Introduce the new index chat settings

crates/milli/src/update/chat.rs

@@ -0,0 +1,45 @@
+use deserr::Deserr;
+use serde::{Deserialize, Serialize};
+use utoipa::ToSchema;
+
+use crate::index::ChatConfig;
+use crate::prompt::{default_max_bytes, PromptData};
+use crate::update::Setting;
+
+#[derive(Debug, Clone, Default, Serialize, Deserialize, PartialEq, Eq, Deserr, ToSchema)]
+#[serde(deny_unknown_fields, rename_all = "camelCase")]
+#[deserr(deny_unknown_fields, rename_all = camelCase)]
+pub struct ChatSettings {
+    #[serde(default, skip_serializing_if = "Setting::is_not_set")]
+    #[deserr(default)]
+    #[schema(value_type = Option<String>)]
+    pub description: Setting<String>,
+
+    /// A liquid template used to render documents to a text that can be embedded.
+    ///
+    /// Meillisearch interpolates the template for each document and sends the resulting text to the embedder.
+    /// The embedder then generates document vectors based on this text.
+    #[serde(default, skip_serializing_if = "Setting::is_not_set")]
+    #[deserr(default)]
+    #[schema(value_type = Option<String>)]
+    pub document_template: Setting<String>,
+
+    /// Rendered texts are truncated to this size. Defaults to 400.
+    #[serde(default, skip_serializing_if = "Setting::is_not_set")]
+    #[deserr(default)]
+    #[schema(value_type = Option<usize>)]
+    pub document_template_max_bytes: Setting<usize>,
+}
+
+impl From<ChatConfig> for ChatSettings {
+    fn from(config: ChatConfig) -> Self {
+        let ChatConfig { description, prompt: PromptData { template, max_bytes } } = config;
+        ChatSettings {
+            description: Setting::Set(description),
+            document_template: Setting::Set(template),
+            document_template_max_bytes: Setting::Set(
+                max_bytes.unwrap_or(default_max_bytes()).get(),
+            ),
+        }
+    }
+}

crates/milli/src/update/mod.rs

@@ -1,4 +1,5 @@
 pub use self::available_ids::AvailableIds;
+pub use self::chat::ChatSettings;
 pub use self::clear_documents::ClearDocuments;
 pub use self::concurrent_available_ids::ConcurrentAvailableIds;
 pub use self::facet::bulk::FacetsUpdateBulk;
@@ -13,6 +14,7 @@ pub use self::words_prefix_integer_docids::WordPrefixIntegerDocids;
 pub use self::words_prefixes_fst::WordsPrefixesFst;
 
 mod available_ids;
+mod chat;
 mod clear_documents;
 mod concurrent_available_ids;
 pub(crate) mod del_add;

crates/milli/src/update/settings.rs

@@ -13,7 +13,7 @@ use time::OffsetDateTime;
 
 use super::del_add::{DelAdd, DelAddOperation};
 use super::index_documents::{IndexDocumentsConfig, Transform};
-use super::IndexerConfig;
+use super::{ChatSettings, IndexerConfig};
 use crate::attribute_patterns::PatternMatch;
 use crate::constants::RESERVED_GEO_FIELD_NAME;
 use crate::criterion::Criterion;
@@ -22,11 +22,11 @@ use crate::error::UserError;
 use crate::fields_ids_map::metadata::{FieldIdMapWithMetadata, MetadataBuilder};
 use crate::filterable_attributes_rules::match_faceted_field;
 use crate::index::{
-    IndexEmbeddingConfig, PrefixSearch, DEFAULT_MIN_WORD_LEN_ONE_TYPO,
+    ChatConfig, IndexEmbeddingConfig, PrefixSearch, DEFAULT_MIN_WORD_LEN_ONE_TYPO,
     DEFAULT_MIN_WORD_LEN_TWO_TYPOS,
 };
 use crate::order_by_map::OrderByMap;
-use crate::prompt::default_max_bytes;
+use crate::prompt::{default_max_bytes, PromptData};
 use crate::proximity::ProximityPrecision;
 use crate::update::index_documents::IndexDocumentsMethod;
 use crate::update::{IndexDocuments, UpdateIndexingStep};
@@ -185,6 +185,7 @@ pub struct Settings<'a, 't, 'i> {
     localized_attributes_rules: Setting<Vec<LocalizedAttributesRule>>,
     prefix_search: Setting<PrefixSearch>,
     facet_search: Setting<bool>,
+    chat: Setting<ChatSettings>,
 }
 
 impl<'a, 't, 'i> Settings<'a, 't, 'i> {
@@ -223,6 +224,7 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
             localized_attributes_rules: Setting::NotSet,
             prefix_search: Setting::NotSet,
             facet_search: Setting::NotSet,
+            chat: Setting::NotSet,
             indexer_config,
         }
     }
@@ -453,6 +455,14 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
         self.facet_search = Setting::Reset;
     }
 
+    pub fn set_chat(&mut self, value: ChatSettings) {
+        self.chat = Setting::Set(value);
+    }
+
+    pub fn reset_chat(&mut self) {
+        self.chat = Setting::Reset;
+    }
+
     #[tracing::instrument(
         level = "trace"
         skip(self, progress_callback, should_abort, settings_diff),
@@ -1238,6 +1248,45 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
         Ok(())
     }
 
+    fn update_chat_config(&mut self) -> heed::Result<bool> {
+        match &mut self.chat {
+            Setting::Set(ChatSettings {
+                description: new_description,
+                document_template: new_document_template,
+                document_template_max_bytes: new_document_template_max_bytes,
+            }) => {
+                let mut old = self.index.chat_config(self.wtxn)?;
+                let ChatConfig {
+                    ref mut description,
+                    prompt: PromptData { ref mut template, ref mut max_bytes },
+                } = old;
+
+                match new_description {
+                    Setting::Set(d) => *description = d.clone(),
+                    Setting::Reset => *description = Default::default(),
+                    Setting::NotSet => (),
+                }
+
+                match new_document_template {
+                    Setting::Set(dt) => *template = dt.clone(),
+                    Setting::Reset => *template = Default::default(),
+                    Setting::NotSet => (),
+                }
+
+                match new_document_template_max_bytes {
+                    Setting::Set(m) => *max_bytes = NonZeroUsize::new(*m),
+                    Setting::Reset => *max_bytes = Some(default_max_bytes()),
+                    Setting::NotSet => (),
+                }
+
+                self.index.put_chat_config(self.wtxn, &old)?;
+                Ok(true)
+            }
+            Setting::Reset => self.index.delete_chat_config(self.wtxn),
+            Setting::NotSet => Ok(false),
+        }
+    }
+
     pub fn execute<FP, FA>(mut self, progress_callback: FP, should_abort: FA) -> Result<()>
     where
         FP: Fn(UpdateIndexingStep) + Sync,
@@ -1275,6 +1324,7 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
         self.update_facet_search()?;
         self.update_localized_attributes_rules()?;
         self.update_disabled_typos_terms()?;
+        self.update_chat_config()?;
 
         let embedding_config_updates = self.update_embedding_configs()?;