I'm assuming that whoever wrote these Perl Judy scripts in the first place has now left your company, and without providing a comprehensive handover. Am I close?

You've asked two questions now on Judy arrays and neither has convinced me you have the technical chops required to take over this project. Maybe it's time to step back and look at the bigger picture. Do you really really need the performance boost that Judy arrays provide? Is this required performance boost clearly documented? Are benchmarks provided demonstrating why? Is there a strong business case? Or was it just a fun project to play with by a clever former employee?

If you don't really need the performance boost of Judy arrays, you might consider converting the code to simply use Perl hashes instead because that would make long term code maintenance easier for your company as employees come and go over time.

Having said that, the Judy array code looks like it was written by an expert (Doug Baskins) and appears to be very stable ... still, you may struggle to get quick support if you hit a bug.

BTW, one quote from the infamous BrowserUk scared me:

A supersearch for my name and judy arrays will turn up a compact, single file version of the Judy array code that compiled clean and worked very well for that application. Still slower than hashes, but far more compact. Just hope you don't find any bugs, because the Judy code is horribly complex.

Having got that off my chest, I find high performance fun, and I've never used Judy arrays, so I've put together some basic references below that I might play with later when I get more time.

• Judy array (wikipedia)

BrowserUk on Judy arrays:

• Bidirectional lookup algorithm? (Updated: further info.) by BrowserUk (search for Judy in this very long thread)
• [OT] The statistics of hashing. by BrowserUk
• Heap structure for lookup? by BrowserUk
• Re^6: Heap structure for lookup? by BrowserUk

Judy arrays on PM:

• Re: Challenge: Ricochet Robots (more edits) by vr (Feb 2021)
• need help with judy array searching by expo1967 (Dec 2020, question around using Judy arrays in Perl)
• need help with Tie::Judy by expo1967 (Dec 2020, earlier question around using Judy arrays in Perl)
• Re: Memory efficient way to deal with really large arrays? by LanX (2020)
• Re: Improving speed match arrays with fuzzy logic by Your Mother (2019)
• Re: What is the most memory efficient way to (sort) and print a hash? (Judy::SL) by tye (2014)
• Re: Bidirectional lookup algorithm? (judy trie?) by anonymonk (2015)
• Bit-vector compression by baxy77bax (Mar 2022)

• Unsigned 64-bit integer as Judy key by anonymonk (Dec 2022)
• Re^3: Unsigned 64-bit integer as Judy key by anonymonk (Dec 2022)
• Re^2: Rosetta Code: Long List is Long by anonymonk (Dec 2022) - using Judy Arrays with Rosetta Code: Long List is Long
• Re^2: Rosetta Code: Long List is Long - JudySL summary by marioroy (Jan 2023) - "Most amazingly is the Judy array consuming the least amount of memory"

• Size of Judy::HS array: where is MemUsed()? by kcott (Apr 2023)
• Re^5: Size of Judy::HS array: where is MemUsed()? - perldelta, Perl Releases and Building Perl by me - perl 5.38 (due for release June 30) should improve memory usage of large Perl hashes

Judy arrays on CPAN:

• Judy by jjore (Josh ben Jore)
• Judy::HS - efficient string to integer map (part of Judy)
• Judy::SL - efficient null-terminated string to integer map (part of Judy)

• Alien::Judy by jjore. Helps installing the Judy C library from sourceforge (old link, this one seems more recent: sourceforge judy by Douglas L Baskins)
• Tie::Judy by bholzman (Benjamin Holzman)

Building the Judy C Library on Linux

Notes below based on Re: Rosetta Code: Long List is Long - JudySL code by marioroy.

Introduction to Judy C Library: Judy Arrays Web Page

Download the Judy library (Judy-1.0.5.tar.gz) from: sourceforge Judy

Example build of the Judy C Library on Ubuntu Linux:

cd $HOME/local-judy
tar -xzf Judy-1.0.5.tar.gz
cd judy-1.0.5
./configure --enable-64-bit
make
sudo make install
[download]

Example compile and run of a C++ program on Linux that uses the Judy library (taken from my-rosetta-code-notes):

clang++ -o llil4judy -std=c++20 -fopenmp -Wall -O3 llil4judy.cpp -I "$
+HOME/local-fast_io/fast_io/include" -I "$HOME/local-boost/boost_1_81_
+0" -I /usr/local/include -L /usr/local/lib -l Judy

LD_LIBRARY_PATH=/usr/local/lib ./llil4judy big1.txt big2.txt big3.txt 
+>f.tmp
[download]