jinja2 - Jinja is a template engine written in pure Python. It provides a non-XML syntax that supports inline expressions and an optional sandboxed environment._identifier - Undocumentedasync_utils - Undocumentedbccache - The optional bytecode cache system. This is useful if you have very complex template situations and the compilation of all those templates slows down your application too much.compiler - Compiles nodes from the parser into Python code.constants - Undocumenteddebug - No module docstring; 3/4 functions documenteddefaults - Undocumentedenvironment - Classes for managing templates and their runtime and compile time options.exceptions - No module docstring; 9/9 classes documentedext - Extension API for adding custom tags and behavior.filters - Built-in template filters used with the | operator.idtracking - No module docstring; 0/4 constant, 0/2 function, 1/3 class documentedlexer - No summaryloaders - API and implementations for loading templates from different data sources.meta - Functions that expose information about templates that might be interesting for introspection.nativetypes - No module docstring; 1/1 function, 2/3 classes documentednodes - AST nodes generated by the parser for the compiler. Also provides some node tree helper functions used by the parser and compiler in order to normalize nodes.optimizer - The optimizer tries to constant fold expressions and modify the AST in place so that it should be faster to evaluate.parser - Parse tokens from the lexer into nodes for the compiler.runtime - The runtime functions and state used by compiled templates.sandbox - A sandbox layer that ensures unsafe operations cannot be performed. Useful when the template itself comes from an untrusted source.tests - Built-in template tests used with the is operator.utils - No module docstring; 0/5 variable, 0/1 constant, 19/21 functions, 4/6 classes documentedvisitor - API for traversing the AST nodes. Implemented by the compiler and meta introspection.