miniZinc plan

Author: radevgit

.gitignore

@@ -5,6 +5,7 @@
 **/*.rs.bk
 clippy.txt
 /coverage/results
+/src/zinc/flatzinc
 
 # Binary executables and object files
 *.exe

docs/development/ZINC.md

@@ -1,162 +1,22 @@
-## 2a. Survey of Existing Parsers and Design Notes
+Integration with miniZinc directory: /src/zinc
 
-- **Rust ecosystem:** No mature Rust crates exist for FlatZinc or MiniZinc parsing as of 2025. Most solvers in other languages (C++, Python, Java) implement their own parser or use the official C++ library (libminizinc).
-- **Reference implementations:**
-	- [MiniZinc/libminizinc](https://github.com/MiniZinc/libminizinc): Official C++ library with FlatZinc parser (useful for grammar/structure reference).
-	- [Chuffed](https://github.com/chuffed/chuffed): C++ solver with FlatZinc parser.
-	- [Google OR-Tools](https://github.com/google/or-tools): C++ FlatZinc parser.
-- **Design takeaways:**
-	- FlatZinc is line-oriented and regular, making it feasible to hand-write a parser in Rust.
-	- The official FlatZinc BNF grammar is a good starting point for tokenizer and parser design.
-	- Most solvers use a simple recursive-descent parser or state machine for FlatZinc.
-- **No external dependencies:** All parsing and lexing will be implemented manually in Rust, using only the standard library.
+MiniZinc can export constraint satisfaction problem to FlatZinc format (*.fzn):
+	- https://docs.minizinc.dev/en/stable/flattening.html
+	- fzn specification: https://docs.minizinc.dev/en/latest/fzn-spec.html  BNF grammar syntax at document end.
 
----
 
-## 2b. Crate Organization: Standalone vs Integrated Parser
+FlatZinc examples: https://github.com/google/or-tools/tree/stable/examples/flatzinc
+Examples explanation: https://www.hakank.org/minizinc/
+Local examples are in /src/zinc/flatzinc
 
-**Option 1: Separate Crate**
-- Pros:
-	- Parser can be reused in other projects or solvers.
-	- Clear separation of concerns; easier to test and document parser independently.
-	- Encourages clean API boundaries.
-- Cons:
-	- Slightly more maintenance overhead (versioning, publishing, documentation).
-	- May be overkill if parser is tightly coupled to Selen's internal model.
+1. Import .fzn model.
+- Not sure what is the latest spec version? The MiniZinc latest release: 2.9.4
 
-**Option 2: Integrated in Selen Crate**
-- Pros:
-	- Simpler project structure; no need for cross-crate dependencies.
-	- Easier access to Selen's internal types and APIs.
-	- Faster iteration for project-specific needs.
-- Cons:
-	- Harder to reuse parser in other projects.
-	- Parser code may become entangled with solver logic.
+2. Create AST.
 
-**Recommendation:**
-- If you anticipate reusing the FlatZinc parser in other Rust projects or want to encourage community adoption, a separate crate is preferable.
-- If the parser will be tightly integrated with Selen's internal model and not reused elsewhere, keep it as a module within this crate for simplicity.
-# MiniZinc Import: Detailed Implementation Plan
+3. Map AST to programmable Selen API.
 
-## 1. Scope and Requirements
+4. Solve problem.
 
-- **Goal:** Enable parsing and importing of MiniZinc (.mzn) model files (and optionally .dzn data files) into the Selen CSP solver, mapping them to internal model structures.
-- **Directory:** Implementation is scoped to `docs/development/` (for planning/design) and the relevant Rust source directory for code.
-- **Constraints:** No external dependencies (no crates for parsing, lexing, or MiniZinc).
+5. Output result/error in FlatZinc format
 
----
-
-
-## 2. MiniZinc and FlatZinc Standards and References
-
-- **MiniZinc Language Reference (2.8.4):**  
-	- [MiniZinc 2.8.4 Language Reference](https://www.minizinc.org/doc-2.8.4/en/index.html)
-	- [MiniZinc Grammar (BNF)](https://github.com/MiniZinc/libminizinc/blob/master/doc/grammar/minizinc.bnf)
-- **FlatZinc Specification (2.8.4):**  
-	- [FlatZinc 2.8.4 Specification](https://www.minizinc.org/doc-2.8.4/en/fzn-spec.html)
-- **File Types:**  
-	- `.mzn` — Model files (constraints, variables, parameters)
-	- `.dzn` — Data files (parameter assignments)
-- **Key Language Features:**  
-	- Variable declarations (int, bool, set, array)
-	- Constraints (global, arithmetic, logical)
-	- Parameters and data separation
-	- Solve annotations (satisfy, minimize, maximize)
-	- Comments (`% ...`)
-- **Subset Recommendation:**  
-	- Start with a subset: integer/boolean variables, basic constraints, arrays, and parameter assignment.
-
----
-
-## 3. Implementation Complexity
-
-- **Parsing:**  
-	- Must hand-write a recursive-descent parser or a simple tokenizer and parser for the MiniZinc subset.
-	- Handle comments, whitespace, identifiers, literals, arrays, and basic expressions.
-- **Mapping:**  
-	- Map MiniZinc constructs to Selen’s internal model (variables, constraints, objectives).
-- **Error Handling:**  
-	- Provide clear error messages for unsupported or malformed input.
-- **Extensibility:**  
-	- Design parser to allow future support for more MiniZinc features.
-
-**Estimated Complexity:**  
-- **Minimal Subset:** Moderate (basic parser, mapping, error handling)
-- **Full MiniZinc:** High (complex grammar, global constraints, advanced types)
-
----
-
-## 4. Implementation Plan
-
-### Step 1: Research and Design
-
-- Review MiniZinc language reference and grammar.
-- Identify the minimal viable subset to support (variables, constraints, arrays, basic arithmetic).
-- Document mapping from MiniZinc constructs to Selen’s API.
-
-### Step 2: Write a MiniZinc Tokenizer
-
-- Implement a tokenizer for MiniZinc syntax:
-	- Recognize keywords, identifiers, numbers, symbols, comments, and whitespace.
-	- Output a stream of tokens for the parser.
-
-### Step 3: Implement a Recursive-Descent Parser
-
-- Parse MiniZinc model files into an AST (abstract syntax tree).
-- Support:
-	- Variable declarations (int, bool, array)
-	- Parameter assignments
-	- Constraint statements
-	- Solve annotations (optional, for future)
-- Ignore unsupported features with clear errors.
-
-### Step 4: Map AST to Selen Model
-
-- Translate parsed MiniZinc AST into Selen’s internal model:
-	- Create variables, post constraints, set objectives.
-- Handle arrays and parameter substitution.
-
-### Step 5: Integrate and Test
-
-- Add import API (e.g., `Model::import_minizinc(path: &str) -> Result<Model, Error>`).
-- Write unit tests with sample MiniZinc files.
-- Document supported features and limitations.
-
----
-
-
-## 5. References and Resources
-
-- [MiniZinc 2.8.4 Language Reference](https://www.minizinc.org/doc-2.8.4/en/index.html)
-- [MiniZinc BNF Grammar](https://github.com/MiniZinc/libminizinc/blob/master/doc/grammar/minizinc.bnf)
-- [FlatZinc 2.8.4 Specification](https://www.minizinc.org/doc-2.8.4/en/fzn-spec.html)
-- [MiniZinc Example Models](https://www.minizinc.org/examples.html)
-- [MiniZinc Standard Library](https://www.minizinc.org/doc-2.8.4/en/lib-globals.html)
-
----
-
-## 6. No-Dependency Considerations
-
-- All parsing and lexing must be implemented manually in Rust.
-- Avoid using crates like `nom`, `pest`, or `lalrpop`.
-- Use Rust’s standard library only.
-
----
-
-## 7. Example: Minimal Supported MiniZinc
-
-```minizinc
-int: n;
-array[1..n] of var 1..n: x;
-constraint all_different(x);
-solve satisfy;
-```
-
----
-
-## 8. Future Extensions
-
-- Support for `.dzn` data files.
-- More global constraints.
-- Objective functions (minimize/maximize).
-- Full MiniZinc grammar coverage.