Noodle currently supports the Full-Text Search features of Oracle and MSSQL.

Additionally 3rd party filters in the form of CLI tools are supported for Postgres or any RDBMS.

An extensible full-text search is available for Postgres.

MSSQL full text search

1) Install the 3MB MSOffice IFilters not included in the 3GB MSSQL (instructions)

Other formats can be supported, like PDF, ODF, etc.

2) You can verify file type support with this:

select * from sys.fulltext_document_types order by document_type;

3) Optionally if you have already uploaded files to noodle you should right click 'rebuild' the 'Databases/noodledb/Storage/Full Text Catalogs/BINARY_DATA' table.

Keep in mind you have to use the advanced search in Noodle and select the Full Text option to use the MSSQL FTS.

Related: configure MSSQL and migrate

The noodle.properties and server.xml text files are in the conf directory.

If there are problems with the conf the logs will tell you about them.

Changes to the conf file will take effect the next time Noodle starts.

Full-text search is extensible to any format, so long as there is a CLI tool to convert it to text.

Configuration of ORACLE and MSSQL built-in indexers differ and can be used in combination with this method.

To index the contents of popular formats like office documents on linux add the following to your .conf:

#tested
#apt/yum install -y poppler-utils pstotext antiword html2text unrtf python-excelerator libwpd-tools unzip catdoc
Indexer.IndexContent = ALL
Indexer.toTxt.Enabled = true
Indexer.toTxt.pdf = "pdftotext -q -eol unix -enc UTF-8 $IN $OUT"
Indexer.toTxt.doc = "antiword $IN > $OUT"
Indexer.toTxt.html = "html2text -nobs -o $OUT $IN"
Indexer.toTxt.xls = "xls2csv $IN > $OUT"
Indexer.toTxt.mp3 = "id3info $IN | grep '===' | grep -v 'PRIV' | grep -v 'image\/' | perl -p -e 's/^.+(\)|\])\:/ /g' > $OUT"
Indexer.toTxt.rtf = "unrtf --nopict --text $IN 2>/dev/null | grep -v '^### ' > $OUT"
Indexer.toTxt.docx = "unzip -p $IN word/document.xml | perl -p -e 's/<.+?>/ /g' > $OUT"
Indexer.toTxt.pptx = "unzip -p $IN ppt/slides/*.xml | perl -p -e 's/<.+?>/ /g' > $OUT"
Indexer.toTxt.xlsx = "unzip -p $IN xl/sharedStrings.xml | perl -p -e 's/<.+?>/ /g' > $OUT"
Indexer.toTxt.odt = "unzip -p $IN content.xml | perl -p -e 's/<.+?>/ /g' > $OUT"
Indexer.toTxt.ods = "unzip -p $IN content.xml | perl -p -e 's/<.+?>/ /g' > $OUT"
Indexer.toTxt.odp = "unzip -p $IN content.xml | perl -p -e 's/<.+?>/ /g' > $OUT"

# TensorFlow image search
#Indexer.toTxt.jpg = "python classify.py --image $IN | grep -P '^1\. ' > $OUT"

## others
#Indexer.toTxt.wpd = "wpd2text $IN > $OUT"
#Indexer.toTxt.jpg = "exiftool $IN > $OUT #for camera type or gps location"
#Indexer.toTxt.xls = "py_xlstoTxt $IN > $OUT # supports sheets but adds sheet = ----"
#Indexer.toTxt.docx = "unzip -p $IN word/document.xml | sed -e 's/<[^>]\{1,\}>/ /g' > $OUT"

Windows examples

• Xpdf tools
• Antiword
• strawberryperl

Indexer.IndexContent = ALL
Indexer.Interval = 30
Indexer.toTxt.Enabled = true
Indexer.toTxt.pdf = "\"C:/Program Files/xpdf/pdftotext.exe\" $IN $OUT"
Indexer.toTxt.doc = "C:\antiword\antiword.exe $IN > $OUT"

You can populate your conf file with command line method of converting the file types of your choice to text.

Be sure "System Tools > Settings > General > Enable Full Text Search" is set to "Yes".

Make sure you adhere to formatting restrictions.

1. Add or un-comment the following line in the noodle.properties file:

ScriptFile = “noodleScript.js”

2. Add custom JavaScript to noodleScript.js

3. Restart Noodle.

To test put this in the noodleScript.js:

alert("It works");

Save it and re-login to Noodle.

Other ways to insert custom JavaScript here.

You may want to wait for top.noodle like this.