o )iZ!@sdZddlmZmZmZmZmZmZmZm Z ddl m Z ddl m Z mZddlmZmZmZddlmZGdd d e ZGd d d e Zd S) z6Compiling grammar for efficient token mask generation.)AnyDictListOptionalTupleTypeUnionoverload) BaseModel) XGRObject_core)GrammarStructuralTagItem_convert_schema_to_str) TokenizerInfoc@sneZdZdZedefddZedefddZede fddZ de fd d Z e d e d eddfd dZdS)CompiledGrammara#This is the primary object to store compiled grammar. A CompiledGrammar can be used to construct GrammarMatcher to generate token masks efficiently. Notes ----- Do not construct this class directly, instead use :class:`GrammarCompiler` to construct the object. returncCt|jjS)zThe original grammar.)r_create_from_handle_handlegrammarselfr]/home/app/PaddleOCR-VL-test/.venv_paddleocr/lib/python3.10/site-packages/xgrammar/compiler.pyrzCompiledGrammar.grammarcCr)z8The tokenizer info associated with the compiled grammar.)rrrtokenizer_inforrrrrrzCompiledGrammar.tokenizer_infocC|jjS)z>The approximate memory usage of the compiled grammar in bytes.)rmemory_size_bytesrrrrr"sz!CompiledGrammar.memory_size_bytescC |jS)z0Serialize the compiled grammar to a JSON string.)rserialize_jsonrrrrr!' zCompiledGrammar.serialize_jsonjson_strrcCsttj||jS)zGDeserialize the compiled grammar from a JSON string and tokenizer info.)rrr deserialize_jsonr)r#rrrrr$+sz CompiledGrammar.deserialize_jsonN)__name__ __module__ __qualname____doc__propertyrrrrintrstrr! staticmethodr$rrrrr s rc @sJeZdZdZdddddededed efd d Zdd d dd dee e e e e e ffdedeedeee e fdedef ddZdefddZde defddZdeedee defddZedd d!e d"e defd#d$Zed%edefd&d$Zdd d%ee efd"e defd'd$Zd.d(d)Zdefd*d+Zedefd,d-Zd S)/GrammarCompilera:The compiler for grammars. It is associated with a certain tokenizer info, and compiles grammars into CompiledGrammar with the tokenizer info. It allows parallel compilation with multiple threads, and has a cache to store the compilation result, avoiding compiling the same grammar multiple times. T) max_threads cache_enabledcache_limit_bytesrr0r1r2cCs.t|ts td|t|j|||dS)aConstruct the compiler. Parameters ---------- tokenizer_info : TokenizerInfo The tokenizer info. max_threads : int, default: 8 The maximum number of threads used to compile the grammar. cache_enabled : bool, default: True Whether to enable the cache. cache_limit_bytes : int, default: -1 The maximum memory usage for the cache in the specified unit. Note that the actual memory usage may slightly exceed this value. zSPlease convert the tokenizer to TokenizerInfo before passing it to GrammarCompiler.N) isinstancer ValueErrorZ _init_handler r-r)rrr0r1r2rrr__init__:s  zGrammarCompiler.__init__N)any_whitespaceindent separators strict_modeschemar6r7r8r9rc Cs"t|}t|j|||||S)aGet CompiledGrammar from the specified JSON schema and format. The indent and separators parameters follow the same convention as in json.dumps(). Parameters ---------- schema : Union[str, Type[BaseModel], Dict[str, Any]] The schema string or Pydantic model or JSON schema dict. indent : Optional[int], default: None The number of spaces for indentation. If None, the output will be in one line. separators : Optional[Tuple[str, str]], default: None Two separators used in the schema: comma and colon. Examples: (",", ":"), (", ", ": "). If None, the default separators will be used: (",", ": ") when the indent is not None, and (", ", ": ") otherwise. strict_mode : bool, default: True Whether to use strict mode. In strict mode, the generated grammar will not allow properties and items that is not specified in the schema. This is equivalent to setting unevaluatedProperties and unevaluatedItems to false. This helps LLM to generate accurate output in the grammar-guided generation with JSON schema. Returns ------- compiled_grammar : CompiledGrammar The compiled grammar. )rrrrcompile_json_schema)rr:r6r7r8r9Z schema_strrrrr;_s & z#GrammarCompiler.compile_json_schemacCst|jS)zGet CompiledGrammar from the standard JSON. Returns ------- compiled_grammar : CompiledGrammar The compiled grammar. )rrrcompile_builtin_json_grammarrrrrr<sz,GrammarCompiler.compile_builtin_json_grammarregexcCst|j|S)zGet CompiledGrammar from the specified regex. Parameters ---------- regex : str The regex string. Returns ------- compiled_grammar : CompiledGrammar The compiled grammar. )rrr compile_regex)rr=rrrr>s zGrammarCompiler.compile_regextagstriggerscCs"dd|D}t|j||S)aCompile a grammar from structural tags. See Grammar.from_structural_tag() for more details. Parameters ---------- tags : List[StructuralTagItem] The structural tags. triggers : List[str] The triggers. Returns ------- compiled_grammar : CompiledGrammar The compiled grammar. cSs g|] }|jt|j|jfqSr)beginrZschema_end).0tagrrr s z:GrammarCompiler.compile_structural_tag..)rrrcompile_structural_tag)rr?r@Z tags_tuplerrrrFs z&GrammarCompiler.compile_structural_tagrootroot_rule_name ebnf_stringrIcCdSNr)rrJrIrrrcompile_grammarszGrammarCompiler.compile_grammarrcCrKrLr)rrrrrrMscCs,t|tr tj||d}t|j|jS)amCompile a grammar object. Overloads: 1. ``compile_grammar(ebnf_string: str, *, root_rule_name: str = "root") -> CompiledGrammar`` - Compile a grammar from an EBNF string. The string should follow the format described in https://github.com/ggerganov/llama.cpp/blob/master/grammars/README.md. 2. ``compile_grammar(grammar: Grammar) -> CompiledGrammar`` - Compile a grammar from a Grammar object. Parameters ---------- ebnf_string : str The grammar string in EBNF format. root_rule_name : str, default: "root" The name of the root rule in the grammar. grammar : Union[str, Grammar] The grammar string or Grammar object. Returns ------- compiled_grammar : CompiledGrammar The compiled grammar. rH)r3r+rZ from_ebnfrrrrM)rrrIrrrrMs cCs|jdS)z#Clear all cached compiled grammars.N)r clear_cacherrrrrNszGrammarCompiler.clear_cachecCr )z3The approximate memory usage of the cache in bytes.)rget_cache_size_bytesrrrrrOr"z$GrammarCompiler.get_cache_size_bytescCr)zw The maximum memory usage for the cache in bytes. Returns -1 if the cache has no memory limit. )rr2rrrrr2sz!GrammarCompiler.cache_limit_bytes)rN)r%r&r'r(rr*boolr5rr+rr rrrrrr;r<r>rrrFr rMrrNrOr)r2rrrrr-3s~  ) -     r-N)r(typingrrrrrrrr Zpydanticr baser r rrrrrrrr-rrrrs(  '