GitHub - jtbraun/nutrimatic: The source behind nutrimatic.org.

jtbraun / nutrimatic Public

forked from PuzzleTechHub/nutrimatic

Notifications You must be signed in to change notification settings
Fork 0
Star 0

The source behind nutrimatic.org.

View license

0 stars 12 forks Branches Tags Activity

Star

Notifications

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 27 Commits
openfst-1.7.0		openfst-1.7.0
.gitignore		.gitignore
COPYING		COPYING
README		README
build.py		build.py
cgi-search.py		cgi-search.py
cgi-slack.py		cgi-slack.py
dump-index.cpp		dump-index.cpp
explore-index.cpp		explore-index.cpp
expr-anagram.cpp		expr-anagram.cpp
expr-filter.cpp		expr-filter.cpp
expr-intersect.cpp		expr-intersect.cpp
expr-optimize.cpp		expr-optimize.cpp
expr-parse.cpp		expr-parse.cpp
expr.h		expr.h
find-anagrams.cpp		find-anagrams.cpp
find-expr.cpp		find-expr.cpp
find-phone-words.cpp		find-phone-words.cpp
index-reader.cpp		index-reader.cpp
index-walker.cpp		index-walker.cpp
index-writer.cpp		index-writer.cpp
index.h		index.h
make-index.cpp		make-index.cpp
memoize.py		memoize.py
merge-indexes.cpp		merge-indexes.cpp
remove-markup.cpp		remove-markup.cpp
search-driver.cpp		search-driver.cpp
search-printer.cpp		search-printer.cpp
search.h		search.h
test-expr.cpp		test-expr.cpp

Repository files navigation

This is Nutrimatic (http://nutrimatic.org/usage.html).

To build the source, run "./build.py".  You will need the following installed:
   * Python
   * g++
   * libxml2 (ubuntu: apt-get install libxml2-dev; osx: pip install lxml)
   * libtre (ubuntu: apt-get install libtre-dev; osx: brew install tre)

To do anything useful, you will need to build an index from Wikipedia.

1. Download the latest Wikipedia database dump (this is a ~13GB file!):

     wget https://dumps.wikimedia.org/enwiki/latest/enwiki-latest-pages-articles.xml.bz2

2. Extract the text from the articles using Wikipedia Extractor
   (this generates ~12GB of text, and can take several hours!):

     # See http://medialab.di.unipi.it/wiki/Wikipedia_Extractor
     wget https://raw.githubusercontent.com/attardi/wikiextractor/master/WikiExtractor.py
     python WikiExtractor.py enwiki-latest-pages-articles.xml.bz2

   This will write many files named text/??/wiki_??.

3. Index the text (this generates ~50GB of data, and can also take hours!):

     find text -type f | xargs cat | bin/make-index wikipedia

   This will write many files named wikipedia.?????.index.
   (You can break this up; run make-index several times with different
   sets of input files, replacing "wikipedia" with unique names each time.)

4. Merge the indexes; I normally do this in two stages:

     for x in 0 1 2 3 4 5 6 7 8 9
     do bin/merge-indexes 2 wikipedia.????$x.index wiki-merged.$x.index
     done

     bin/merge-indexes 5 wiki-merged.*.index wiki-merged.index

   There's nothing magical about this appproach with 10 batches, you can use
   any way you like to merge the files. The 2 and 5 numbers are minimum phrase
   frequency cutoffs (how many times a string must occur to be included).

5. Enjoy your new index:

     bin/find-expr wiki-merged.index '<aciimnrttu>'

If you want to set up the web interface, write a short shell wrapper that runs
cgi-search.py with arguments pointing it at your binaries and data files, e.g.:

     #!/bin/sh

     export NUTRIMATIC_FIND_EXPR=/path/to/nutrimatic/bin/find-expr
     export NUTRIMATIC_INDEX=/path/to/nutrimatic/data/wiki-merged.index
     exec /path/to/nutrimatic/cgi-search.py

Then arrange for your web server to invoke that shell wrapper as a CGI script.

Have fun,

-- [email protected]