From 5e8ff4b3ad92b1bbd4f2f789f4f0f45b21ae08cf Mon Sep 17 00:00:00 2001
From: Joshua Klein <mobiusklein@gmail.com>
Date: Mon, 1 Dec 2025 17:57:04 -0500
Subject: [PATCH] feature: support latin1 characters for identifiers in spectra
 and chromatograms for mzML

---
 Cargo.lock                    |  1 +
 Cargo.toml                    |  3 ++-
 src/io/mzml/reader.rs         | 14 +++++++++++---
 src/io/mzml/reading_shared.rs | 10 ++++++++--
 4 files changed, 22 insertions(+), 6 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 34f84f7..012ee1c 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -2071,6 +2071,7 @@ dependencies = [
  "chrono",
  "clap",
  "criterion",
+ "encoding_rs",
  "env_logger",
  "filename",
  "flate2",
diff --git a/Cargo.toml b/Cargo.toml
index bd2162a..c5eaa9b 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -60,7 +60,7 @@ default = ["zlib-ng-compat", "mgf", "mzml"]
 checksum = ["dep:md5", "dep:sha1", "dep:base16ct"]
 
 mgf = []
-mzml = ["dep:quick-xml", "checksum", "dep:memchr"]
+mzml = ["dep:quick-xml", "checksum", "dep:memchr", "dep:encoding_rs"]
 imzml = ["mzml", "dep:uuid"]
 
 # mzsignal's main functionality requires a linear algebra backend.
@@ -201,6 +201,7 @@ pin-project-lite = { version = "0.2.16", optional = true }
 memchr = { version = "2.7.4", optional = true }
 libloading = { version = "0.8.6", optional = true }
 zstd = { version = "0.13.3", optional = true }
+encoding_rs = { version = "0.8.35", optional = true }
 
 
 [dev-dependencies]
diff --git a/src/io/mzml/reader.rs b/src/io/mzml/reader.rs
index 87ff9b3..e16a3e5 100644
--- a/src/io/mzml/reader.rs
+++ b/src/io/mzml/reader.rs
@@ -738,7 +738,12 @@ impl<C: CentroidLike + BuildFromArrayMap, D: DeconvolutedCentroidLike + BuildFro
                     match attr_parsed {
                         Ok(attr) => match attr.key.as_ref() {
                             b"id" => {
-                                self.entry_id = match attr.unescape_value() {
+                                self.entry_id = match attr.unescape_value()
+                                    .map(|v| v.to_string())
+                                    .or_else(|_| -> Result<String, quick_xml::Error> {
+                                        log::warn!("Detected non-UTF8 character in spectrum id");
+                                        Ok(quick_xml::escape::escape(encoding_rs::mem::decode_latin1(&attr.value).as_ref()).into())
+                                    }) {
                                     Ok(value) => value.to_string(),
                                     Err(e) => {
                                         return Err(xml_error!(
@@ -898,8 +903,11 @@ impl<C: CentroidLike + BuildFromArrayMap, D: DeconvolutedCentroidLike + BuildFro
                             b"id" => {
                                 self.entry_id = attr
                                     .unescape_value()
-                                    .expect("Error decoding id")
-                                    .to_string();
+                                    .map(|v| v.to_string())
+                                    .or_else(|_| -> Result<String, quick_xml::Error> {
+                                        log::warn!("Detected non-UTF8 character in chromatogram id");
+                                        Ok(quick_xml::escape::escape(encoding_rs::mem::decode_latin1(&attr.value).as_ref()).into())
+                                    }).unwrap();
                                 trace!("Stored chromatogram id = {}", self.entry_id);
                             }
                             b"index" => {
diff --git a/src/io/mzml/reading_shared.rs b/src/io/mzml/reading_shared.rs
index dd3b077..e38bad3 100644
--- a/src/io/mzml/reading_shared.rs
+++ b/src/io/mzml/reading_shared.rs
@@ -461,8 +461,14 @@ impl IndexedMzMLIndexExtractor {
                             if attr.key.as_ref() == b"idRef" {
                                 self.last_id = attr
                                     .unescape_value()
-                                    .expect("Error decoding idRef")
-                                    .to_string();
+                                    .map(|v| v.to_string())
+                                    .or_else(|_| -> Result<String, quick_xml::Error> {
+                                        log::warn!("Detected non-UTF8 character in idRef");
+                                        Ok(quick_xml::escape::escape(encoding_rs::mem::decode_latin1(&attr.value).as_ref()).into())
+                                    })
+                                    .unwrap_or_else(|e| {
+                                        panic!("Error decoding idRef on offset {e} from bytes {:?}", attr.value)
+                                    });
                             }
                         }
                         Err(err) => {