csv2ldif2.pl README
-------------------
This is a brief description of csv2ldif2.pl, a tool written in perl
that you can use to convert CSV files to LDIF files that you can
load into a directory server (like OpenLDAP).
Please also have a look to other related tools hosted on sf.net:
ldap-preg-replace: Change entries in ldap online with regexp
ldap-searchEntries: Mass check existence / enrich of entries based on csv
ldap-collate: Group and count entries by attributes
ldap-csvexport: Easily export LDAP to abitary csv-formats
ldif-preg-replace: Convert and modify LDIF files
csv2ldif2: Convert arbitary CSV files to LDIF
ldif-extract: grep-like filter for entries in LDIF
TODOs:
- none at the moment :)
I. Prerequisites and installing
Installation is not neccessary. Just make the file executable
if it is not already, or run the script through `perl`.
However, before you can run this program, you need:
* PERL installed (perl.org, on linux probably already installed)
* PERL modules 'Encode', 'MIME::Base64' and 'Getopt::Std'; all should be core
modules and already be present in your installation.
II. General
The tool supports many command line options that allow you to fine-tune
the generated LDIF file.
All command line options are shown if you call csv2ldif2.pl without
parameters.
Additional help and some hints is available by calling `csv2ldif2.pl -h`.
Simply spoken, the tool reads CSV data and parses it to LDIF data.
The first line is threaten as header line describing the attributes.
Each records data is mapped to the corresponding attribute.
If you want to write more than one csv-fields data into one
single attribute (multivalue) then you simply need to use the same attribute
name for the affected columns:
--- CSV DATA ---
1: cn,attr1,attr_mv,attr_mv
2: foo,test,mv1,mv2
Calling `csv2ldif2.pl -b 'o=example,dc=com'` will parse this into:
--- LDIF DATA ---
1: dn: cn=foo,o=example,dc=com
2: attr1: test
3: attr_mv: mv1
4: attr_mv: mv2
One of csv2ldif2.pl's features is consolidation of the csv data.
If you have several records for the same entry this is considered.
Only one LDIF entry will be generated for the same dn regardless how often
it will appear in the csv data. As shown below, we have two records
for "foo" (line 2 and 4):
--- CSV DATA ---
1: cn,attr1,attr_mv,attr_mv
2: foo,test,mv1,mv2
3: bar,test2,some_mv1,some_other_mv
4: foo,test,mv3,mv4
This will generate:
--- LDIF DATA ---
1: dn: cn=foo,o=example,dc=com
2: attr1: test
3: attr_mv: mv1
4: attr_mv: mv2
5: attr_mv: mv3
6: attr_mv: mv4
7:
8: dn: cn=bar,o=example,dc=com
9: attr1: test2
10: attr_mv: some_mv1
11: attr_mv: some_other_mv
III. Running and examples
The tool is written to read from STDIN and write to STDOUT (errors go
to STDERR) so you can pipe and redirect if you wish. This gives you
the flexibility to invoke the LDIF generation inside a chain
of different tasks like:
`cat data.csv | head -n 100 | grep -v "foo" | csv2ldif2.pl -b "ou=example,dc=com"`
(convert first 100 lines of data.csv, but ignore records containing "foo".
Result is printed to standard output for human validation first)
Or another example:
`cat data.csv | head -n 100 | grep -v "foo"
| csv2ldif2.pl -b "ou=example,dc=com" > data.ldif`
(same as above, but now write LDIF data to file "data.ldif")
Of course you can use it standalone too:
`csv2ldif2.pl -b "ou=example,dc=com" < data.csv > data.ldif`
Have fun!
