v2.2.1

• 456b8b1 fix: update findlike requirement to 1.4.2

There was a bug in the code where the reference document or query was not being preprocessed when the BM25 algorithm was selected.

v2.2.0

7b7643c feat: add ignore front-matter option

This option can be useful if, depending on the scenario, the Org front-matter (i.e. properties drawer) is interfering with the results. Smaller notes are more susceptible to this. Because they bear less content, files with the same #+FILETAGS but radically different subjects can still be seen as quite similar.

To enable this option, set org-similarity-ignore-frontmatter to t (default is nil).

v2.1.1

• 5f20aeb fix: remove leading spaces from org ID inside link

v2.1.0

• 5e8c382 feat: lookup python3 instead of python
• 7549c32 docs: update installation instructions

v2.0.0: Merge pull request #29 from brunoarine/develop

This release version basically replaces all the Python code that was shipped along with the Elisp package with an external source. Effectively, the entire Python code has been extracted from org-similarity and re-baptized as findlike, so that it can also be re-used by other projects.

Given that findlike operates as a command-line program and returns the results in plain text format (just like org-similarity Python routines did), there should be no major interoperability hiccups. Results may be slightly different because of findlike's different tokenization procedure and list of stopwords.

Despite the big changes, with this separation of concerns, both packages will be easier to maintain and extend.

v1.0.0

• Added BM25 as an alternative algorithm.
• Added heading and prefix options.
• Formatted the score as a floating point number with two decimal places.
• Implemented a filter for minimum words.
• Added the org-similarity-remove-first option.
• Changed the default directory to ~/org.
• Decoupled the interpreter and dependency checks from the main function.
• Renamed predicate functions for clarity.
• Refactored command, executable, and dependency checks.
• Removed null entries from junkchars and stopwords.
• Implemented a benchmarking routine.
• Several bug fixes.

v0.3

New features

• The org-similarity-sidebuffer function was added to display the similarity results in a side-window.

Fixes

• The check for the availability of Python was updated to check for "python3" instead of "python".
• Fixed typo in the coded that prevented the results lists to be rendered correctly when org-similarity-use-id-links was set to t.
• The installation instructions for org-similarity were updated.
• The format_results function was modified to use the get_relative_path function.

v0.2

Automated installation of Python dependencies

Users no longer need to worry about installing org-similarity dependencies manually, and risking breaking something in their system. With the help of virtual environments, all dependencies will be downloaded during the first run.

Better org-roam v2 compatibility

By the time I implemented the first version of org-similarity, org-roam v2 wasn't released yet. The newer version heavily relies on org PROPERTIES, linking documents through their IDs rather than filenames. Now, org-similarity correctly parses the properties drawer and offers the option to create links using IDs or filenames (as before).

Refactored and optimized Python code

I must level with you: the first version of org-similarity was made in a hurry. The documents pre-processing step was kinda hacky and less than ideal. Now, the preprocessing steps are leveraged by orgparse, a nifty Python package to handle org-mode files.