Semantic Interpretation Tester

short description or question.

detailed description

information text

Editor Shortcuts

CTRL + #: comment / un-comment line
CTRL + a: select all
CTRL + d: delete line
CTRL + f: search / replace dialog
CTRL + l: "go to line" dialog
CTRL + q: jump to last edited position
CTRL + y: redo
CTRL + z: undo

Preview masking (for non-ASCII characters)

Warning: applying masking will re-format the JSON grammar.

If the JSON values contain unicode characters (i.e. non-ASCII chars), they will get masked before creating the grammar definition (intermediate grammar), and compiling the executable JavaScript grammar.

You can preview the effects of masking by using mask and unmask.

Note: that the grammar will not be able to detect tokens that correspond to the masking pattern itself, e.g. the token ~~751F~~ will not be detected, but instead its unmasked representation 生.

The Grammar ID is internally used to identify the grammar, that is, the ID is used to manage multiple grammars.
If a grammar with the same ID already exists, it will replaced.

Upgrading old umlaut-encoded JSON grammars

Warning: applying upgrading/downgrading will re-format the JSON grammar.

The old file format for JSON grammars used a special masking for non-ASCII umlauts (e.g. ö or Ä; also for ß).

For example: ö → __oe__

The old file format was targeted at ASCII encoded files which required that the masking was already used in the JSON files themselves.
For example:

 "stop_word": [
	    "ich",
	    "m__oe__chte",
	    ...

The new file format internally uses a generic masking for all non-ASCII characters. But more importantly, it targets UTF-8 encoded files which allows using (almost all) unicode characters directly within the JSON files; i.e. no masking in the files themselves is required.
For example:

 "stop_word": [
	    "ich",
	    "möchte",
	    ...

or alternatively with new masking:

 "stop_word": [
	    "ich",
	    "m~~00F6~~chte",
	    ...

Unfortunately the new masking-mechanism is incompatible with the old one, i.e. JSON files which contain the old masking will not work correctly with the new masking.
As a result, old-style formatted JSON files (with masking) must be converted, i.e. old-style masking must be reverted (unmasked) so that the either no masking is done or, alternatively, the new masking style may be used, too.

For this purpose, the upgrade button will convert old-style masked umlauts back to the unmasked, original umlauts; and the downgrade button will reverse that effect (i.e. apply old-stlye masking).

Introduction

NOTE: On initial load of the page, the Language list is loaded from the file config/directories.json: for each found language directory (config/languages/), an entry is generated -- on selection, the corresponing JSON file is attempted to be read and compiled.
Note, that there may not be a JSON file for the lanuage available -- in this case an error message will be shown on selection.

You can use the file selector, for loading a JSON grammar from the local file system

Alternatively, you can enter (e.g. copy&paste) a JSON-formatted grammar into the JSON Grammar Definition field and then press the parse button.

Grammar ID information

JSON Grammar Tools

Masking information

Upgrade information

top

Compiled Parser Code:

This is the compiled JavaScript code for the grammar.

This code can be copied into a file: on loading (i.e. interpreting) the JavaScript file, the grammar will be registered with the SemanticInterpreter for the Grammar ID that is specified above in the Grammar Definition section.

The default location for the compiled JavaScript grammar files is

assets/www/gen/grammar/[Grammar ID]_grammar.json

Test Semantics:

Result:

Alternate Interpreter Result: (using alternate method for stopword removal)

Test Stopword Removal:

After Stopword Removal:

After Stopword Removal (alternative stopword processing method) :