Refinements for regular release

Miscellaneous refinements, enhancements, etc.

• New option --query-file
• Revamped new, larger model files
• The model files are downloaded from an external website
• File format changes of the index DB

Full Changelog: v1.0.0-beta.6...v1.0.0-beta.11

Support macOS and revise installation instructions

In this release:

• the installation instructions have been revised.
• now officially supporting macOS 🎉.

Full Changelog: v1.0.0-beta.5...v1.0.0-beta.6

Bugfix release

This release includes the following fixes:

• When password protected pdf or docx files are found, the program now continues execution with a warning instead of stopping with an error message.
• Unused functions and classes have been removed, and the module structure has been optimized.

Full Changelog: v1.0.0-beta.4...v1.0.0-beta.5

Bugfix release

This release fixes the errors that occurred when on Windows:

• Reading PDF files
• Encoding error (invalid chars)

Full Changelog: v1.0.0-beta.1...v1.0.0-beta.4

Add indexing command

This release introduces the indexing command as an experimental feature.
(Unlike the indexing of the previous tool (d2vg), it is only about twice faster. Also, indexing is no longer incremental and must be done explicitly.)

Full Changelog: v1.0.0-alpha.5...v1.0.0-beta.1

Ease installation

• Simplified installation by making searching in PDF files optional.
• Model files were tuned to reduce file size.

Full Changelog: v1.0.0-alpha.4...v1.0.0-alpha.5

Refactoring, optimizations etc.

Some tunings of the code for beta release, including:

• Remove corner-case bugs.
• Remove redundant message.
• Refactoring
• Remove some (potential) performance glitches.

Full Changelog: v1.0.0-alpha.3...v1.0.0-alpha.4

First alpha release

This release is the first alpha version after the tool was renamed to dvg.

The difference from the previous d2vg are:

• The models have been changed to be based on SCDV.
• The model files became smaller ones and are now included in the release.
• Indexing was dropped because the speed improvement by parallelization (without indexing) has been achieved.