Merge pull request #21 from 0xrsydn/fix/ksei-spaced-headers

fix: handle spaced KSEI PDF headers + bump to v0.2.3

src/api/msn/map.rs

@@ -186,13 +186,7 @@ fn normalize_percentish(value: Option<f64>) -> Option<f64> {
 }
 
 fn sanitize_current_ratio(value: Option<f64>) -> Option<f64> {
-    value.and_then(|number| {
-        if !number.is_finite() || number < 0.01 {
-            None
-        } else {
-            Some(number)
-        }
-    })
+    value.filter(|number| number.is_finite() && *number >= 0.01)
 }
 
 fn round_price(value: f64) -> i64 {

src/cli/stocks.rs

@@ -983,10 +983,10 @@ fn sort_screener_quotes(quotes: &mut [Quote], filter: &str) {
             });
         }
         "high-volume" => {
-            quotes.sort_by(|a, b| b.volume.cmp(&a.volume));
+            quotes.sort_by_key(|quote| std::cmp::Reverse(quote.volume));
         }
         "large-cap" => {
-            quotes.sort_by(|a, b| b.market_cap.unwrap_or(0).cmp(&a.market_cap.unwrap_or(0)));
+            quotes.sort_by_key(|quote| std::cmp::Reverse(quote.market_cap.unwrap_or(0)));
         }
         _ => {}
     }

src/ownership/db.rs

@@ -196,10 +196,7 @@ fn insert_ksei_holdings_rows(
             )
             .map_err(|e| IdxError::DatabaseError(e.to_string()));
 
-        match changed {
-            Ok(n) => inserted += n,
-            Err(err) => return Err(err),
-        }
+        inserted += changed?;
     }
 
     Ok(inserted)

src/ownership/graph.rs

@@ -122,7 +122,7 @@ pub fn format_graph_text(nodes: &[GraphNode], edges: &[GraphEdge]) -> String {
     out.push_str(&format!("nodes: {}  edges: {}\n", nodes.len(), edges.len()));
 
     for (ticker_id, mut rels) in ticker_to_entities {
-        rels.sort_by(|a, b| b.percentage_bps.cmp(&a.percentage_bps));
+        rels.sort_by_key(|edge| std::cmp::Reverse(edge.percentage_bps));
         let ticker_label = labels.get(ticker_id).map(|v| v.0).unwrap_or(ticker_id);
         out.push_str(&format!("\n{ticker_label} [TICKER]\n"));
 

src/ownership/parser.rs

@@ -17,6 +17,7 @@ const HEADER_LABELS: &[&str] = &[
     "ISSUERNAME",
     "INVESTORNAME",
     "INVESTORTYPE",
+    "INVESTORCLASSIFICATION",
     "LOCALFOREIGN",
     "NATIONALITY",
     "DOMICILE",
@@ -31,14 +32,15 @@ const HOLDER_REGISTER_SCHEMA_MARKERS: &[&str] = &[
     "TEXT=\"SHARE_CODE\"",
     "TEXT=\"INVESTOR_NAME\"",
     "TEXT=\"INVESTOR_TYPE\"",
+    "TEXT=\"INVESTOR_CLASSIFICATION\"",
     "TEXT=\"LOCAL_FOREIGN\"",
     "TEXT=\"TOTAL_HOLDING_SHARES\"",
     "TEXT=\"PERCENTAGE\"",
 ];
 const ANNOUNCEMENT_WRAPPER_SCHEMA_MARKERS: &[&str] =
-    &["TEXT=\"PENGUMUMAN\"", "PT BURSA EFEK INDONESIA (BEI)"];
+    &["TEXT=\"PENGUMUMAN\"", "PT_BURSA_EFEK_INDONESIA_(BEI)"];
 const ABOVE_FIVE_SCHEMA_MARKERS: &[&str] =
-    &["TEXT=\"INVS\"", "REKENING TAMPUNGAN KSEI", "CLOSED MEMBER-"];
+    &["TEXT=\"INVS\"", "REKENING_TAMPUNGAN_KSEI", "CLOSED_MEMBER-"];
 const INVESTOR_TYPE_SCHEMA_MARKERS: &[&str] = &[
     "TEXT=\"STOCK_CODE\"",
     "TEXT=\"NUMBER_OF_SHARES\"",
@@ -103,7 +105,7 @@ pub fn extract_pdf_stext(path: &Path) -> Result<String, IdxError> {
 
 /// Classify a PDF schema from mutool stext XML before the row parser runs.
 pub fn classify_stext_xml(xml: &str) -> OwnershipPdfSchema {
-    let normalized = xml.to_ascii_uppercase();
+    let normalized = normalize_stext_for_classification(xml);
 
     if count_schema_markers(&normalized, HOLDER_REGISTER_SCHEMA_MARKERS) >= 5 {
         return OwnershipPdfSchema::HolderRegister;
@@ -121,6 +123,41 @@ pub fn classify_stext_xml(xml: &str) -> OwnershipPdfSchema {
     OwnershipPdfSchema::Unknown
 }
 
+/// Normalize stext XML for schema classification.
+///
+/// Newer KSEI PDFs use spaces in header text (e.g. `SHARE CODE` instead of
+/// `SHARE_CODE`, `INVESTOR CLASSIFICATION` instead of `INVESTOR_TYPE`).  The
+/// schema markers use the underscored form, so we replace spaces with
+/// underscores inside `text="..."` attribute values to match both layouts.
+fn normalize_stext_for_classification(xml: &str) -> String {
+    let upper = xml.to_ascii_uppercase();
+    let mut result = String::with_capacity(upper.len());
+    let bytes = upper.as_bytes();
+    let mut i = 0;
+
+    while i < bytes.len() {
+        if bytes[i..].starts_with(b"TEXT=\"") {
+            result.push_str("TEXT=\"");
+            i += 6; // skip past TEXT="
+            // Collect everything until the closing quote.
+            while i < bytes.len() && bytes[i] != b'"' {
+                let ch = bytes[i] as char;
+                result.push(if ch == ' ' { '_' } else { ch });
+                i += 1;
+            }
+            if i < bytes.len() {
+                result.push('"');
+                i += 1; // skip closing quote
+            }
+        } else {
+            result.push(bytes[i] as char);
+            i += 1;
+        }
+    }
+
+    result
+}
+
 /// Parse a KSEI ownership PDF into raw rows.
 /// Shells out to `mutool` for XML extraction, classifies the schema,
 /// and only parses the supported holder-register layout.
@@ -684,6 +721,12 @@ mod tests {
         );
     }
 
+    #[test]
+    fn classify_stext_xml_detects_spaced_holder_register_schema() {
+        let xml = include_str!("../../tests/fixtures/ksei_above1_spaced_stext_excerpt.xml");
+        assert_eq!(classify_stext_xml(xml), OwnershipPdfSchema::HolderRegister);
+    }
+
     #[test]
     fn classify_stext_xml_detects_legacy_above5_schema() {
         let xml = include_str!("../../tests/fixtures/ksei_above5_stext_excerpt.xml");