[Bug]: Formula Environments are Not Distinguished by slicer

[emstruong]

What happened?

Using the VS Code version of flowr for slicing, the slicer does not show that bill_len in lm() refers to a variable in penguins and not the 'fake' bill_len just before it.

penguins

bill_len <- "Fake bill_len"

lm(bill_len ~ bill_dep , data = penguins)

Perhaps this is related to #2005 ?

What version of R do you have installed?

4.5.1

What version of flowR did you use?

v2.6.0 (Oct 13, 2025)

Relevant log output

flowR version: 2.6.0  
Extension version: 0.6.3 (Production mode)  
VS Code version: 1.105.0 (web false)  
Session: internal (active)  
OS: linux  
Extension config:  

{
  "server": {
    "host": "localhost",
    "port": 1042,
    "autoConnect": false,
    "connectionType": "auto"
  },
  "r": {
    "executable": "",
    "engine": "tree-sitter"
  },
  "style": {
    "sliceOpacity": 0.25,
    "sliceDisplay": "text",
    "tokenBackgroundColor": "rgba(0, 127, 0, 0.5)",
    "onlyHighlightActiveSelection": false,
    "mermaidMaxTextLength": 500000
  },
  "tree-sitter": {
    "timeout": 60000,
    "lax": true
  },
  "slice": {
    "automaticReconstruct": true,
    "revisitThreshold": 12
  },
  "dependencyView": {
    "autoReveal": 10,
    "updateType": "adaptive",
    "updateInterval": 10,
    "adaptiveCharacterLimit": 5000,
    "cacheLimit": 5,
    "keepOnError": true,
    "query": {
      "ignoreDefaults": false,
      "overrides": {
        "libraryFunctions": [],
        "sourceFunctions": [],
        "readFunctions": [],
        "writeFunctions": []
      },
      "enabledCategories": [
        "library",
        "read",
        "source",
        "write",
        "visualize"
      ]
    }
  },
  "verboseLog": false,
  "debug": {
    "flowrLogLevel": "Error"
  },
  "errorMessage": {
    "Timer": 5000
  },
  "config": {
    "ignoreSourceCalls": false,
    "solver": {
      "variableHandling": "alias",
      "pointerTracking": false,
      "resolveSource": {
        "ignoreCapitalization": true,
        "inferWorkingDirectory": "active-script",
        "searchPath": [],
        "dropPaths": "once"
      }
    },
    "overwriteBuiltIns": {
      "loadDefaults": true,
      "definitions": []
    }
  },
  "linter": {
    "enabledRules": [
      "deprecated-functions",
      "file-path-validity",
      "seeded-randomness",
      "absolute-file-paths",
      "unused-definitions",
      "dataframe-access-validation",
      "dead-code",
      "useless-loop"
    ],
    "ruleConfigs": {},
    "updateType": "adaptive",
    "updateInterval": 10,
    "adaptiveCharacterLimit": 5000
  }
}

Meta Information

• The issue is urgent
• I used flowR as a docker image

[EagleoutIce]

First of all thank you for reporting!
In general i'd say this is more related to the fact that we are still working on getting to know the things exported by packages on the system (once we have namespace file support in #1983, we intend to re-mine all packages to get these + the shape information - which is indeed related to #2005). Once we have this we know that bill_len refers to penguins otherwise we cannot be sure (given that penguins "could" be anything).
Yet we should make this clearer in the presentation of the slicing result, no matter what we "know".

(By the way, thank you for choosing the penguins dataset for the example, I have a personal knack for penguins :D)