NLP++ Language Extension
Only Computer Language Dedicated to Natural Language Processing
NLP++ is the only computer language in the world exclusively dedicated to natural language processing. It allows for creating digital human readers that use linguistic and world knowledge to parse and understand text. Whether it be for extracting or marking up "messy" text or full-blown NLP understanding, this language allows for "anything thinkable".
NLP++ and VisualText Now Unicode
The NLP-Engine for NLP++ now works with Unicode (UTF8 via the ICU C++ Package) including the ability to work with emojis.
Quick Video Guides
Find quick video guides on how to install and use VisualText including a "hello world" video as well as an in-depth tour of VisualText.
This is a VSCode Language Extension for NLP++ that recreates the functionality of VisualText which has run on Microsoft Windows for the last two decades. NLP++ is a open source computer language specifically dedicated to creating text analyzers that mimic human readers and includes the NLP++ language and knowledge based system called the "conceptual grammar". NLP++ is used for any type of text processing from simple tagging or extraction, to full language parsing. There is a full english parser that is free an available for use (see information below).
The language extension and the required NLP-ENGINE run on Linux, Windows, and MacOS.
The VSCode NLP++ Language Extension allows for the fast development of NLP++ analyzers allowing users to:
NLP++ Example Analyzers
Example analyzers can be found in the "analyzers" folder in the NLP-ENGINE folder.
In order to use the VSCode NLP++ Language Extension, the NLP-ENGINE which is in the form of an executable and directory need to be present. Version one now includes this as part of the NLP language extension.
NLP Engine Overview
The NLP-ENGINE now comes with the NLP++ Language extension but is available separately from the VisualText github repository. The engine can run as a standalone executable (nlp.exe) that runs on Linux, Windows, and MacOS or it can be embedded into c++ code.
Types of Analyzers Commonly Written Using NLP++
There are many types of analyzers that are written by NLP++ programmers including:
This file will automatically get generated when a new analyzer is created in VisualText VSCode. It is located in the .vscode directory under the folder for an individual analyzer.
There are many details in the windows version of VisualText that are yet to be implemented in the VSCode version.
Added chartok tokenize which breaks apart all characters
Added windows Help.chm file in context menu for windows only
Added cancel analyzer during processing
Don't open text file when analyzing. Also @MULTI fixes and other fixes.
Fixed numerous problems with fileOps
Added "Generate @PATH" from tree
Added icons to output view
Fixed version number bug in linux and mac
Added unicode character offsets to trees
Mac and Linux fixes
Now uses the unicode version of the nlp-engine
Now handles utf8 files and characters (unicode)
Various minor issues fixed
Added stable file and directory deletes and copies
Added copy single analyzer or all analyzers to chosen folder
Added analyzer operation queue
Added sort & unique to text files
Major overhaul on updater
Reveals sequence number for an nlp file. This will eventually select the sequence item once Microsoft fully implements the reveal function for tree items.
Added delete logs to textview. Initialization messages now to LOGGING tab.
Simple search for function. This is a workable hack.
Added online browser lookup help where links work. This is to fix the fact that links don't work in the WebView inside VSCode.
Add clearing log folders for all analyzers for archiving analyzers
Copies settings.json file to the current work file for special NLP++ colorization
Added comment / uncomment lines
Added first version progress bar
Fixed autogenerate rule
Analyzer log files now have the extension .tree, added duplicate pass, update nlp-engine fix
Added updating VisualText files version separately allowing for independently updating
Now checks nlp-engine version and updates if newer exists. v1.5.2 improved lexer.
Allow for creating different type analyzers including basic and English
Changed all the pat files to nlp but still will process both extensions. Now downloads the latest version of nlp-engine at installation time, no longer included in the extension.
Fixed one line bug for the sequence file and allow for multiple pass selections for inserting.
Allows for dicttok, dicttokz, and cmltok variations of the tokenizer. Also delete and remame files also changes the log files directory name correctly in the textView.
Analyze a folder of texts
Fixed color highlights rule rewrite line
Updated NLP Engine with empty tmp folders
Added duplicate line for rule editor
The NLP engine executable and engine directory are now included in the NLP language extension for VSCode. The location of the engine is located in a subdirectory of the nlp extensions located in the extension directory located on the local computer. See documentation for VSCODE if you want to know the location of the directory which is different for different platforms.
Fixing minor problems
Fixing paths for calling nlp.exe and missing npm package
Fixed file path problem across platforms
Initial version with ongoing minor fixes
VSCode Language Extension
Follow these instructions to install the development code for VSCode extension:
At which point you can edit the sources and launch debug sessions via F5 and included launch configurations.
The NLP Engine which is written in C++ is also open source and can be downloaded in development mode. You can find how to download and develop the NLP Engine code at: https://github.com/VisualText/nlp-engine.