/indexes/' + self.template_name, ))
+            data['text'] = t.render(Context({'object': obj,
+                                             'extracted': extracted_data}))
+
+            return data
+
+        return self.prepared_data
+
+
+class MateriaLegislativaIndex(DocumentoAcessorioIndex):
+    text = indexes.CharField(document=True, use_template=True)
+
+    filename = 'texto_original'
+    model = MateriaLegislativa
+    template_name = 'materia/materialegislativa_text.txt'

sapl/base/templatetags/common_tags.py

@@ -2,6 +2,7 @@ from compressor.utils import get_class
 from django import template
 
 from sapl.base.models import AppConfig
+from sapl.materia.models import DocumentoAcessorio, MateriaLegislativa
 from sapl.parlamentares.models import Filiacao
 
 register = template.Library()
@@ -137,3 +138,13 @@ def cronometro_to_seconds(value):
 @register.filter
 def to_list_pk(object_list):
     return [o.pk for o in object_list]
+
+
+@register.filter
+def search_get_model(object):
+    if type(object) == MateriaLegislativa:
+        return 'm'
+    elif type(object) == DocumentoAcessorio:
+        return 'd'
+
+    return None

sapl/base/urls.py

@@ -99,4 +99,6 @@ urlpatterns = [
         name='login'),
     url(r'^logout/$', views.logout, {'next_page': '/login'}, name='logout'),
 
+    url(r'^search/', include('haystack.urls')),
+
 ] + recuperar_senha

sapl/settings.py

@@ -76,12 +76,23 @@ INSTALLED_APPS = (
     'crispy_forms',
     'easy_thumbnails',
     'floppyforms',
+    'haystack',
     'sass_processor',
     'rest_framework',
     'reversion',
+    'whoosh',
 
 ) + SAPL_APPS
 
+
+HAYSTACK_CONNECTIONS = {
+    'default': {
+        'ENGINE': 'haystack.backends.whoosh_backend.WhooshEngine',
+        'PATH': PROJECT_DIR.child('whoosh'),
+    },
+}
+
+
 if DEBUG:
     INSTALLED_APPS += ('debug_toolbar', 'rest_framework_docs',)
 

sapl/templates/search/indexes/materia/documentoacessorio_text.txt

@@ -0,0 +1,7 @@
+{% for k, v in extracted.metadata.items %}
+    {% for val in v %}
+        {{ k }}: {{ val|safe }}
+    {% endfor %}
+{% endfor %}
+
+{{ extracted|striptags|safe }}

sapl/templates/search/indexes/materia/materialegislativa_text.txt

@@ -0,0 +1,7 @@
+{% for k, v in extracted.metadata.items %}
+    {% for val in v %}
+        {{ k }}: {{ val|safe }}
+    {% endfor %}
+{% endfor %}
+
+{{ extracted|striptags|safe }}

sapl/templates/search/search.html

@@ -0,0 +1,70 @@
+{% extends 'crud/form.html' %}
+{% load crispy_forms_tags %}
+{% load common_tags %}
+
+{% block base_content %}
+    <h1><legend>Pesquisa Textual</legend></h1>
+    </br>
+
+    <form method="get" action=".">
+        <div class="row container-detail clearfix">
+
+            <div  class="row-fluid">
+                <div class="col-md-8">
+                    {{ form.q|as_crispy_field }}
+                </div>
+            </div>
+
+            <div  class="row-fluid">
+                <div class="col-md-8">
+                    </br>
+                    <h4> Em quais tipos de documento deseja pesquisar?</h4>
+                </div>
+            </div>
+
+            <div  class="row-fluid">
+                <div class="col-md-8">
+                    {{ form.models }}
+                </div>
+            </div>
+
+
+            <div  class="row-fluid">
+                <div class="col-md-12">
+                    <input class="btn btn-primary pull-right" type="submit" value="Pesquisar">
+                </div>
+            </div>
+
+        </div>
+
+        {% if query %}
+            <h3>Resultados</h3>
+
+            {% for result in page.object_list %}
+                {% if result.object|search_get_model == 'm' %}
+                    <p>
+                      <a href="{% url 'sapl.materia:materialegislativa_detail' result.object.pk %}">{{ result.object }}</a>
+                    </p>
+                
+                {% elif result.object|search_get_model == 'd' %}
+                    <p>
+                      <a href="{% url 'sapl.materia:documentoacessorio_detail' result.object.pk %}">{{ result.object }}</a>
+                    </p>
+                {% endif %}
+
+            {% empty %}
+                <p>No results found.</p>
+            {% endfor %}
+
+            {% if page.has_previous or page.has_next %}
+                <div>
+                    {% if page.has_previous %}<a href="?q={{ query }}&amp;page={{ page.previous_page_number }}">{% endif %}&laquo; Previous{% if page.has_previous %}</a>{% endif %}
+                    |
+                    {% if page.has_next %}<a href="?q={{ query }}&amp;page={{ page.next_page_number }}">{% endif %}Next &raquo;{% if page.has_next %}</a>{% endif %}
+                </div>
+            {% endif %}
+        {% else %}
+            {# Show some example queries to run, maybe query syntax, something else? #}
+        {% endif %}
+    </form>
+{% endblock %}