Here the complete procedure for Linux users :

1/ import all your products

2/ export your products : this step is mandatory to get the external ID of your product. In my case I exported Name (that is mandatory) and ean13 since the picture's filename was name as follow :

ean13_code.jpg

3/ open your exported file (product.product.csv) with libreoffice and save it again with option "edit filter settings" in order to change the CSV separator to whatever else than ",". This is a little trick to avoid complicated script to deal with record that embed a coma.

4/ use the following script :

    #!/bin/bash
    # Usage : MakeBase64CSV.sh infile.csv outfile.csv
    # infile.csv columns are : externalID, name, filename or identifier
    # infile.csv separator MUST BE |


echo \"External ID\",\"Name\",\"image\" > $2

while IFS="|" read f1 f2 f3; do

    # recopy external ID and name
    echo -n $f1,\"$f2\", >> $2

    #If third column represents the picture's filename (not the key), please use this command
    #cat $(echo ${f3} | tr -d '\r' | tr -d '"') | base64 --wrap=0 >> $2

    #If third column represents the key to match with the filename, please use this command
    cat $(echo ${f3} | tr -d '\r' | tr -d '"').jpg | base64 --wrap=0 >> $2

    #Carrier return at end of line
    echo  >> $2
done < $1

5/ then use your script

./MakeBase64CSV.sh product.product.csv out.csv

In my example, all the pictures are in the same directory.

6/ you can now import your file out.csv into openERP. (Don't forget to change the CSV separator).

7/ If you run into the following error :

field larger than field limit (131072)

Please do the following :

• modify the file addons/base_import/models.py

• add the following line at the very begining of the _convert_import_data function :

  csv.field_size_limit(2097152)

• delete the file addons/base_import/models.pyc

• restart your server
• redo the import of your out.csv file