在Microsoft Bot Framework中读取用户字符串列表及客户技能收集表单咨询

Hey there! Let's tackle your chatbot project questions—first fixing that form model code, then covering how to collect a list of skills from users in Microsoft Bot Framework.

I spotted a syntax issue in your current code: the last [Prompt] attribute isn't closed, and you're missing the getter/setter for that final property. Also, if you want to collect skills directly in the form, you can add a dedicated property for that. Here's the corrected, expanded version:

[Serializable]
public class ExperiencedDetailFormModel
{
    [Prompt("Please enter your {&}")]
    public string EducationalQualification { get; set; }

    [Prompt("Please enter name of your {&}")]
    public string CurrentOrganization { get; set; }

    [Prompt("Please enter your {&}")]
    public string CurrentPositionInTheOrganization { get; set; }

    // Add this property to collect raw skill input (we'll process it into a list later)
    [Prompt("Please list your skills separated by commas (e.g., C#, Python, SQL)")]
    public string SkillsInput { get; set; }
}

Quick note: The {&} in the Prompt attribute automatically inserts the property's name into the prompt text, which keeps your code DRY and consistent.

There are three common, user-friendly ways to collect a list of skills (or any string list) from users—pick the one that fits your bot's flow best:

Option 1: Single Text Input with Delimiter #

Ask users to enter all skills in one message separated by commas (or another delimiter), then split the input into a list. This works great for quick, straightforward input:

Example in a Waterfall Dialog #

// Prompt the user for skills
private async Task<DialogTurnResult> GetSkillsStepAsync(WaterfallStepContext stepContext, CancellationToken cancellationToken)
{
    return await stepContext.PromptAsync(
        "textPrompt",
        new PromptOptions { Prompt = MessageFactory.Text("Please list your skills separated by commas (e.g., C#, Python, SQL)") },
        cancellationToken);
}

// Process the input into a list
private async Task<DialogTurnResult> ProcessSkillsStepAsync(WaterfallStepContext stepContext, CancellationToken cancellationToken)
{
    var rawSkillsInput = (string)stepContext.Result;
    
    // Split, clean up whitespace, and remove empty entries
    List<string> userSkills = rawSkillsInput.Split(',')
                                        .Select(skill => skill.Trim())
                                        .Where(skill => !string.IsNullOrEmpty(skill))
                                        .ToList();
    
    // Save to your form model if needed
    var formModel = (ExperiencedDetailFormModel)stepContext.Options;
    formModel.SkillsInput = rawSkillsInput; // Store raw input, or track the list separately
    
    // Continue your dialog flow
    return await stepContext.NextAsync(userSkills, cancellationToken);
}

Option 2: Collect Skills One-by-One #

If you want to guide users to enter skills individually (and let them stop when done), use a loop in your waterfall dialog:

Example Loop Flow #

// Initialize list and prompt for first skill
private async Task<DialogTurnResult> CollectSingleSkillStepAsync(WaterfallStepContext stepContext, CancellationToken cancellationToken)
{
    if (!stepContext.Values.ContainsKey("UserSkills"))
    {
        stepContext.Values["UserSkills"] = new List<string>();
    }

    return await stepContext.PromptAsync(
        "textPrompt",
        new PromptOptions { Prompt = MessageFactory.Text("Enter one of your skills, or type 'done' when you're finished") },
        cancellationToken);
}

// Process each skill and loop until user says "done"
private async Task<DialogTurnResult> ProcessSingleSkillStepAsync(WaterfallStepContext stepContext, CancellationToken cancellationToken)
{
    var skillInput = (string)stepContext.Result;
    var userSkills = (List<string>)stepContext.Values["UserSkills"];

    if (skillInput.Equals("done", StringComparison.OrdinalIgnoreCase))
    {
        if (userSkills.Count == 0)
        {
            await stepContext.Context.SendActivityAsync(MessageFactory.Text("You haven't entered any skills yet—let's try again!"), cancellationToken);
            return await stepContext.ReplaceDialogAsync(nameof(CollectSingleSkillStepAsync), null, cancellationToken);
        }
        // Move to next step with the full skills list
        return await stepContext.NextAsync(userSkills, cancellationToken);
    }
    else
    {
        userSkills.Add(skillInput.Trim());
        await stepContext.Context.SendActivityAsync(MessageFactory.Text($"Added skill: {skillInput.Trim()}. Enter another or type 'done'."), cancellationToken);
        return await stepContext.ReplaceDialogAsync(nameof(CollectSingleSkillStepAsync), null, cancellationToken);
    }
}

Option 3: Use Adaptive Cards for Visual Input #

For a more polished experience, use an Adaptive Card with multiple input fields. Users can fill in skills directly, and you can parse the submitted values into a list:

Example Adaptive Card (JSON) #

{
    "type": "AdaptiveCard",
    "version": "1.0",
    "body": [
        {
            "type": "TextBlock",
            "text": "Tell us about your skills",
            "size": "Large",
            "weight": "Bolder"
        },
        {
            "type": "Input.Text",
            "id": "skill1",
            "placeholder": "Skill 1 (e.g., JavaScript)"
        },
        {
            "type": "Input.Text",
            "id": "skill2",
            "placeholder": "Skill 2 (optional)"
        },
        {
            "type": "Input.Text",
            "id": "skill3",
            "placeholder": "Skill 3 (optional)"
        }
    ],
    "actions": [
        {
            "type": "Action.Submit",
            "title": "Submit Skills"
        }
    ]
}

Parse Card Submission in Dialog #

// Send the Adaptive Card to the user
private async Task<DialogTurnResult> GetSkillsViaCardStepAsync(WaterfallStepContext stepContext, CancellationToken cancellationToken)
{
    var cardAttachment = new Attachment
    {
        ContentType = "application/vnd.microsoft.card.adaptive",
        Content = JsonConvert.DeserializeObject(/* Insert your Adaptive Card JSON string here */)
    };

    var promptOptions = new PromptOptions
    {
        Prompt = MessageFactory.Attachment(cardAttachment)
    };

    return await stepContext.PromptAsync("adaptiveCardPrompt", promptOptions, cancellationToken);
}

// Process the submitted card values into a list
private async Task<DialogTurnResult> ProcessCardSkillsStepAsync(WaterfallStepContext stepContext, CancellationToken cancellationToken)
{
    var submittedValues = (IDictionary<string, object>)stepContext.Result;
    List<string> userSkills = new List<string>();

    // Collect non-empty skill inputs
    for (int i = 1; i <= 3; i++)
    {
        if (submittedValues.TryGetValue($"skill{i}", out var skillObj) && 
            skillObj is string skill && 
            !string.IsNullOrEmpty(skill.Trim()))
        {
            userSkills.Add(skill.Trim());
        }
    }

    // Use the skills list in your bot logic
    return await stepContext.NextAsync(userSkills, cancellationToken);
}

内容的提问来源于stack exchange，提问作者Akbar Badhusha