See example/films/README.txt The “name” field is guessed incorrectly (because the first film has name=“.45”, so indexing errors once it hits a name value that is no longer numeric. The README provides a command to define the name field *before* indexing. If you’ve indexed and had the name field guessed incorrectly and created, you’ll need to delete and recreate the collection, then define the name field, then reindex.
We used to have a fake film at the top to allow field guessing to “work”, but I felt that was too fake and that the example should be true to what happens with real world data and the pitfalls of allowing field type guessing to guess incorrectly. — Erik Hatcher, Senior Solutions Architect http://www.lucidworks.com > On Sep 2, 2015, at 5:17 AM, Long Yan <long....@robotron.de> wrote: > > Hey, > I have created a core with > bin\solr create -c mycore > > I want to index the csv sample files from solr-5.2.1 > > If I index film.csv under solr-5.2.1\example\films\, solr can only index this > file until the line > "2046,Wong Kar-wai,Romance Film|Fantasy|Science > Fiction|Drama,,/en/2046_2004,2004-05-20" > > But if I at first index books.csv under solr-5.2.1\example\exampledocs and > then index film.csv, solr can index all lines in film.csv > > Why? > > Regards > Long Yan > >
