Yehowshua/ZIm
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Have read_file support images (#30435)

Browse source 
This is very basic support for them. There are a number of other TODOs
before this is really a first-class supported feature, so not adding any
release notes for it; for now, this PR just makes it so that if
read_file tries to read a PNG (which has come up in practice), it at
least correctly sends it to Anthropic instead of messing up.

This also lays the groundwork for future PRs for more first-class
support for images in tool calls across more image file formats and LLM
providers.

Release Notes:

- N/A

---------

Co-authored-by: Agus Zubiaga <hi@aguz.me>
Co-authored-by: Agus Zubiaga <agus@zed.dev>
This commit is contained in:
Richard Feldman 2025-05-13 10:58:00 +02:00 • committed by GitHub
parent f01af006e1
commit 8fdf309a4a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
30 changed files with 557 additions and 194 deletions
Download patch file Download diff file Expand all files Collapse all files

48
crates/language_models/src/provider/copilot_chat.rs
View file

@ -5,8 +5,9 @@ use std::sync::Arc;
use anyhow::{Result, anyhow};
use collections::HashMap;
use copilot::copilot_chat::{
ChatMessage, ChatMessageContent, CopilotChat, ImageUrl, Model as CopilotChatModel, ModelVendor,
Request as CopilotChatRequest, ResponseEvent, Tool, ToolCall,
ChatMessage, ChatMessageContent, ChatMessagePart, CopilotChat, ImageUrl,
Model as CopilotChatModel, ModelVendor, Request as CopilotChatRequest, ResponseEvent, Tool,
ToolCall,
};
use copilot::{Copilot, Status};
use futures::future::BoxFuture;
@ -20,12 +21,14 @@ use language_model::{
AuthenticateError, LanguageModel, LanguageModelCompletionError, LanguageModelCompletionEvent,
LanguageModelId, LanguageModelName, LanguageModelProvider, LanguageModelProviderId,
LanguageModelProviderName, LanguageModelProviderState, LanguageModelRequest,
LanguageModelRequestMessage, LanguageModelToolChoice, LanguageModelToolSchemaFormat,
LanguageModelToolUse, MessageContent, RateLimiter, Role, StopReason,
LanguageModelRequestMessage, LanguageModelToolChoice, LanguageModelToolResultContent,
LanguageModelToolSchemaFormat, LanguageModelToolUse, MessageContent, RateLimiter, Role,
StopReason,
};
use settings::SettingsStore;
use std::time::Duration;
use ui::prelude::*;
use util::debug_panic;
use super::anthropic::count_anthropic_tokens;
use super::google::count_google_tokens;
@ -198,6 +201,10 @@ impl LanguageModel for CopilotChatLanguageModel {
self.model.supports_tools()
}
fn supports_images(&self) -> bool {
self.model.supports_vision()
}
fn tool_input_format(&self) -> LanguageModelToolSchemaFormat {
match self.model.vendor() {
ModelVendor::OpenAI | ModelVendor::Anthropic => {
@ -447,9 +454,28 @@ fn into_copilot_chat(
Role::User => {
for content in &message.content {
if let MessageContent::ToolResult(tool_result) = content {
let content = match &tool_result.content {
LanguageModelToolResultContent::Text(text) => text.to_string().into(),
LanguageModelToolResultContent::Image(image) => {
if model.supports_vision() {
ChatMessageContent::Multipart(vec![ChatMessagePart::Image {
image_url: ImageUrl {
url: image.to_base64_url(),
},
}])
} else {
debug_panic!(
"This should be caught at {} level",
tool_result.tool_name
);
"[Tool responded with an image, but this model does not support vision]".to_string().into()
}
}
};
messages.push(ChatMessage::Tool {
tool_call_id: tool_result.tool_use_id.to_string(),
content: tool_result.content.to_string(),
content,
});
}
}
@ -460,18 +486,18 @@ fn into_copilot_chat(
MessageContent::Text(text) | MessageContent::Thinking { text, .. }
if !text.is_empty() =>
{
if let Some(ChatMessageContent::Text { text: text_content }) =
if let Some(ChatMessagePart::Text { text: text_content }) =
content_parts.last_mut()
{
text_content.push_str(text);
} else {
content_parts.push(ChatMessageContent::Text {
content_parts.push(ChatMessagePart::Text {
text: text.to_string(),
});
}
}
MessageContent::Image(image) if model.supports_vision() => {
content_parts.push(ChatMessageContent::Image {
content_parts.push(ChatMessagePart::Image {
image_url: ImageUrl {
url: image.to_base64_url(),
},
@ -483,7 +509,7 @@ fn into_copilot_chat(
if !content_parts.is_empty() {
messages.push(ChatMessage::User {
content: content_parts,
content: content_parts.into(),
});
}
}
@ -523,9 +549,9 @@ fn into_copilot_chat(
messages.push(ChatMessage::Assistant {
content: if text_content.is_empty() {
None
ChatMessageContent::empty()
} else {
Some(text_content)
text_content.into()
},
tool_calls,
});