feat(import): support X-Bichon-Metadata and optimize CLI progress reporting

crates/core/src/envelope/extractor.rs

@@ -17,6 +17,7 @@
 // along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 use crate::common::AddrVec;
+use crate::envelope::meta::parse_bichon_metadata;
 use crate::envelope::utils::normalize_subject;
 use crate::error::code::ErrorCode;
 use crate::error::BichonResult;
@@ -33,6 +34,7 @@ use async_imap::types::Fetch;
 use bytes::Bytes;
 use mail_parser::{Address, HeaderName, Message, MessageParser, MimeHeaders};
 use tantivy::TantivyDocument;
+use tantivy::schema::Facet;
 use tracing::error;
 use uuid::Uuid;
 
@@ -159,6 +161,36 @@ async fn extract_envelope_core(
     let envelope_id = Uuid::new_v4().to_string();
     let now = utc_now!();
 
+
+    let mut final_tags = Vec::new();
+
+    if let Some(meta_header) = message.header_raw("X-Bichon-Metadata") {
+        if let Some(bmd) = parse_bichon_metadata(meta_header) {
+            if let Some(tags) = bmd.tags {
+                let validated_tags: Result<Vec<String>, _> = tags
+                    .iter()
+                    .map(|tag| {
+                        Facet::from_text(tag)
+                            .map(|_| tag.clone()) 
+                            .map_err(|e| e)
+                    })
+                    .collect();
+
+                match validated_tags {
+                    Ok(valid_list) => {
+                        final_tags = valid_list;
+                    }
+                    Err(e) => {
+                        eprintln!(
+                            "Tag validation failed, ignoring all tags: {:#?}",
+                            e
+                        );
+                    }
+                }
+            }
+        }
+    }
+
     let attachment_docs: Vec<TantivyDocument> = attachments
         .iter()
         .filter(|a| !a.inline || a.content_id.is_none())
@@ -210,7 +242,7 @@ async fn extract_envelope_core(
         thread_id,
         attachment_count,
         regular_attachment_count: attachment_docs.len(),
-        tags: None,
+        tags: (!final_tags.is_empty()).then_some(final_tags),
         account_email: None,
         mailbox_name: None,
         content_hash: email_content_hash,

crates/core/src/envelope/meta.rs

@@ -0,0 +1,15 @@
+use serde::{Deserialize, Serialize};
+
+use crate::base64_decode;
+
+#[derive(Clone, Debug, Default, Eq, PartialEq, Deserialize, Serialize)]
+pub struct BichonMetadata {
+    pub account_email: Option<String>,
+    pub mailbox_name: Option<String>,
+    pub tags: Option<Vec<String>>,
+}
+
+pub fn parse_bichon_metadata(header_value: &str) -> Option<BichonMetadata> {
+    let decoded = base64_decode!(header_value.trim());
+    serde_json::from_slice(&decoded).ok()
+}

crates/core/src/envelope/mod.rs

@@ -17,4 +17,5 @@
 // along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 pub mod extractor;
+pub mod meta;
 pub mod utils;

crates/core/src/store/tantivy/model.rs

@@ -106,6 +106,12 @@ impl EnvelopeWithAttachments {
             }
         }
 
+        if let Some(tags) = &self.envelope.tags {
+            for tag in tags {
+                doc.add_facet(fields.f_tags, tag);
+            }
+        }
+
         doc.add_u64(
             fields.f_attachment_count,
             self.envelope.attachment_count as u64,
@@ -166,7 +172,7 @@ impl EnvelopeWithAttachments {
                 fields.f_regular_attachment_count,
                 F_REGULAR_ATTACHMENT_COUNT,
             )? as usize,
-            tags: if tags.is_empty() { None } else { Some(tags) },
+            tags: (!tags.is_empty()).then_some(tags),
             content_hash: extract_string_field(doc, fields.f_content_hash, F_CONTENT_HASH)?,
             ingest_at: extract_i64_field(doc, fields.f_ingest_at, F_INGEST_AT)?,
         };