[ALPHA] Upcoming Plugin: HiddenLexers

PeterJones

NOTE: I split this out from the Using Notepad++ to Organize Game Guides and Documentation because it distracts from the original conversation, and is really about something separate from that discussion.

PeterJones

Updated artifacts: https://github.com/pryrt/NppPlugin-HiddenLexers/actions/runs/31041591210 (again, limited time availability, only for alpha testing)

Got rid of the “Hello” actions. ;-)
Added About box
Added Edit Config File to open the correct JSON file
Added Reread Config to reload the settings from the JSON
- when you are editing the JSON in Notepad++, when you save that file, it should automatically reread the config. (you will have to activate the tab with the example file to see the update in highlighting)

This build should make it much easier for incremental changes to the JSON, without having to reload Notepad++ every time. :-)

Also, now I’ve updated the README at https://github.com/pryrt/NppPlugin-HiddenLexers

Call For Testers

It seems to be basically working at this point. If @mpheath and others could give it a test, that would be appreciated. I don’t personally need this plugin, as I don’t have any hidden lexers I want to activate, so I don’t know if I’ve missed anything critical or obvious. But from my tests with STATA (one of the languages that the original HiddenLexers.py handled) and ASCIIDoc (per the start of this conversation), I think it’s working reasonably.

I know that Lexilla has various config-variables for some of the lexers. If not having config variables available is a deal breaker for you, let me know which lexer you need it for, and give an example file and JSON to show me why that option is critical, and I can prioritize getting it added before I do the v1.0 release. Otherwise, the lexilla-options will probably not be included until a future version.

PeterJones

I was able to find an example lexer that had an option that I could see the effect of, so was able to confirm that I seemed to be processing correctly.

v0.0.1.1-dev

add options processing

example JSON:

{
  "extensions": {
    "md": "markdown",
    "markdown": "markdown"
  },
  "lexers": {
    "markdown": {
        "0  SCE_MARKDOWN_DEFAULT"    : { "fgColor": "", "bgColor": "", "fontStyle": "0" },
        "1  SCE_MARKDOWN_LINE_BEGIN" : { "fgColor": "DADADA", "bgColor": "", "fontStyle": "0" },
        "2  SCE_MARKDOWN_STRONG1"    : { "fgColor": "EECCAA", "bgColor": "", "fontStyle": "1" },
        "3  SCE_MARKDOWN_STRONG2"    : { "fgColor": "EECCAA", "bgColor": "4F4F4F", "fontStyle": "1" },
        "4  SCE_MARKDOWN_EM1"        : { "fgColor": "AADDFF", "bgColor": "", "fontStyle": "2" },
        "5  SCE_MARKDOWN_EM2"        : { "fgColor": "AADDFF", "bgColor": "4F4F4F", "fontStyle": "2" },
        "6  SCE_MARKDOWN_HEADER1"    : { "fgColor": "FFFFFF", "bgColor": "8F8F8F", "fontStyle": "1" },
        "7  SCE_MARKDOWN_HEADER2"    : { "fgColor": "CCffff", "bgColor": "7F7F7F", "fontStyle": "1" },
        "8  SCE_MARKDOWN_HEADER3"    : { "fgColor": "CCbbbb", "bgColor": "6F6F6F", "fontStyle": "1" },
        "9  SCE_MARKDOWN_HEADER4"    : { "fgColor": "CC7777", "bgColor": "5F5F5F", "fontStyle": "1" },
        "10 SCE_MARKDOWN_HEADER5"    : { "fgColor": "CC3333", "bgColor": "4F4F4F", "fontStyle": "1" },
        "11 SCE_MARKDOWN_HEADER6"    : { "fgColor": "CC0000", "bgColor": "3F3F3F", "fontStyle": "1" },
        "12 SCE_MARKDOWN_PRECHAR"    : { "fgColor": "5F5F5F", "bgColor": "4F4F4F", "fontStyle": "0" },
        "13 SCE_MARKDOWN_ULIST_ITEM" : { "fgColor": "0000CC", "bgColor": "4F4F4F", "fontStyle": "0" },
        "14 SCE_MARKDOWN_OLIST_ITEM" : { "fgColor": "0000CC", "bgColor": "4F4F4F", "fontStyle": "0" },
        "15 SCE_MARKDOWN_BLOCKQUOTE" : { "fgColor": "00CC00", "bgColor": "4F4F4F", "fontStyle": "0" },
        "16 SCE_MARKDOWN_STRIKEOUT"  : { "fgColor": "6F6F6F", "bgColor": "", "fontStyle": "0" },
        "17 SCE_MARKDOWN_HRULE"      : { "fgColor": "FFFFFF", "bgColor": "0020C0", "fontStyle": "0" },
        "18 SCE_MARKDOWN_LINK"       : { "fgColor": "8F8FFF", "bgColor": "", "fontStyle": "4" },
        "19 SCE_MARKDOWN_CODE"       : { "fgColor": "FFFF00", "bgColor": "6F6F6F", "fontStyle": "0" },
        "20 SCE_MARKDOWN_CODE2"      : { "fgColor": "FFFF00", "bgColor": "6F6F6F", "fontStyle": "0" },
        "21 SCE_MARKDOWN_CODEBK"     : { "fgColor": "FFFF00", "bgColor": "6F6F6F", "fontStyle": "0" },
        "options": { "lexer.markdown.header.eolfill": "1" },
        "keywords": [ "", "", "", "", "", "", "", "", "" ]
    }
  }
}

I verified the functionality using AllStyles.md from the lexilla test suite in N++ codebase; I can change from "1" to "0" and watch the header lines change:

“1”:
“0”:

So I believe that options are implemented now.

Also, this one is in a release tag, so you can either find the artifacts as before, or you can download a normal PluginsAdmin-style zipfile which can be unzipped and the DLL installed, as with any normal manual plugin installation.

mpheath

@PeterJones You have enabled the lexer property of lexer.markdown.header.eolfill though the styles are not using it as I see in the images. For clarity, I am focussing on eolfilled on the style id rather then the property itself so could an extra attribute not available currently.

Here is a example in SciTE on the level 1 header:

As the styling shows:

{6}# Tools
{1}
{0}Lua scripts that can be executed by dofile.{1}


{7}## Ideas
{1}

there is nothing preventing style id 6 to fill the whole line with the background colour if the eolfilled style is used.

I added this to SciTE’s Option -> Open User Options File

# SCE_MARKDOWN_HEADER1: "#" - Level-one header
style.markdown.6=fore:#FF7766,bold,back:#700000,eolfilled

The json file could have eolFilled attribute setting like the fontStyle attribute so is applied for the style id.

PeterJones

@mpheath said:

@PeterJones You have enabled the lexer property of lexer.markdown.header.eolfill though the styles are not using it as I see in the images.

Well, it partially used it, in that the style went to the end of the header text instead of just the end of the ### symbols.

The json file could have eolFilled attribute setting like the fontStyle attribute so is applied for the style id.

I didn’t know/remember that SCI_STYLESETEOLFILLED existed until your post made me go look for it. I will have to add that attribute to the style info in the JSON

mpheath

@PeterJones I just did an edit before you replied. The property does change the processing of headers the eolfilled though on the style itself is absent.

PeterJones

v0.0.1.2-dev

Add the "eolFilled": "1" attribute to the JSON (expected values: not supplied, "","0","1"). This will allow lexers that support the SCI_STYLESETEOLFILLED style to fill the styling to the end of the window instead of the end of the text.

        "6  SCE_MARKDOWN_HEADER1"    : { "fgColor": "FFFFFF", "bgColor": "8F8F8F", "fontStyle": "1", "eolFilled": "1" },
        "7  SCE_MARKDOWN_HEADER2"    : { "fgColor": "CCffff", "bgColor": "7F7F7F", "fontStyle": "1", "eolFilled": "1" },
        "8  SCE_MARKDOWN_HEADER3"    : { "fgColor": "CCbbbb", "bgColor": "6F6F6F", "fontStyle": "1", "eolFilled": "1" },
        "9  SCE_MARKDOWN_HEADER4"    : { "fgColor": "CC7777", "bgColor": "5F5F5F", "fontStyle": "1", "eolFilled": "1" },
        "10 SCE_MARKDOWN_HEADER5"    : { "fgColor": "CC3333", "bgColor": "4F4F4F", "fontStyle": "1", "eolFilled": "1" },
        "11 SCE_MARKDOWN_HEADER6"    : { "fgColor": "CC0000", "bgColor": "474747", "fontStyle": "1", "eolFilled": "1" },

PeterJones

@mpheath ,

I am curious: do you think it’s enough to warrant trying to get v1 in Plugins Admin for the upcoming v8.9.8 RC? Or should I give more time for more people to test? (Or do you think it’s really worth publishing at all?)

(And secondary question: for getting the styler output from SciTE, are you just running TestLexers.exe from a N++ or just-scintilla build? or is there another way of generating that? I have my PythonScript styleDebugger script for doing it inside N++, but was curious if there was an easy way to do it directly from SciTE)

mpheath

@PeterJones said:

@mpheath ,

I am curious: do you think it’s enough to warrant trying to get v1 in Plugins Admin for the upcoming v8.9.8 RC? Or should I give more time for more people to test? (Or do you think it’s really worth publishing at all?)

Rushing anything can lead to a buggy release. You may want to make sure decisions like file format with the json … is good which is important. If OK, then release. I cannot come up with anything that prevents it. I do not see anyone desperate for a rushed release so you choose.

(And secondary question: for getting the styler output from SciTE, are you just running TestLexers.exe from a N++ or just-scintilla build? or is there another way of generating that? I have my PythonScript styleDebugger script for doing it inside N++, but was curious if there was an easy way to do it directly from SciTE)

I do it from Notepad++ with PythonScript, SciTE with Lua or if want to do it the hard way, with TestLexers.exe.

TestLexers.exe should be used for actual Lexilla lexer changes as it does additional things like checking CRLF/LF modes … so is official for Lexilla contributions. From SciTE, I use a Lua script to output the data to the console pane. In my previous comment where I posted the styling information was printed from a Lua function that uses CharAt and StyleIndexAt to know where to insert the {N} style ids.

I did a quick hack as my script has MsgBoxes and other incompatable code for an uncustomized SciTE.

File: printStyles.lua in same directory as SciTE:

-- about: Print Style IDs for the current document.

do
    local lastID = -1
    local text = ''
    local startPos = 0
    local endPos = editor.Length

    -- Set to selection positions.
    if editor.SelectionEnd > editor.SelectionStart then
        startPos = editor.SelectionStart
        endPos = editor.SelectionEnd
    end

    -- Process each char.
    for i = startPos, endPos do
        local ch = editor.CharAt[i]

        if ch < 0 then
            ch = ch + 256
        end

        local character = string.char(ch)
        local styleId = editor.StyleIndexAt[i]

        if styleId ~= lastID then
            text = text .. '{' .. tostring(styleId) .. '}'
        end

        text = text .. character

        lastID = styleId
    end

    -- Output the reformatted text.
    if text ~= '' then
        print(text)
    end
end

and add to the User Options File:

# 01 PrintDocStyles similar to TestLexers.exe
command.name.1.*.*=PrintDocStyles
command.1.*.*=dofile $(SciteDefaultHome)\printStyles.lua
command.subsystem.1.*.*=3
command.save.before.1.*.*=2

or if using Sc1, can just add it to the Global Options File to save in same directory as the Sc1 executable file.

In the SciTE Tools menu drop down, should then show an item named PrintDocStyles or press Ctrl+1 keys to run the function.

mpheath

@PeterJones Looking through the readme.md. Perhaps could make it easier for users to have a properties list for the lexer options, like possibly from

https://github.com/mpheath/xconfig-for-notepad-plus-plus/blob/bed2f415648623b6ce0f49b6ed333fbedfdba8be/xconfig.properties#L122-L585

Possibly, a little bit of your regex tricks could rearrange to get the list of options into a file for users to view.

SciTE Doc has a good lexer property listing:
https://scintilla.org/SciTEDoc.html#property-xml.auto.close.tags
which is where the source of information can be viewed in HTML.

Getting all the information to insert into the json file is the challenge.

I just thought of the Nim lexer. Notepad++ still uses the deprecated Nimrod lexer though I do not use Nim. You can see some of the hackery I had to do to recognize the Nim lexer from the Nimrod lexer here with PythonScript. Nim users who want an improved Nim lexer can access it with the HiddenLexers plugin. So benefits exist for those with the interest in the hidden lexers.