Include range line detection with : instead of only -.

### Describe the feature or problem you'd like to solve

Some IDEs like Zed use a colon for range lines (e.g., L25:30), whereas Copilot only accepts hyphens (e.g., L25-30). Please add support for colon-based range recognition.

### Proposed solution

Supporting colon-based line ranges (e.g., L25:30) would provide the following benefits:
- Improved IDE Interoperability: Modern editors like Zed and certain terminal emulators use the file:line:column or Lstart:end convention as a standard. Supporting this would allow users to copy-paste line references directly from their editor into the CLI without manual editing.
- Reduced Friction & Better UX: Currently, users coming from environments that default to colons face a "syntax error" or ignored input when they use their habitual formatting. Adding : as a recognized separator makes the tool more intuitive and "just works."
- Consistency with POSIX and LSP Standards: Many developer tools and Language Server Protocol (LSP) outputs use colons to denote ranges. Aligning Copilot CLI with these existing ecosystems ensures it feels like a native part of the modern developer's toolkit.
- Wider Adoption: By removing small but repetitive manual tasks (like changing a : to a -), you lower the barrier to entry for power users who rely on fast, CLI-based workflows.

### Example prompts or workflows

```json
{
"sessionId": "c9b.....9b",
"prompt": [
{
"type": "resource",
"resource": {
"text": "

\n",
"uri": "file:///.../page.tsx#L6:8"
}
},
{
"type": "text",
"text": " This line starts from where to where?"
}
]
}
```

### Additional context

selected line in zed not readable by agent

Start by locating the CLI code that recognizes file line ranges and any tests covering hyphenated references such as L25-30. Check how resource URIs like file:///.../page.tsx#L6:8 are parsed, then add coverage for colon-separated ranges while preserving existing behavior. Done means the colon form is accepted and the relevant tests pass.

Rédigé par le modèle d'indexation à partir du texte de l'issue.