German translation... completed unicode

.gitignore

@@ -1,4 +1,5 @@
 .idea
+.DS_Store
 target
 build.rs.orig
 Python/libmathcat.pyd

.vscode/launch.json

@@ -11,6 +11,7 @@
             "program": "${workspaceRoot}/target/debug/mathcat.exe",
             "args": [],
             "cwd": "${workspaceRoot}",
+            "env": {"RUST_LOG": "DEBUG"},
             "preLaunchTask": "cargo build",
         },
         {
@@ -20,7 +21,8 @@
             "program": "${workspaceRoot}/target/debug/mathcat.exe",
             "args": [],
             "cwd": "${workspaceFolder}",
+            "env": {"RUST_LOG": "DEBUG"},
             "preLaunchTask": "cargo test build",
         }
-    ]
+    ],
 }

.vscode/settings.json

@@ -517,5 +517,4 @@
         "bzlib_private.h": "c",
         "stdlib.h": "c"
     },
-
 }

.vscode/tasks.json

@@ -24,8 +24,8 @@
             "type": "cargo",
             "command": "build",
             "args": [
-                "--features",
-                "include-zip" // Replace with your actual feature name
+                // "--features", "include-zip",
+                "--profile", "dev",
             ],
             "group": {
                 "kind": "build",

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "mathcat"
-version = "0.7.2-rc.1"
+version = "0.7.2-rc.3"
 authors = ["Neil Soiffer <soiffer@alum.mit.edu>"]
 license = "MIT"
 description = "MathCAT: Math Capable Assistive Technology ('Speech and braille from MathML')"

README.md

@@ -1,20 +1,19 @@
-# MathCAT: Math Capable Assistive Technology
+# MathCATde: German Translation for MathCAT: Math Capable Assistive Technology
 
 <img src="logo.png" style="position: relative; top: 16px; z-index: -1;">
 is a library that supports conversion of MathML to:
 
 
 * Speech strings (in several languages) with embedded speech engine commands
-* Braille (Nemeth, UEB Technical, CMU, and many others)
+* (later: Braille (Nemeth, UEB Technical, CMU, and many others))
 * Navigation of math (in multiple ways including overviews)
 
+This is the fork of [MathCAT](https://github.com/NSoiffer/MathCAT) of Neil Soiffer focussed on the translations to the German speaking contexts. 
 
-There are four related projects that make use of MathCAT:
-- [MathCATDemo](https://nsoiffer.github.io/MathCATDemo/) -- an online demonstration of some of what can be done with MathCAT
-- [A python interface for MathCAT](https://github.com/NSoiffer/MathCATForPython) -- used by a [MathCAT NVDA add-on](https://addons.nvda-project.org/addons/MathCAT.en.html).
-- [A C/C++ interface for MathCAT](https://github.com/NSoiffer/MathCATForC)
-- [A Java interface for MathCAT](https://github.com/mwhapples/MathCAT4J) (thanks to Michael Whapples for working on that)
+## Status
+This translation project was launched in the summer 2025 it is quite new still.
 
-MathCAT is used in many assistive technologies including NVDA and JAWS.
+## How to contribute?
+This project is untertaken by the team of IT&Tech at the [IU University of Applied Sciences](https://www.iu.org/) and welcomes test-reports in the form of issues and proposed changes in the form of pull-requests.
 
-For more information, see the [full documentation](https://nsoiffer.github.io/MathCAT/).
+For process questions, raising an issue is probably the best idea too.

Rules/Braille/UEB/unicode.yaml

@@ -453,6 +453,7 @@
  - "γ": [t: "GL⠛"]              # 0x3b3 
  - "δ": [t: "GL⠙"]              # 0x3b4 
  - "ε": [t: "GL⠑"]              # 0x3b5 
+ - "ϵ": [t: "GL⠑"]              # 0x3f5 \varepsilon
  - "ζ": [t: "GL⠵"]              # 0x3b6 
  - "η": [t: "GL⠱"]             # 0x3b7 
  - "θ": [t: "GL⠹"]              # 0x3b8 

Rules/Languages/de/unicode.yaml

@@ -467,7 +467,7 @@
  - "⊃":                                         	#  0x2283
      - test: 
          if: "$Verbosity!='Terse'"
-         then: [T: "ist ein"]                   	# 	(en: 'is a', google translation)
+         then: [T: "ist eine"]                   	# 	(en: 'is a', google translation)
      - T: "obermenge von"                 	# 	(en: 'superset of', MathPlayer: 'Obermenge von', google: 'superset von')
  - "⊄":                                         	#  0x2284
      - test: 

docs/index.md

@@ -175,7 +175,7 @@ Using open source math books, the initial load should cover at least 99.99% of t
 
 The library is about ~3mb in size.
 
-If you are working on an in-browser solution (i.e, you are using JavaScript or some other browser-based language), MathCAT is probably not the best tool for you (although I will probably factor the [MathCATDemo](https://github.com/NSoiffer/MathCATDemo/) into a Javascript interface which the demo is built on top of). Instead, take a look at [Speech rule engine](https://github.com/zorkow/speech-rule-engine) (SRE) by Volker Sorge. It is written in TypeScript and will likely meet your needs for an in-browser solution unless UEB braille is important.
+If you are working on an in-browser solution (i.e, you are using JavaScript or some other browser-based language), MathCAT is probably not the best tool for you (although I will probably factor the [MathCATDemo](https://github.com/NSoiffer/MathCATDemo/) into a Javascript interface which the demo is built on top of). Instead, take a look at [Speech rule engine](https://github.com/zorkow/speech-rule-engine) (SRE) by Volker Sorge. It is written in TypeScript and will likely meet your needs for an in-browser solution unless braille is important; MathCAT supports multiple braille codes and at least for Nemeth Code, is higher quality.
 
 # Acknowledgements
 Several people helped out in various ways with the project. I am very grateful for all their help!
@@ -192,7 +192,7 @@ Translators:
 * Finnish -- Sami Määttä, Accessibility Library Celia, and Essi Viippola, freelancer
 * German --Nazli Andjic, Robert Graf and Paul Libbrecht (IU International University of Applied Sciences)
 * Indonesian -- Dr. Pinta Deniyanti Sampoerno, M.Si; Dr. Meiliasari, S.Pd., M.Sc; and Ari Hendarno, S.Pd., M.kom
-* Norwegian -- Marthe Gjelstad, National Library of Norway
+* Norwegian -- Marthe Gjelstad, National Library of Norway, Kvile
 * Spanish -- Noelia Ruiz Martínez (also help with NVDA addon development) and María Allo Roldán
 * Swedish -- Tim Arborealis Lötberg, Swedish Agency for Accessible Media (MTM) and Anders Eklund, SPSM
 * Vietnamese -- Dang Hoai Phúc and Trang Pham

docs/nav-commands.md

@@ -301,22 +301,20 @@ Note: while navigating an expression, "control+c" copies the math content of the
 
 ## Navigation Modes
 
-<b>Enhanced mode</b>: &nbsp;navigation is by mathematically
-meaningful pieces (operators, delimiters, and operands)
+MathCAT supports three different navigation modes: enhanced, simple, and character. The first two modes of navigation follow the semantics of what was read for the entire expression except if LiteralSpeech is selected. For example $|x+y| > 0$ will not read the vertical lines that are used for the absolute value notation, but instead will say "absolute value". Zooming in will move directly saying "x plus y". In contrast, character mode will read this as "vertical line", "x", "plus", "y" "vertical line", "is greater than", "zero" as you move through the expression.
 
-
-<b>Simple mode</b>: this moves by words except when you get to a
-2D notation (fractions, roots, ...), then it speaks the entire notation.
-Zooming in lets you explore the 2D notation in the same mode. Zooming out or
-moving out of the 2D notation brings you back to the outer/higher level of
-navigation.
-
-
-<b>Character mode</b>: &nbsp;this is actually two useful modes --
-word mode and character mode (zoom in to get &quot;real&quot; character mode).
-&nbsp;Moves by words/characters. &nbsp;This differs for numbers of more than
-one digit and function names such as &quot;sin&quot; that are multiple
-characters. Otherwise, word and character navigation is the same.
+* _Enhanced mode_: navigation is by mathematically meaningful pieces (operators, delimiters, and operands)
+* _Simple mode_: this moves by words except when you get to a
+    2D notation (fractions, roots, ...), then it speaks the entire notation.
+    Zooming in lets you explore the 2D notation in the same mode. Zooming out or
+    moving out of the 2D notation brings you back to the outer/higher level of
+    navigation.
+* _Character mode_: this is actually two useful modes --
+    word mode and character mode (zoom in to get &quot;real&quot; character mode).
+    &nbsp;Moves by words/characters. &nbsp;This differs for numbers of more than
+    one digit and function names such as &quot;sin&quot; that are multiple
+    characters. Otherwise, word and character navigation is the same. Both will
+    automatically zoom into fractions, etc.
 
 ## Typical Use
 

src/canonicalize.rs

@@ -3161,7 +3161,7 @@ impl CanonicalizeContext {
 			// FIX: MathType generates the wrong version of union and intersection ops (binary instead of unary)
 		} else if !is_base && (parent_name == "msup" || parent_name == "msubsup") {
 			mo_text = match mo_text {
-				"\u{00BA}"| "\u{2092}"| "\u{20D8}"| "\u{2218}" => "\u{00B0}",		// circle-like objects -> degree
+				"\u{00BA}"| "\u{2092}"| "\u{20D8}"| "\u{2218}" | "\u{25E6}" => "\u{00B0}",		// circle-like objects -> degree
 				_ => mo_text,
 			};
 		} else {

src/definitions.rs

@@ -29,8 +29,9 @@ use yaml_rust::yaml::Hash;
 use yaml_rust::Yaml;
 use crate::errors::*;
 use crate::prefs::*;
-use std::{cell::RefCell, cell::Ref, cell::RefMut, collections::HashSet,  rc::Rc};
-use std::{collections::HashMap, path::Path, path::PathBuf};
+use std::{cell::RefCell, cell::Ref, cell::RefMut, rc::Rc};
+use std::path::{Path, PathBuf};
+use std::collections::{HashMap, HashSet};
 use crate::shim_filesystem::read_to_string_shim;
 
 /// An enum to paper over the different types of data access needed.

src/interface.rs

@@ -539,6 +539,7 @@ pub fn get_supported_languages() -> Vec<String> {
         file_name.truncate(file_name.len() - "_Rules.yaml".len())
     }
     speech_styles.sort();
+    // remove duplicates -- shouldn't be any, but just in case
     let mut i = 1;
     while i < speech_styles.len() {
         if speech_styles[i-1] == speech_styles[i] {

src/main.rs

@@ -18,6 +18,8 @@ cfg_if::cfg_if! {
         }
     } else {
         fn get_rules_dir() -> String {
+          // for testing with zipped rules dir
+          // let rules_path = std::env::current_exe().unwrap().parent().unwrap().join("../../../MathCATForPython/addon/globalPlugins/MathCAT/Rules");
           let rules_path = std::env::current_exe().unwrap().parent().unwrap().join("../../Rules");
           return rules_path.as_os_str().to_str().unwrap().to_string();
         }
@@ -192,17 +194,18 @@ fn main() {
   let expr = r#"
 <math xmlns="http://www.w3.org/1998/Math/MathML"><mo>(</mo><mn>1</mn><mo>)</mo></math>
                    "#;
-  let instant = Instant::now();
+  // let instant = Instant::now();
 
   // let rules_dir = "".to_string();    // Use MathCATRulesDir, potentially pointing to a zipped version
   if let Err(e) = set_rules_dir(get_rules_dir()) {
     panic!("Error: exiting -- {}", errors_to_string(&e));
   }
+  debug!("Languages: {}", libmathcat::interface::get_supported_languages().join(", "));
 
   #[cfg(feature = "include-zip")]
   info!("***********include-zip is present**********");
   info!("Version = '{}' using Rules dir {}", get_version(), get_rules_dir());
-  set_preference("Language".to_string(), "en-gb".to_string()).unwrap();
+  set_preference("Language".to_string(), "en".to_string()).unwrap();
   set_preference("DecimalSeparator".to_string(), "Auto".to_string()).unwrap();
   set_preference("BrailleCode".to_string(), "Nemeth".to_string()).unwrap();
   set_preference("TTS".to_string(), "None".to_string()).unwrap();
@@ -291,7 +294,7 @@ fn main() {
   debug!("...using BrailleCode: {:?}", get_preference("BrailleCode".to_string()).unwrap());
   // let xpath_counts = libmathcat::speech::xpath_count();
   // info!("#xpath = {}; duplicates = {}", xpath_counts.0, xpath_counts.1);
-  info!("Time taken (second time for speech + braille): {}ms", instant.elapsed().as_millis());
+  // info!("Time taken (second time for speech + braille): {}ms", instant.elapsed().as_millis());
   // debug!("Hashmap sizes:\n{}", libmathcat::speech::SpeechRules::print_sizes());
   timing_test(expr, 0);
 

src/prefs.rs

@@ -398,7 +398,8 @@ impl PreferenceManager {
     fn set_style_file(&mut self, language_dir: &Path, language: &str, style_file_name: &str) -> Result<()> {
         let style_file_name = style_file_name.to_string() + "_Rules.yaml";
         self.speech = PreferenceManager::find_file(language_dir, language, Some("en"), &style_file_name)?;
-
+        // debug!("set_style_file: language_dir: {}, language: {}, style_file_name: {}, self.speech: {}",
+        //        language_dir.display(), language, style_file_name, self.speech.display());
         return Ok( () );
     }
 
@@ -477,6 +478,16 @@ impl PreferenceManager {
                     PreferenceManager::unzip_files(path, language, default_lang)
                                                 .chain_err(|| format!("Couldn't open zip file {zip_file_string} in parent {language}: {e}."))?
                 } else {
+                    // maybe just regional dialects
+                    let mut regional_dirs = Vec::new();
+                    find_all_dirs_shim(&dir, &mut regional_dirs);
+                    for dir in regional_dirs {
+                        // debug!("unzip_files: trying again in subdir: {}", dir.display());
+                        let language = format!("{}-{}", lang, dir.file_name().unwrap().to_str().unwrap());
+                        if let Ok(result) =PreferenceManager::unzip_files(path, &language, default_lang) {
+                            return Ok(result);
+                        }
+                    }
                     bail!("Couldn't open zip file {}: {}.", zip_file_string, e)
                 }
             },
@@ -569,6 +580,7 @@ impl PreferenceManager {
             };
             if looking_for_style_file && alternative_style_file.is_none() {
                 if let Ok(alt_file_path) = find_any_style_file(os_path) {
+                    // debug!("find_file: found alternative style file '{}'", alt_file_path.display());
                     alternative_style_file = Some(alt_file_path);
                 }
             }
@@ -578,11 +590,26 @@ impl PreferenceManager {
             }
         }
 
+
         if let Some(result) = alternative_style_file {
             // debug!("find_file: found alternative_style_file '{}'", result.to_string_lossy());
             return Ok(result);     // found an alternative style file in the same lang dir
         }
 
+        // try a subdir (regional dialect) of the language dir
+        let mut regional_dirs = Vec::new();
+        find_all_dirs_shim(&lang_dir, &mut regional_dirs);
+        for dir in regional_dirs {
+            // debug!("find_file: trying again in subdir: {}", dir.display());
+            // debug!(" ... files found = {:?}", find_files_in_dir_that_ends_with_shim(&dir, file_name));
+            if find_files_in_dir_that_ends_with_shim(&dir, ".yaml").contains(&file_name.to_string()) {
+                let path = dir.join(file_name);
+                if is_file_shim(&path) {
+                    return Ok(path);
+                }
+            }
+        }
+
         if let Some(default_lang) = default_lang {
             // try again with the default language (we're likely in trouble)
             return PreferenceManager::find_file(rules_dir, default_lang, None, file_name);
@@ -594,7 +621,8 @@ impl PreferenceManager {
                Looking for file: {}",
             rules_dir.to_str().unwrap(), lang, file_name);
 
-        // try to find a xxx_Rules.yaml file -- returns an error if none is found ()
+
+        /// try to find a xxx_Rules.yaml file -- returns an error if none is found ()
         fn find_any_style_file(path: &Path) -> Result<PathBuf> {    
             // try to find a xxx_Rules.yaml file
             // we find the first file because this is the deepest (most language specific) speech rule file