Project info

Protex is a programmable text extractor, the project ships with a configuration editor allowing users to create, modify and test their custom text extraction and transformation configurations. After a configuration is created it can be exported as an artifact and used to run on documents at scale with it's extraction api.

Protex changelog

This project adheres to semver

[0.5.2-alpha] - Unreleased

Added

Theme selector
Offset slider input box

Removed

Ascii fox RIP: 02-14-23 - 03-22-23

[0.5.1-alpha] - 03-20-23

Changed

Updated to zig 0.11.0-dev.2191
Updated build.zig
Organize project structure

[0.5.0-alpha] - 03-17-23

Added

Visual scripting system
Operation queues

Removed

The category paradigm in favor of operation queues.

[0.4.1-alpha] - 02-16-23

Added

Toolbar and actions changing what is displayed to the user ( example: target specific extractor to work on ).

[0.4.0-alpha] - 02-14-23

Added

a Fox
Extract/Tag conditions can now specify specific patterns.
Script editing.
Dedicated pattern editing area.

Fixed

Text tabs/new-lines are now properly displayed.
Out of bounds index when using UNTIL OFFSET within a conditional statement.
App writing artifacts to wrong dir.

Removed

Generate config button and text that would appear because configurations are now written by main thread instead of a web server.

Changed

Text scrollbars are now hidden
Project name is now "protex"
Extractor now allways resolves with a script.

Added

Pressing '~' during conditional statement editing now prompts an auto complete for the pattern to include within the statement.

[0.3.1-alpha] - 02-08-23

Removed

Country/Language inputs within config, and refrences within extractor and packager.

Changed

Don't let index cache renderer.js

Fixed

User provided text is now properly rendering after clicking away.

[0.3.0-alpha] - 02-06-23

Changed

Packager now returns an error on failed hs compilation rather than a crash.
Web client is now a standalone electron application

Removed

Web server

[0.2.0-alpha] - 01-29-23

Added

Basic tagging conditions
Delete buttons for categories and extractors

[0.1.1-alpha] - 01-23-23

Added

.gitignore
Regex highlighting for selected patterns
Server writes configurations built in the web interface to disk.
Server can run packager and create artifacts from a web request.
Server can call extractor from a web request.

Changed

Extractor's Zig Init API no longer requires null terminated artifact path.
Packager exposes artifact creation functionality.
Refactoring of python module selector population.
When selecting text the pattern input which was last selected (if any) is filled with the user's text selection converted to a basic regex pattern.
Set resuse address flag when initializing stream server.
Refactor HTTP response creation.

Removed

Needless async functions

[0.1.0-alpha] - 01-12-23

Added

Web client and server

Fixed

Throwing away scratch space refrence per database load.

[0.0.5-alpha] - 01-06-23

Added

a README
Full support for multiple extractor definitions.
Json parsing on returned value from python callbacks.

Changed

Sempy run now writes to a utf8 encoded byte buffer rather than a wchar buffer.

Fixed

Extractor Unit test (again)
sempy init/deinit crash

[0.0.4-alpha] - 01-04-23

Fixed

Extractor Unit test
Memory leaks caused by moving off of a backing buffer.

Added

Loading of arbitrary python modules
CatBoxes...

Changed

Instances of Text paramaters are now marked as constant.
Decoupled python modules from category definition blocks.
Sempy now writes main module output to a buffer which is then exposed through extract api.

[0.0.3-alpha] - 12-31-22

Changed

Artifact structure, the packager and deserializer now handle new artifact structure
Improved error handling
Simplified backing allocator logic
Slab allocator now stores entire backing buffer slize instead of just it's ptr as an int

[0.0.2-alpha] - 12-15-22

Added

VERY EARLY embeded python 11 support

[0.0.1-alpha] - 11-25-22

Added

Slab alloactor
Debug slab visualizer
Init,Extract,Deinit