Fulltext Search in arangodb using AQL and python

Question

i have stored the data in arangodb in the following format:

{"data": [
{
  "content": "maindb",
  "type": "string",
  "name": "db_name",
  "key": "1745085839"
},
{
  "type": "id",
  "name": "rel",
  "content": "1745085840",
  "key": "1745085839"
},
{
  "content": "user",
  "type": "string",
  "name": "rel_name",
  "key": "1745085840"
},
{
  "type": "id",
  "name": "tuple",
  "content": "174508584001",
  "key": "1745085840"
},
{
  "type": "id",
  "name": "tuple",
  "content": "174508584002",
  "key": "1745085840"
},
{
  "type": "id",
  "name": "tuple",
  "content": "174508584003",
  "key": "1745085840"
},
{
  "type": "id",
  "name": "tuple",
  "content": "174508584004",
  "key": "1745085840"
},
{
  "type": "id",
  "name": "tuple",
  "content": "174508584005",
  "key": "1745085840"
},
{
  "type": "id",
  "name": "tuple",
  "content": "174508584006",
  "key": "1745085840"
},
{
  "type": "id",
  "name": "tuple",
  "content": "174508584007",
  "key": "1745085840"
},
{
  "content": "dspclient",
  "type": "varchar",
  "name": "username",
  "key": "174508584001"
},
{
  "content": "12345",
  "type": "varchar",
  "name": "password",
  "key": "174508584001"
},
{
  "content": "12345",
  "type": "varchar",
  "name": "cpassword",
  "key": "174508584001"
},
{
  "content": "n",
  "type": "varchar",
  "name": "PostgreSQL",
  "key": "174508584001"
},
{
  "content": "n",
  "name": "IBMDB2",
  "type": "varchar",
  "key": "174508584001"
},
{
  "content": "n",
  "name": "MySQL",
  "type": "varchar",
  "key": "174508584001"
},
{
  "content": "n",
  "type": "varchar",
  "name": "SQLServer",
  "key": "174508584001"
},
{
  "content": "n",
  "name": "Hadoop",
  "type": "varchar",
  "key": "174508584001"
},
{
  "content": "None",
  "name": "dir1",
  "type": "varchar",
  "key": "174508584001"
},
{
  "content": "None",
  "name": "dir2",
  "type": "varchar",
  "key": "174508584001"
},
{
  "content": "None",
  "name": "dir3",
  "type": "varchar",
  "key": "174508584001"
},
{
  "content": "None",
  "name": "dir4",
  "type": "varchar",
  "key": "174508584001"
},
{
  "type": "inet",
  "name": "ipaddr",
  "content": "1921680103",
  "key": "174508584001"
},
{
  "content": "y",
  "name": "status",
  "type": "varchar",
  "key": "174508584001"
},
{
  "content": "None",
  "type": "varchar",
  "name": "logintime",
  "key": "174508584001"
},
{
  "content": "None",
  "type": "varchar",
  "name": "logindate",
  "key": "174508584001"
},
{
  "content": "None",
  "type": "varchar",
  "name": "logouttime",
  "key": "174508584001"
},
{
  "content": "client",
  "type": "varchar",
  "name": "user_type",
  "key": "174508584001"
},
{
  "content": "royal",
  "type": "varchar",
  "name": "username",
  "key": "174508584002"
},
{
  "content": "12345",
  "type": "varchar",
  "name": "password",
  "key": "174508584002"
},
{
  "content": "12345",
  "type": "varchar",
  "name": "cpassword",
  "key": "174508584002"
},
{
  "content": "n",
  "type": "varchar",
  "name": "PostgreSQL",
  "key": "174508584002"
},
{
  "content": "n",
  "name": "IBMDB2",
  "type": "varchar",
  "key": "174508584002"
},
{
  "content": "n",
  "name": "MySQL",
  "type": "varchar",
  "key": "174508584002"
},
{
  "content": "n",
  "type": "varchar",
  "name": "SQLServer",
  "key": "174508584002"
},
{
  "content": "n",
  "name": "Hadoop",
  "type": "varchar",
  "key": "174508584002"
},
{
  "content": "None",
  "name": "dir1",
  "type": "varchar",
  "key": "174508584002"
},
{
  "content": "None",
  "name": "dir2",
  "type": "varchar",
  "key": "174508584002"
},
{
  "content": "None",
  "name": "dir3",
  "type": "varchar",
  "key": "174508584002"
},
{
  "content": "None",
  "name": "dir4",
  "type": "varchar",
  "key": "174508584002"
},
{
  "type": "inet",
  "name": "ipaddr",
  "content": "1921680105",
  "key": "174508584002"
},
{
  "content": "y",
  "name": "status",
  "type": "varchar",
  "key": "174508584002"
},
{
  "content": "190835899000",
  "type": "varchar",
  "name": "logintime",
  "key": "174508584002"
},
{
  "content": "20151002",
  "type": "varchar",
  "name": "logindate",
  "key": "174508584002"
},
{
  "content": "None",
  "type": "varchar",
  "name": "logouttime",
  "key": "174508584002"
},
{
  "content": "client",
  "type": "varchar",
  "name": "user_type",
  "key": "174508584002"
},
{
  "content": "abc",
  "type": "varchar",
  "name": "username",
  "key": "174508584003"
},
{
  "content": "12345",
  "type": "varchar",
  "name": "password",
  "key": "174508584003"
},
{
  "content": "12345",
  "type": "varchar",
  "name": "cpassword",
  "key": "174508584003"
},
{
  "content": "n",
  "type": "varchar",
  "name": "PostgreSQL",
  "key": "174508584003"
},
{
  "content": "n",
  "name": "IBMDB2",
  "type": "varchar",
  "key": "174508584003"
}]}

In order to perform fulltext search, I have created an index on content attribute by using the syntax from a python script:

c.DSP.ensureFulltextIndex("content");

Where, c is database, and DSP is the collection name. Now, I am trying to perform a search operation in the above data set by using the syntax:

FOR doc IN FULLTEXT(DSP, "content", "username") RETURN doc

Then, an error occure:

[1571] in function 'FULLTEXT()': no suitable fulltext index found for fulltext query on 'DSP' (while executing)

Please tell me the problem, and also tell me what will be the syntax when i will try this query with a python script.

Thanks...

did you have a look at the management interface? Collections->yourCollection[ (I) ]->Indexes - You should find your index in that list. — dothebart, Nov 10 '15 at 12:54
Thanks dothebert, Now i have created the index through arangodb interface...But the above query is returning an empty list [] as result.... — Mrityunjay Singh, Nov 10 '15 at 18:05
If my eyeball grep works properly, you don't have any `content` attribute containing the string `username`? See [the fulltext index example](https://docs.arangodb.com/IndexHandling/Fulltext.html) — dothebart, Nov 13 '15 at 14:09

score 1 · Answer 1 · answered Nov 23 '15 at 15:18

Working with the 10 minutes tutorial and the driver documentation

I got it working like this:

from pyArango.connection import *
c = Connection()
db = c.createDatabase(name = "testdb")
DSP= db.createCollection(name = "DSP")

DSP.ensureFulltextIndex(fields=["content"])

doc = DSP.createDocument({"content": "test bla"})
doc.save()

print db.AQLQuery('''FOR doc IN FULLTEXT(DSP, "content", "bla") RETURN doc''', 10)

Resulting in:

[{u'_key': u'1241175138503', u'content': u'test bla', u'_rev': u'1241175138503', u'_id': u'DSP/1241175138503'}]

I've used arangosh to revalidate the steps from the python prompt:

arangosh> db._useDatabase("testdb")
arangosh [testdb]> db.DSP.getIndexes()
[ 
  { 
    "id" : "DSP/0", 
    "type" : "primary", 
    "fields" : [ 
      "_key" 
    ], 
    "selectivityEstimate" : 1, 
    "unique" : true, 
    "sparse" : false 
  }, 
  { 
    "id" : "DSP/1241140928711", 
    "type" : "hash", 
    "fields" : [ 
      "content" 
    ], 
    "selectivityEstimate" : 1, 
    "unique" : false, 
    "sparse" : true 
  }, 
  { 
    "id" : "DSP/1241142960327", 
    "type" : "fulltext", 
    "fields" : [ 
      "content" 
    ], 
    "unique" : false, 
    "sparse" : true, 
    "minLength" : 2 
  } 
]
arangosh [testdb]> db.testdb.toArray()
[ 
  { 
    "content" : "test bla", 
    "_id" : "DSP/1241175138503", 
    "_rev" : "1241175138503", 
    "_key" : "1241175138503" 
  } 
]
db._query('FOR doc IN FULLTEXT(DSP, "content", "bla") RETURN doc')

I've been trying this using python 2.7.9 on a debian jessie; /usr/local/lib/python2.7/dist-packages/pyArango-1.0.3-py2.7.egg; ArangoDB 2.6.7 and ArangoDB 2.8 beta, with both I get something like that from the python script: `python /tmp/test.py [{u'_key': u'22654861461', u'content': u'test bla', u'_rev': u'22654861461', u'_id': u'DSP/22654861461'}]` — dothebart, Jan 04 '16 at 08:45

Fulltext Search in arangodb using AQL and python

1 Answers1