in progress, need search engine
Browse files- app.py +53 -0
- models.json +149 -0
- packages.txt +15 -0
- requirements.txt +4 -0
app.py
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import re
|
| 2 |
+
import streamlit as st
|
| 3 |
+
import textract
|
| 4 |
+
import tempfile
|
| 5 |
+
import random
|
| 6 |
+
from pathlib import Path
|
| 7 |
+
import spacy
|
| 8 |
+
from spacy.tokens import DocBin
|
| 9 |
+
import srsly
|
| 10 |
+
from spacy.matcher import Matcher
|
| 11 |
+
|
| 12 |
+
st.title('Index and Search a Collection of Documents')
|
| 13 |
+
|
| 14 |
+
@st.cache
|
| 15 |
+
def download_model(select_model:str):
|
| 16 |
+
with st.spinner(f'Loading model {select_model}'):
|
| 17 |
+
spacy.cli.download(select_model)
|
| 18 |
+
return True
|
| 19 |
+
|
| 20 |
+
doc_bin = DocBin()
|
| 21 |
+
models = srsly.read_json('models.json')
|
| 22 |
+
models[''] = [] #require the user to choose a language
|
| 23 |
+
languages = models.keys()
|
| 24 |
+
language = st.selectbox("Language", languages, index=len(models.keys())-1, help="Select the language of your materials.")
|
| 25 |
+
if language:
|
| 26 |
+
select_model = st.selectbox("Model", models[language], help="spaCy model")
|
| 27 |
+
if select_model:
|
| 28 |
+
model_downloaded = download_model(select_model)
|
| 29 |
+
|
| 30 |
+
if model_downloaded:
|
| 31 |
+
|
| 32 |
+
nlp = spacy.load(select_model)
|
| 33 |
+
|
| 34 |
+
nlp.max_length = 1200000
|
| 35 |
+
|
| 36 |
+
uploaded_files = st.file_uploader("Select files to process", accept_multiple_files=True)
|
| 37 |
+
|
| 38 |
+
for uploaded_file in uploaded_files:
|
| 39 |
+
file_type = uploaded_file.type
|
| 40 |
+
|
| 41 |
+
temp = tempfile.NamedTemporaryFile()
|
| 42 |
+
temp.write(uploaded_file.getvalue())
|
| 43 |
+
try:
|
| 44 |
+
text = textract.process(temp.name)
|
| 45 |
+
text = text.decode('utf-8')
|
| 46 |
+
doc = nlp(text)
|
| 47 |
+
st.write(text)
|
| 48 |
+
except Exception as e:
|
| 49 |
+
st.error(e)
|
| 50 |
+
|
| 51 |
+
|
| 52 |
+
#st.download_button('Download', '', 'text/plain')
|
| 53 |
+
|
models.json
ADDED
|
@@ -0,0 +1,149 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"Catalan": [
|
| 3 |
+
"",
|
| 4 |
+
"ca_core_news_sm",
|
| 5 |
+
"ca_core_news_md",
|
| 6 |
+
"ca_core_news_lg",
|
| 7 |
+
"ca_core_news_trf"
|
| 8 |
+
],
|
| 9 |
+
"Chinese":[
|
| 10 |
+
"",
|
| 11 |
+
"zh_core_web_sm",
|
| 12 |
+
"zh_core_web_md",
|
| 13 |
+
"zh_core_web_lg",
|
| 14 |
+
"zh_core_web_trf"
|
| 15 |
+
],
|
| 16 |
+
"Croatian": [
|
| 17 |
+
"",
|
| 18 |
+
"hr_core_news_sm",
|
| 19 |
+
"hr_core_news_md",
|
| 20 |
+
"hr_core_news_lg"
|
| 21 |
+
],
|
| 22 |
+
"Danish": [
|
| 23 |
+
"",
|
| 24 |
+
"da_core_news_sm",
|
| 25 |
+
"da_core_news_md",
|
| 26 |
+
"da_core_news_lg",
|
| 27 |
+
"da_core_news_trf"
|
| 28 |
+
],
|
| 29 |
+
"Dutch":[
|
| 30 |
+
"",
|
| 31 |
+
"nl_core_news_sm",
|
| 32 |
+
"nl_core_news_md",
|
| 33 |
+
"nl_core_news_lg"
|
| 34 |
+
],
|
| 35 |
+
"English": [
|
| 36 |
+
"",
|
| 37 |
+
"en_core_web_sm",
|
| 38 |
+
"en_core_web_md",
|
| 39 |
+
"en_core_web_lg",
|
| 40 |
+
"en_core_web_trf"
|
| 41 |
+
],
|
| 42 |
+
"Finnish":[
|
| 43 |
+
"",
|
| 44 |
+
"fi_core_news_sm",
|
| 45 |
+
"fi_core_news_md",
|
| 46 |
+
"fi_core_news_lg"
|
| 47 |
+
],
|
| 48 |
+
"French":[
|
| 49 |
+
"",
|
| 50 |
+
"fr_core_news_sm",
|
| 51 |
+
"fr_core_news_md",
|
| 52 |
+
"fr_core_news_lg",
|
| 53 |
+
"fr_dep_news_trf"
|
| 54 |
+
],
|
| 55 |
+
"German": [
|
| 56 |
+
"",
|
| 57 |
+
"de_core_news_sm",
|
| 58 |
+
"de_core_news_md",
|
| 59 |
+
"de_core_news_lg",
|
| 60 |
+
"de_dep_news_trf"
|
| 61 |
+
],
|
| 62 |
+
"Greek":[
|
| 63 |
+
"",
|
| 64 |
+
"el_core_news_sm",
|
| 65 |
+
"el_core_news_md",
|
| 66 |
+
"el_core_news_lg"
|
| 67 |
+
],
|
| 68 |
+
"Italian":[
|
| 69 |
+
"",
|
| 70 |
+
"it_core_news_sm",
|
| 71 |
+
"it_core_news_md",
|
| 72 |
+
"it_core_news_lg"
|
| 73 |
+
],
|
| 74 |
+
"Japanese":[
|
| 75 |
+
"",
|
| 76 |
+
"ja_core_news_sm",
|
| 77 |
+
"ja_core_news_md",
|
| 78 |
+
"ja_core_news_lg",
|
| 79 |
+
"ja_core_news_trf"
|
| 80 |
+
],
|
| 81 |
+
"Korean":[
|
| 82 |
+
"",
|
| 83 |
+
"ko_core_news_sm",
|
| 84 |
+
"ko_core_news_md",
|
| 85 |
+
"ko_core_news_lg"
|
| 86 |
+
],
|
| 87 |
+
"Lithuanian":[
|
| 88 |
+
"",
|
| 89 |
+
"lt_core_news_sm",
|
| 90 |
+
"lt_core_news_md",
|
| 91 |
+
"lt_core_news_lg"
|
| 92 |
+
],
|
| 93 |
+
"Macedonian":[
|
| 94 |
+
"",
|
| 95 |
+
"mk_core_news_sm",
|
| 96 |
+
"mk_core_news_md",
|
| 97 |
+
"mk_core_news_lg"
|
| 98 |
+
],
|
| 99 |
+
"Multi-language":[
|
| 100 |
+
"",
|
| 101 |
+
"xx_ent_wiki_sm",
|
| 102 |
+
"xx_sent_ud_sm"
|
| 103 |
+
],
|
| 104 |
+
"Norwegian Bokmål":[
|
| 105 |
+
"",
|
| 106 |
+
"nb_core_news_sm",
|
| 107 |
+
"nb_core_news_md",
|
| 108 |
+
"nb_core_news_lg"
|
| 109 |
+
],
|
| 110 |
+
"Polish":[
|
| 111 |
+
"",
|
| 112 |
+
"pl_core_news_sm",
|
| 113 |
+
"pl_core_news_md",
|
| 114 |
+
"pl_core_news_lg"
|
| 115 |
+
],
|
| 116 |
+
"Portuguese":[
|
| 117 |
+
"",
|
| 118 |
+
"pt_core_news_sm",
|
| 119 |
+
"pt_core_news_md",
|
| 120 |
+
"pt_core_news_lg"
|
| 121 |
+
],
|
| 122 |
+
"Romanian":[
|
| 123 |
+
"",
|
| 124 |
+
"ro_core_news_sm",
|
| 125 |
+
"ro_core_news_md",
|
| 126 |
+
"ro_core_news_lg"
|
| 127 |
+
],
|
| 128 |
+
"Russian":[
|
| 129 |
+
"",
|
| 130 |
+
"ru_core_news_sm",
|
| 131 |
+
"ru_core_news_md",
|
| 132 |
+
"ru_core_news_lg"
|
| 133 |
+
],
|
| 134 |
+
"Spanish":[
|
| 135 |
+
"",
|
| 136 |
+
"es_core_news_sm",
|
| 137 |
+
"es_core_news_md",
|
| 138 |
+
"es_core_news_lg",
|
| 139 |
+
"es_dep_news_trf"
|
| 140 |
+
],
|
| 141 |
+
"Swedish":[
|
| 142 |
+
"",
|
| 143 |
+
"sv_core_news_sm",
|
| 144 |
+
"sv_core_news_md",
|
| 145 |
+
"sv_core_news_lg"
|
| 146 |
+
|
| 147 |
+
]
|
| 148 |
+
|
| 149 |
+
}
|
packages.txt
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
python-dev
|
| 2 |
+
libxml2-dev
|
| 3 |
+
libxslt1-dev
|
| 4 |
+
antiword
|
| 5 |
+
unrtf
|
| 6 |
+
poppler-utils
|
| 7 |
+
tesseract-ocr
|
| 8 |
+
flac
|
| 9 |
+
ffmpeg
|
| 10 |
+
lame
|
| 11 |
+
libmad0
|
| 12 |
+
libsox-fmt-mp3
|
| 13 |
+
sox
|
| 14 |
+
libjpeg-dev
|
| 15 |
+
swig
|
requirements.txt
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
streamlit==1.11.1
|
| 2 |
+
textract==1.6.5
|
| 3 |
+
spacy==3.4.1
|
| 4 |
+
#spacy-streamlit==1.0.4
|