Retrieve input name or input path

[ghost]

I'm searching through 2 txt files for words. Is there any way to retrieve as result except the row ,with the word in, of the file and its name?
Currently i am running the project like this :

 hadoop jar solr-hadoop-job-2.2.5.jar 
    com.lucidworks.hadoop.ingest.IngestJob  
    -Dlww.commit.on.close=true -DcsvDelimiter= 
   -cls com.lucidworks.hadoop.ingest.CSVIngestMapper -c spyros1  
    - i  /usr/local/hadoop/input 
    -of com.lucidworks.hadoop.io.LWMapRedOutputFormat 
    -s http://127.0.1.1:8983/solr

Is somehow possible make it?

[ctargett]

I'm not sure I understand what you would like to see. Are you saying you just want to see the row number and file name?

[ghost]

@ctargett Yes i want to see also the file's name that the results came from.

[acesar]

I'm searching through 2 txt files for words.

you are ingesting the 2 txt files and then doing a query Solr to do the search, correct?

are you searching for a specify word?

You can try the RegexIngestMapper.

• The com.lucidworks.hadoop.ingest.RegexIngestMapper.regex argument is a regex, in the example "\\w+", will add every word as a field.

   hadoop jar solr-hadoop-job-2.2.5.jar com.lucidworks.hadoop.ingest.IngestJob  
    -Dlww.commit.on.close=true 
    -Dcom.lucidworks.hadoop.ingest.RegexIngestMapper.regex="\\w+" 
    -Dcom.lucidworks.hadoop.ingest.RegexIngestMapper.groups_to_fields=0=match_s  
    -cls com.lucidworks.hadoop.ingest.RegexIngestMapper  
    -c collection1 -i /path/* -zk 127.0.0.1:2181/solr 
    -of com.lucidworks.hadoop.io.LWMapRedOutputFormat 

Result the Solr document:

id=/path/file-name.txt-<ROW>, match_s=[wordA, wordB, wordC], path=/path/file-name.txt

[ghost]

@acesar When i try your command with the zk i am taking this message :

Solr server not available on: http://127.0.1.1:9983/solr
Make sure that collection [spyros44] exists

when i use -s http://127.0.1.1:8983/solr everything runs fine, but no data is sent to Solr...
I create the collections with this: bin/solr -e cloud
Any ideas?

[acesar]

@SpyrosAv -zk is for the zookeeper String, your port is 9983.

everything runs fine, but no data is sent to Solr...

Can you share the output of the job?
do you have a field name path in your Solr Schema?

[ghost]

@acesar I have the field path at my schema
Whit this :
-zk http://127.0.1.1:9983/solr

whit this :
-s http://127.0.1.1:8983/solr

[acesar]

@SpyrosAv

-zk http://127.0.1.1:9983/solr

The zk string should be -zk 127.0.1.1:9983 but if you only have one Solr node. It should be the same as the -s option.

The output seems to be correct.

it has a DOCS_ADDED value of 12761, you should have those docs in Solr.

Go to http://127.0.1.1:8983/solr and look at the collection.

[ghost]

@acesar In this case i have 2 nodes. But i have tried with 1 also.
Looking for "book" word

no data retrieved

[acesar]

@SpyrosAv Can you try again with the field match_ss?

 hadoop jar solr-hadoop-job-2.2.5.jar com.lucidworks.hadoop.ingest.IngestJob  
    -Dlww.commit.on.close=true 
    -Dcom.lucidworks.hadoop.ingest.RegexIngestMapper.regex="\\w+" 
   -Dcom.lucidworks.hadoop.ingest.RegexIngestMapper.groups_to_fields=0=match_ss  
   -cls com.lucidworks.hadoop.ingest.RegexIngestMapper  
   -c collection1 -i /path/* -zk 127.0.0.1: 9983
   -of com.lucidworks.hadoop.io.LWMapRedOutputFormat 

Can you first search with the *:* value to see how many docs you have.

The next step should be to search with match_ss:book

[ghost]

@acesar match_ss worked!!!
Thanks for the help and the your time!!!

[acesar]

Sure np!

[ghost]

@acesar If you could provide some information why match_ss worked and match_s didn't, i would appreciate it.

[acesar]

@SpyrosAv TheRegexIngestMapper add the matches from the regex to an Array of values (multivalue in Solr). A Dynamic multivalue in Solr has the suffix _ss by default.

[ghost]

@acesar Which Ingest Mapper should i use if i am using .pdf or .doc files? Is GrokIngestMapper the right one ?
Also is there any documentation about the project that will help me figure out queries like this?
Thanks in advance.
(I am not opening a new issue but if you feel is needed just tell me...)