Ng d dl mZ d dlZd dlmZ d dlmZ d dlmZm Z m
Z
m Z ej e eed g d Z
d Zd
Zd ZdS ) )partialN)chunk_elements)chunk_by_title)ElementMetadata
NarrativeTextTextTitle)combine_text_under_n_chars)paramsc | j S )N)param)requests g/var/www/html/ai-engine/env/lib/python3.11/site-packages/test_unstructured/chunking/test_html_output.py chunking_fnr
s
= c j d}d}d}d |||g }t dt | t dt | t d t | g} | | }t | d
k sJ |d j j |k sJ d S ) N%
Header
z@z Header text_as_htmltextmetadatazDate: October 30, 2023z Form field name Example value r )joinr r r lenr r )r
metadata_1
metadata_2
metadata_3combined_metadataelementschunkss r Htest_combining_html_metadata_when_multiple_elements_in_composite_elementr$ s 8JSJ *j*!EFF 8o:&N&N&NOOO *_R\5]5]5]^^^ 0?Xb;c;c;c
H [
"
"F v;;! !9 *.? ? ? ? ? ? ?r c d}d}d}t dt | t dt |d t d
t |d g} | |d }t | d
k sJ |d j dk sJ |d j d
k sJ |d j j |dz |z k sJ |d j j |k sJ dS )aH
Ground truth
r r First1)r parent_idSecond max_characters r r r N)r r r r r r r )r r r r r" r# s r Ftest_combining_html_metadata_with_nested_relationship_between_elementsr/ & s$ $ 2J9J:J
"J G G GHHH?
VY#Z#Z#Z
OWZ$[$[$[
H [!
4
4
4F v;;! !9>W $ $ $ $ !9>X % % % % !9 *j3.>.K K K K K !9 *j 8 8 8 8 8 8r c : d}t dt | g} | |d }t | dk sJ |d j d k sJ |d
j d k sJ |d j j dk sJ |d
j j dk sJ d S )
z2Mimic behaviour of elements with non-html metadatar r r r r, r. r Hear derN)r r r r r r )r r r" r# s r ;test_html_metadata_exist_in_both_element_when_text_is_splitr4 N s 8J
8o:&N&N&NOOOH [!
4
4
4F v;;! !9>U " " " " !9>U " " " " !9 *.U U U U U !9 *.U U U U U U Ur ) functoolsr pytestunstructured.chunking.basicr unstructured.chunking.titler unstructured.documents.elementsr r r r fixturer r$ r/ r4 r r r< s
6 6 6 6 6 6 6 6 6 6 6 6 W W W W W W W W W W W W [\(](](]^___ `_@ @ @.%9 %9 %9P V V V V Vr