use haystack for cazador

manolo/apps/cazador/forms.py

@@ -0,0 +1,19 @@
+from haystack.forms import HighlightedSearchForm
+
+
+class CazadorForm(HighlightedSearchForm):
+    def search(self):
+        sqs = super(CazadorForm, self).search()
+
+        if not self.is_valid():
+            return self.no_query_found()
+
+        if not self.cleaned_data.get('q'):
+            return self.no_query_found()
+
+        sqs = self.searchqueryset.auto_query(self.cleaned_data['q'])
+
+        if self.load_all:
+            sqs = sqs.load_all()
+
+        return sqs

manolo/apps/cazador/management/commands/import_data.py

@@ -0,0 +1,73 @@
+"""Add deudores al Estado to index"""
+from tqdm import tqdm
+
+from django.core.management.base import BaseCommand, CommandError
+
+from cazador.models import Cazador
+
+
+DBS = ['deudores', 'candidato_2014', 'narcoindultos', 'redam']
+
+
+class Command(BaseCommand):
+    def add_arguments(self, parser):
+        parser.add_argument('-i',
+                            action='store',
+                            dest='input_file',
+                            help='Enter filename with data to import.',
+                            )
+        parser.add_argument('-db',
+                            action='store',
+                            dest='database',
+                            help='Enter database, options: {}'.format(
+                            ', '.join(DBS)))
+
+    def handle(self, *args, **options):
+        if not options['input_file']:
+            raise CommandError("Input Filename does not exist.")
+        if options['database'] not in DBS:
+            raise CommandError("Name of database is incorrect. Use one of the "
+                               "following options: {}".format(
+                               ', '.join(DBS)))
+
+        input_file = options['input_file']
+        database = options['database']
+        self.process_file(input_file, database)
+
+    def process_file(self, input_file, database):
+        print(input_file, database)
+        with open(input_file, "r") as handle:
+            data = handle.readlines()
+
+        entries = []
+        for raw_line in tqdm(data):
+            line = raw_line.strip()
+            fields = line.split("\t")
+            if database == 'candidato_2014':
+                c = Cazador(
+                        raw_data=" ".join([
+                            fields[2],
+                            fields[3],
+                            fields[4],
+                            fields[1],
+                            fields[7],
+                        ]),
+                        source=database,
+                )
+            elif database == 'redam':
+                c = Cazador(
+                        raw_data=", ".join(fields),
+                        source=database,
+                )
+            elif database == 'narcoindultos':
+                c = Cazador(
+                        raw_data=" ".join(fields),
+                        source=database,
+                )
+            else:
+                c = Cazador(
+                        raw_data=" ".join([fields[2], fields[10], fields[1]]),
+                        source=database,
+                )
+            entries.append(c)
+        Cazador.objects.bulk_create(entries)

manolo/apps/cazador/migrations/0001_initial.py

@@ -0,0 +1,21 @@
+# -*- coding: utf-8 -*-
+from __future__ import unicode_literals
+
+from django.db import models, migrations
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='Cazador',
+            fields=[
+                ('id', models.AutoField(serialize=False, primary_key=True)),
+                ('raw_data', models.TextField()),
+                ('source', models.TextField()),
+            ],
+        ),
+    ]

manolo/apps/cazador/models.py

@@ -2,4 +2,8 @@
 
 from django.db import models
 
-# Create your models here.
+
+class Cazador(models.Model):
+    id = models.AutoField(primary_key=True)
+    raw_data = models.TextField()
+    source = models.TextField()

manolo/apps/cazador/search_indexes.py

@@ -0,0 +1,9 @@
+from haystack import indexes
+from .models import Cazador
+
+
+class CazadorIndex(indexes.SearchIndex, indexes.Indexable):
+    text = indexes.CharField(document=True, use_template=True)
+
+    def get_model(self):
+        return Cazador

manolo/apps/cazador/templates/cazador/results.html

@@ -24,8 +24,8 @@
                     <tbody>
                     {% for result in results %}
                     <tr>
-                        <td>{{ result.1 }}</td>
-                        <td>{{ result.3 }}</td>
+                        <td>{{ result.source }}</td>
+                        <td>{{ result.raw_data }}</td>
                     </tr>
                     {% endfor %}
                     </tbody>

manolo/apps/cazador/views.py

@@ -1,13 +1,19 @@
-from django.shortcuts import render_to_response, redirect
+from django.shortcuts import render_to_response
 
-from .utils import search
+from cazador.forms import CazadorForm
+from django.views.decorators.csrf import csrf_exempt
 
 
+@csrf_exempt
 def index(request):
-    if 'q' in request.GET:
-        name = request.GET['q'].strip()
-        results = search(name)
+    form = CazadorForm(request.GET)
+    query = request.GET['q']
 
-        return render_to_response('cazador/results.html', {'results': results})
+    all_items = form.search()
+    all_items = [i.object for i in all_items]
 
-    return render_to_response('cazador/index.html')
+    return render_to_response("cazador/results.html",
+                  {
+                      "results": all_items,
+                  }
+                  )

manolo/apps/visitors/models.py

@@ -1,6 +1,5 @@
 # -*- coding: utf-8 -*-
 from django.db import models
-import hashlib
 
 
 class Visitor(models.Model):

manolo/templates/search/indexes/cazador/cazador_text.txt

@@ -0,0 +1 @@
+{{ object.raw_data }}

requirements/base.txt

@@ -25,3 +25,4 @@ python-slugify==0.1.0
 six==1.9
 elasticsearch==1.4.0
 wheel==0.24.0
+tqdm