feat(xml): read an xml file as its source, not as one very long line

AGENTS.md

@@ -62,6 +62,7 @@ bytes ─▶ magic/open_strategy ─▶ DecodedFile ─▶ Document ─▶ Eleme
 | `src/odr/internal/ooxml/` | OOXML (docx/pptx/xlsx); see [`ooxml/AGENTS.md`](src/odr/internal/ooxml/AGENTS.md) + per-format docs. |
 | `src/odr/internal/oldms/` | **Legacy MS binary** (.doc/.ppt/.xls). |
 | `src/odr/internal/pdf/` | PDF (own parser). |
+| `src/odr/internal/xml/` | XML, rendered as a source view; see [`xml/AGENTS.md`](src/odr/internal/xml/AGENTS.md). |
 | `src/odr/internal/{csv,json,text,svm}/` | Smaller formats. |
 | `cli/src/` | CLI tools: `translate`, `back_translate`, `meta`, `server`. |
 | `python/` | Python bindings (`pyodr`, pybind11); see [`python/AGENTS.md`](python/AGENTS.md). |

CHANGELOG.md

@@ -16,6 +16,9 @@ The release run heads these entries with the version and opens a fresh
 
 ## Unreleased
 
+- An xml file opens as xml and reads as a foldable, highlighted source view
+  rather than as one very long line, in the encoding its declaration names.
+
 ## v6.4.0 - 2026-08-09
 
 - csv opens as a spreadsheet, its dialect probed unless the caller gives one,

CMakeLists.txt

@@ -148,6 +148,7 @@ set(ODR_SOURCE_FILES
         "src/odr/internal/html/media_file.cpp"
         "src/odr/internal/html/pdf_file.cpp"
         "src/odr/internal/html/text_file.cpp"
+        "src/odr/internal/html/xml_file.cpp"
 
         "src/odr/internal/json/json_file.cpp"
         "src/odr/internal/json/json_util.cpp"
@@ -254,6 +255,8 @@ set(ODR_SOURCE_FILES
         "src/odr/internal/util/string_util.cpp"
         "src/odr/internal/util/xml_util.cpp"
 
+        "src/odr/internal/xml/xml_file.cpp"
+
         "src/odr/internal/zip/zip_archive.cpp"
         "src/odr/internal/zip/zip_exceptions.cpp"
         "src/odr/internal/zip/zip_file.cpp"

src/odr/file.cpp

@@ -305,6 +305,10 @@ std::string TextFile::text() const {
   return internal::encoding::to_utf8(bytes, encoding);
 }
 
+std::shared_ptr<internal::abstract::TextFile> TextFile::impl() const {
+  return m_impl;
+}
+
 CsvFile CsvFile::from_file(const File &file, const CsvOptions &options,
                            const Logger &logger) {
   ODR_VERBOSE(logger, "open as csv with options");

src/odr/file.hpp

@@ -144,8 +144,8 @@ enum class FileType {
   // https://en.wikipedia.org/wiki/Windows_Metafile#Enhanced_Metafile
   enhanced_metafile,
 
-  // Classification only - reported under the formats built on it (an svg comes
-  // back as `[text_file, xml, scalable_vector_graphics]`), no decoder yet.
+  // Also reported under the formats built on it - an svg comes back as
+  // `[text_file, xml, scalable_vector_graphics]`.
   // https://en.wikipedia.org/wiki/XML
   xml,
 };
@@ -408,6 +408,8 @@ class TextFile final : public DecodedFile {
   /// decodable, and the raw bytes where it is not.
   [[nodiscard]] std::string text() const;
 
+  [[nodiscard]] std::shared_ptr<internal::abstract::TextFile> impl() const;
+
 private:
   std::shared_ptr<internal::abstract::TextFile> m_impl;
 };

src/odr/html.cpp

@@ -16,6 +16,7 @@
 #include <odr/internal/html/media_file.hpp>
 #include <odr/internal/html/pdf_file.hpp>
 #include <odr/internal/html/text_file.hpp>
+#include <odr/internal/html/xml_file.hpp>
 #include <odr/internal/util/file_util.hpp>
 
 #include <algorithm>
@@ -215,6 +216,12 @@ HtmlService html::translate(const DecodedFile &file, const HtmlConfig &config,
   if (file.is_csv_file()) {
     return translate(file.as_csv_file().document(), config, logger);
   }
+  // and before it for the same reason. Translating it as a text file by hand
+  // still writes the line list.
+  if (file.file_type() == FileType::xml) {
+    return internal::html::create_xml_service(file.as_text_file(), config,
+                                              logger);
+  }
   if (file.is_text_file()) {
     return translate(file.as_text_file(), config, logger);
   }

src/odr/internal/file_type_table.cpp

@@ -659,15 +659,14 @@ constexpr std::array table{
         DocumentType::unknown,
         {.detect_by_content = true, .open = true, .translate_html = true}},
 
-    // Detection reports it, nothing opens it yet - a plain xml file still
-    // decodes as text.
+    // Not decoded: it renders as a source view of itself.
     Row{FileType::xml,
         "xml"sv,
         xml_extensions,
         xml_mimetypes,
         FileCategory::text,
         DocumentType::unknown,
-        {.detect_by_content = true}},
+        {.detect_by_content = true, .open = true, .translate_html = true}},
 };
 
 /// Finds the row whose list, selected by @p list, contains @p needle.

src/odr/internal/html/frontend.cpp

@@ -88,6 +88,41 @@ constexpr std::string_view text_css = R"css(
 [contenteditable]:focus{outline:none}
 )css";
 
+/// No numbered gutter - the numbers would be ours, not the file's. The column
+/// carries the fold handles, and every line reserves it.
+constexpr std::string_view xml_css = R"css(
+:root{
+--odr-xml-text:#1f2328;
+--odr-xml-muted:#6e7781;
+--odr-xml-punct:#57606a;
+--odr-xml-name:#116329;
+--odr-xml-attr:#953800;
+--odr-xml-value:#0a3069;
+--odr-xml-meta:#8250df;
+--odr-xml-mono:ui-monospace,SFMono-Regular,Menlo,Consolas,monospace;
+}
+body{background:#fff}
+.odr-xml{color:var(--odr-xml-text);font:13px/1.6 var(--odr-xml-mono);word-break:break-word;overflow-wrap:anywhere}
+.odr-xml-line,.odr-xml summary{padding-left:1.5em}
+.odr-xml summary{display:block;position:relative;list-style:none;cursor:pointer}
+.odr-xml summary::-webkit-details-marker{display:none}
+/* U+25BE, U+25B8 */
+.odr-xml summary::before{content:"\25BE";position:absolute;left:.35em;color:var(--odr-xml-muted)}
+.odr-xml details:not([open])>summary::before{content:"\25B8"}
+.odr-xml summary:hover{background:rgba(0,0,0,.04)}
+/* Indentation is spaces, not padding, so a copy of the page carries it. */
+.odr-xml-indent{white-space:pre}
+.odr-xml-tag{color:var(--odr-xml-punct)}
+.odr-xml-name{color:var(--odr-xml-name)}
+.odr-xml-attr{color:var(--odr-xml-attr)}
+.odr-xml-value{color:var(--odr-xml-value)}
+/* Source text, everywhere it is written - html would fold a run of spaces. */
+.odr-xml-text,.odr-xml-cdata,.odr-xml-comment,.odr-xml-value,.odr-xml-decl,.odr-xml-doctype,.odr-xml-pi{white-space:pre-wrap}
+.odr-xml-cdata{color:var(--odr-xml-value)}
+.odr-xml-comment{color:var(--odr-xml-muted)}
+.odr-xml-decl,.odr-xml-doctype,.odr-xml-pi{color:var(--odr-xml-meta)}
+)css";
+
 constexpr std::string_view filesystem_css = R"css(
 :root{
 --odr-files-line:#e3e5e8;
@@ -892,6 +927,8 @@ constexpr Asset spreadsheet_css_asset{HtmlResourceType::css, "text/css",
                                       "spreadsheet.css", spreadsheet_css};
 constexpr Asset text_css_asset{HtmlResourceType::css, "text/css", "text.css",
                                text_css};
+constexpr Asset xml_css_asset{HtmlResourceType::css, "text/css", "xml.css",
+                              xml_css};
 constexpr Asset filesystem_css_asset{HtmlResourceType::css, "text/css",
                                      "filesystem.css", filesystem_css};
 constexpr Asset media_css_asset{HtmlResourceType::css, "text/css", "media.css",
@@ -968,6 +1005,10 @@ void html::write_text_style(const WritingState &state) {
   write_style(text_css_asset, state);
 }
 
+void html::write_xml_style(const WritingState &state) {
+  write_style(xml_css_asset, state);
+}
+
 void html::write_filesystem_style(const WritingState &state) {
   write_style(filesystem_css_asset, state);
 }
@@ -993,6 +1034,11 @@ HtmlResources html::locate_text_resources(const HtmlConfig &config) {
   return locate_all(assets, config);
 }
 
+HtmlResources html::locate_xml_resources(const HtmlConfig &config) {
+  static constexpr std::array assets{xml_css_asset};
+  return locate_all(assets, config);
+}
+
 HtmlResources html::locate_media_resources(const HtmlConfig &config) {
   static constexpr std::array assets{media_css_asset};
   return locate_all(assets, config);

src/odr/internal/html/frontend.hpp

@@ -13,6 +13,7 @@ void write_document_style(const WritingState &state);
 /// Written in addition to the document style.
 void write_spreadsheet_style(const WritingState &state);
 void write_text_style(const WritingState &state);
+void write_xml_style(const WritingState &state);
 void write_filesystem_style(const WritingState &state);
 void write_media_style(const WritingState &state);
 
@@ -27,6 +28,7 @@ void write_text_script(const WritingState &state);
 /// a service has to answer for these paths as well as for its views. Every
 /// entry is located `nullopt` when the config embeds them.
 HtmlResources locate_text_resources(const HtmlConfig &config);
+HtmlResources locate_xml_resources(const HtmlConfig &config);
 HtmlResources locate_media_resources(const HtmlConfig &config);
 
 } // namespace odr::internal::html