o qi@s\dZddlZddlZddlmZddlmZddlmZddl m Z m Z Gdd d eZ dS) z Sentence data structures N)Tuple)CoNLL) Conllable)TokenWordc@seZdZdZd#deeeffddZdedefdd Z dede fd d Z d#ded efd dZ ddZ eddZejddZedefddZejddZeddZejddZeddZejddZddZdefdd Zd!d"ZdS)$SentencezY A sentence class that stores attributes of a sentence and carries a list of tokens. NsentencecCs8i|_g|_g|_g|_d|_g|_||_||dS)z Construct a sentence given a dict object which contain sentence meta and tokens in the form of CoNLL-U dicts. N)_meta_tokens_words _dependencies_text_ents_doc_process)selfr docrO/home/ubuntu/.local/lib/python3.10/site-packages/urduhack/core/unit/sentence.py__init__szSentence.__init__keyreturncCs |j|S)ap Returns the value associated with the key in the metadata (comments). Args: key (str): The key whose value to look up. Returns: str: The value associated with the key as a string. If the key is a singleton then None is returned. Raises: KeyError: If the key is not present in the comments. r rrrrr meta_value s zSentence.meta_valuecCs ||jvS)a Check if the key is present as a singleton or as a pair. Args: key (str): The value to check for in the comments. Returns: bool: True if the key was provided as a singleton or as a key value pair. False otherwise. rrrrr meta_present-s zSentence.meta_presentvaluecCs||j|<dS)a# Set the metadata or comments associated with this Sentence. Args: key (str): The key for the comment. value (str): The value to associate with the key. If the comment is a singleton, this field can be ignored or set to None. Nr)rrrrrrset_meta8s zSentence.set_metac Csd}gg|_|_|\}}|D] \}}|||qt|D]C\}}tj|vr2t|d|tj<t|} |j | t | tj} | |krR|jdj | n |j t || gd|jd| _ q dS)Nr)words)tokensr itemsr enumeraterIDstrrappendintgetrparent) rr e_n sentence_metar!rrientrynew_wordidxrrrrCs   zSentence._processcC|jS)z% Access the parent doc of this span. rrrrrrXz Sentence.doccC ||_dS)z" Set the parent doc of this span. Nr1rrrrrr] cCr0)z( Access the raw text for this sentence. rr2rrrtextbr3z Sentence.textcCr4)z% Set the raw text for this sentence. Nr7r5rrrr8gr6cCr0)z. Access the list of tokens for this sentence. r r2rrrr!lr3zSentence.tokenscCr4)z+ Set the list of tokens for this sentence. Nr9r5rrrr!qr6cCr0)z- Access the list of words for this sentence. r r2rrrr vr3zSentence.wordscCr4)z* Set the list of words for this sentence. Nr:r5rrrr {r6cCs g}|jD]}||7}q|S)zV Dumps the sentence into a list of dictionary for each token in the sentence. )r!to_dict)rrettokenrrrr;s zSentence.to_dictcCs|ds|jdur|d|jg}t|jtdd}|D]'}|ddur9tj d|dd|d}n tj d|d}| |q!|j D] }| | qLd |S) z Convert the sentence to a CoNLL-U representation. Returns: str: A string representing the Sentence in CoNLL-U format. r8Nr)rr z =  )rr8rsortedr r"operator itemgetterrCOMMENT_MARKERr&r conlljoin)rlines sorted_metametalinewordrrrrDs     zSentence.conllcCstj|dddS)NF)indent ensure_ascii)jsondumpsr;r2rrr__repr__szSentence.__repr__)N)__name__ __module__ __qualname____doc__rdictlistrr%rboolrrrpropertyrsetterr8r!r r;rDrPrrrrrs4           r) rTrNrAtypingrurduhack.conllrurduhack.conll.conllablerr=rrrrrrrs