Support-old-funz-varform-syntax #59

yannrichet-asnr · 2026-01-24T21:59:22Z

No description provided.

Copilot

Pull request overview

This pull request adds support for the original Java Funz variable and formula syntax to maintain compatibility with existing Java Funz workflows. The PR introduces several significant features:

Changes:

Adds support for () delimiters for variables (e.g., $(var)) alongside the existing {} syntax (e.g., ${var})
Implements static object definitions for constants and functions using #@: prefix syntax
Adds formula evaluation with format specifiers (e.g., @{expr | 0.00})
Introduces multiple model key aliases for backward compatibility (e.g., var_prefix/varprefix, formula_prefix/formulaprefix, etc.)
Modifies fzi() to return a comprehensive dictionary including variables, formulas, and static objects with their values/expressions

Reviewed changes

Copilot reviewed 12 out of 12 changed files in this pull request and generated 14 comments.

Show a summary per file

File	Description
tests/test_static_objects.py	Comprehensive tests for static object definitions (constants and functions) with various interpreters
tests/test_no_variables.py	Minor variable naming changes (variables→result) for clarity
tests/test_model_key_aliases.py	Tests for all model key aliases (var_prefix, formula_prefix, comment)
tests/test_java_funz_compatibility.py	Tests verifying Java Funz syntax compatibility
tests/test_fzi_formulas.py	Tests for formula parsing and evaluation with default values
tests/test_cli_commands.py	Minor comment addition clarifying JSON output structure
fz/interpreter.py	Core implementation of parsing logic, formula evaluation, and static object handling
fz/helpers.py	Updates to use new model key aliases with backward compatibility
fz/core.py	Major changes to `fzi()` function and addition of helper functions for model key resolution
examples/java_funz_syntax_example.py	Example demonstrating Java Funz syntax usage
examples/fzi_static_objects_example.py	Example showing static object definitions
examples/fzi_formulas_example.py	Example demonstrating formula parsing and evaluation

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-01-24T22:04:22Z

fz/core.py

+def _get_comment_char(model: Dict) -> str:
+    """
+    Get comment character from model with support for multiple aliases
+
+    Supported keys (in order of precedence):
+    - commentline
+    - comment_line  
+    - comment_char
+    - commentchar
+    - comment
+
+    Args:
+        model: Model definition dict
+
+    Returns:
+        Comment character (default "#")
+    """
+    return model.get(
+        "commentline",
+        model.get(
+            "comment_line",
+            model.get(
+                "comment_char",
+                model.get(
+                    "commentchar",
+                    model.get("comment", "#")
+                )
+            )
+        )
+    )
+
+
+def _get_var_prefix(model: Dict) -> str:
+    """
+    Get variable prefix from model with support for multiple aliases
+
+    Supported keys (in order of precedence):
+    - var_prefix
+    - varprefix
+    - var_char
+    - varchar
+
+    Args:
+        model: Model definition dict
+
+    Returns:
+        Variable prefix (default "$")
+    """
+    return model.get(
+        "var_prefix",
+        model.get(
+            "varprefix",
+            model.get(
+                "var_char",
+                model.get("varchar", "$")
+            )
+        )
+    )
+
+
+def _get_formula_prefix(model: Dict) -> str:
+    """
+    Get formula prefix from model with support for multiple aliases
+
+    Supported keys (in order of precedence):
+    - formula_prefix
+    - formulaprefix
+    - form_prefix
+    - formprefix
+    - formula_char
+    - form_char
+
+    Args:
+        model: Model definition dict
+
+    Returns:
+        Formula prefix (default "@")
+    """
+    return model.get(
+        "formula_prefix",
+        model.get(
+            "formulaprefix",
+            model.get(
+                "form_prefix",
+                model.get(
+                    "formprefix",
+                    model.get(
+                        "formula_char",
+                        model.get("form_char", "@")
+                    )
+                )
+            )
+        )
+    )


These three helper functions (_get_comment_char, _get_var_prefix, _get_formula_prefix) are duplicated from fz/interpreter.py. This violates the DRY (Don't Repeat Yourself) principle and creates maintenance issues. If the logic needs to change, it must be updated in both places. Consider importing these functions from interpreter.py instead of redefining them here.

Copilot · 2026-01-24T22:04:22Z

fz/interpreter.py

+        if '=' in line and not line.startswith('def ') and '<-' not in line:
+            parts = line.split('=', 1)
+            if len(parts) == 2:
+                name = parts[0].strip()
+                # Remove any type hints
+                if ':' in name:
+                    name = name.split(':')[0].strip()
+                value = parts[1].strip()
+                expressions[name] = value
+                i += 1
+                continue


This check for Python assignment may incorrectly match comparison operators. For example, a line like "if x == value:" contains '=' but should not be treated as an assignment. The check should use a more robust pattern, such as checking for '==' or other comparison operators first, or using a proper AST parser. Consider using regex to match assignment patterns more precisely: r'^\s*([a-zA-Z_][a-zA-Z0-9_])\s=\s*(?!=)' to avoid matching '==', '<=', '>=', '!=', etc.

fz/core.py

Copilot · 2026-01-24T22:04:23Z

fz/core.py

+                    # Try numeric conversion
+                    if '.' in default_value or 'e' in default_value.lower():
+                        variable_defaults[var_name] = float(default_value)
+                    else:
+                        variable_defaults[var_name] = int(default_value)
+                except ValueError:


The numeric conversion logic may fail for certain valid numeric formats. For example, hexadecimal numbers (0x1F), octal numbers (0o77), binary numbers (0b1010), or numbers with underscores (1_000_000) are valid in Python but won't be handled correctly here. Consider using ast.literal_eval() for more robust parsing, which can handle various Python literal formats safely.

Suggested change

# Try numeric conversion

if '.' in default_value or 'e' in default_value.lower():

variable_defaults[var_name] = float(default_value)

else:

variable_defaults[var_name] = int(default_value)

except ValueError:

# Try numeric conversion using ast.literal_eval to support

# various Python numeric literal formats (hex, octal, binary,

# underscores, etc.).

evaluated = ast.literal_eval(default_value)

if isinstance(evaluated, (int, float)):

variable_defaults[var_name] = evaluated

else:

# Not a numeric literal; fall through to string handling

raise ValueError

except (ValueError, SyntaxError):

Copilot · 2026-01-24T22:04:23Z

fz/interpreter.py

+        dedented_code = textwrap.dedent(full_code)
+
+        try:
+            exec(dedented_code, env)


Using exec() to evaluate arbitrary user code from input files poses a significant security risk. If an attacker can control the input file content, they can execute arbitrary Python code on the system. While this may be acceptable for trusted input files in certain use cases, consider adding a warning in the documentation about this security implication, or implementing a sandboxed execution environment. At minimum, document that input files should only come from trusted sources.

Copilot · 2026-01-24T22:04:24Z

fz/interpreter.py

+    # Extract context lines from model if available
+    context_lines = []
+


Variable context_lines is not used.

Suggested change

# Extract context lines from model if available

context_lines = []

Copilot · 2026-01-24T22:04:24Z

fz/core.py


 from .logging import log_error, log_warning, log_info, log_debug, log_progress
-from .config import get_config
+from .config import get_config, get_interpreter


Import of 'get_config' is not used.

Suggested change

from .config import get_config, get_interpreter

from .config import get_interpreter

tests/test_java_funz_compatibility.py

fz/interpreter.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

yannrichet added 2 commits January 24, 2026 22:49

working on true old funz param file

0269c74

better fzi output

e875073

Copilot AI review requested due to automatic review settings January 24, 2026 21:59

Copilot started reviewing on behalf of yannrichet-asnr January 24, 2026 21:59 View session

Copilot AI reviewed Jan 24, 2026

View reviewed changes

Update fz/interpreter.py

9ac1a59

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

yannrichet marked this pull request as draft January 25, 2026 09:37

yannrichet and others added 8 commits January 25, 2026 10:37

Update tests/test_java_funz_compatibility.py

690d8e8

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update fz/core.py

90c16b4

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update tests/test_java_funz_compatibility.py

392664b

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

try fix none/nan testsoutput (depending on availbale deps)

38e47b9

factorize model dict keys equiv.

56a068f

fix slurm ?

3de01be

better support var assign

34c0324

better handle num format

d676803

yannrichet marked this pull request as ready for review January 25, 2026 17:13

yannrichet merged commit 0daf661 into main Jan 25, 2026
22 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Support-old-funz-varform-syntax #59

Support-old-funz-varform-syntax #59

Uh oh!

yannrichet-asnr commented Jan 24, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Jan 24, 2026

Uh oh!

Copilot AI Jan 24, 2026

Uh oh!

Uh oh!

Copilot AI Jan 24, 2026

Uh oh!

Copilot AI Jan 24, 2026

Uh oh!

Copilot AI Jan 24, 2026

Uh oh!

Copilot AI Jan 24, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

-                    # Try numeric conversion
-                    if '.' in default_value or 'e' in default_value.lower():
-                        variable_defaults[var_name] = float(default_value)
-                    else:
-                        variable_defaults[var_name] = int(default_value)
-                except ValueError:
+                    # Try numeric conversion using ast.literal_eval to support
+                    # various Python numeric literal formats (hex, octal, binary,
+                    # underscores, etc.).
+                    evaluated = ast.literal_eval(default_value)
+                    if isinstance(evaluated, (int, float)):
+                        variable_defaults[var_name] = evaluated
+                    else:
+                        # Not a numeric literal; fall through to string handling
+                        raise ValueError
+                except (ValueError, SyntaxError):

		# Extract context lines from model if available
		context_lines = []

	from .config import get_config, get_interpreter
	from .config import get_interpreter

Support-old-funz-varform-syntax #59

Support-old-funz-varform-syntax #59

Uh oh!

Conversation

yannrichet-asnr commented Jan 24, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Jan 24, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 24, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI Jan 24, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 24, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 24, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 24, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants