add new pages + banner

Streamlit/Welcome.py

+"""
+Streamlit Application
+Loïc Chapron
+"""
+
 import streamlit as st
 import pandas as pd
 
@@ -6,6 +11,22 @@ st.set_page_config(
 )
 
 
+# Load Banner image
+st.image('img/gargantool_banner.jpg')
+
+# Reduce size on border
+st.markdown("""
+        <style>
+               .block-container {
+                    padding-top: 2rem;
+                    padding-bottom: 0rem;
+                    padding-left: 1rem;
+                    padding-right: 1rem;
+                }
+        </style>
+        """, unsafe_allow_html=True)
+
+# Load the language file
 def load_bundle(lang):
     df = pd.read_csv("lang/text_Welcome.csv")
     df = df.query(f"locale == '{lang}'")
@@ -15,10 +36,11 @@ def load_bundle(lang):
         tmp[df.key.to_list()[i]] = df.value.to_list()[i]
     return tmp
 
+# Load the language file
 def update_lang():
-    print(st.session_state.general_language)
     st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
 
+# Test if it's first connection on page or else if the last page was this one
 if 'general_session_page' not in st.session_state.keys():
     st.session_state.general_lang_dict = {'Français' : 'fr', 'English': 'en'}
     st.session_state.general_text_dict = load_bundle('fr')
@@ -27,12 +49,16 @@ if 'general_session_page' not in st.session_state.keys():
 elif st.session_state.general_session_page != 'Welcome':
     st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
     st.session_state.general_session_page = 'Welcome'
+    # Delete every key who aren't fron this file
+    for key in st.session_state.keys():
+        if 'general_' not in key:
+            del st.session_state[key]
 
-
+# select the lang
 st.selectbox('Langue', list(st.session_state.general_lang_dict.keys()), list(st.session_state.general_lang_dict.keys()).index(st.session_state.general_language),key='general_language', on_change=update_lang)
 
 
-st.write("# Welcome to GanganText Tools")
+st.write(st.session_state.general_text_dict['title'])
 
 st.write(st.session_state.general_text_dict['welcome'])
 st.write(st.session_state.general_text_dict['tools'])

Streamlit/lang/text_GarganTextJsonToTSV.csv

+locale,key,value
+fr,title,"# Json Vers TSV"
+en,title,"# Json To TSV"
+
+fr,text,"Transforme un corpus Json venant de Gargantext en TSV pour GarganText"
+en,text,"Transform a Json corpus fron GarganText to a TSV file for GarganText"
+
+fr,file,"Choisir un fichier"
+en,file,"Choose a file"
+
+fr,new_file,"Télécharge ton fichier TSV :"
+en,new_file,"Download your TSV file:"
+
+fr,error,"Erreur : le fichier n'est pas valide"
+en,error,"Error : the file isn't valid"
\ No newline at end of file

Streamlit/lang/text_HALToGarganText.csv

+locale,key,value
+fr,title,"**HAL vers GarganText**"
+en,title,"**HAL To GarganText**"
+
+fr,text,"HAL est une base de document scientifique en ligne et libre d'accès contenant plus d'un million de document."
+en,text,"HAL is an online and free access scientific document database containing more than a million documents"
+
+fr,keyword,"Mots clés"
+en,keyword,"Key word"
+
+fr,lang,"Langue des textes (si possible)"
+en,lang,"Text languages (if possible)"
+
+fr,submit,"Soumettre"
+en,submit,"Submit"
+
+fr,load_api,"Chargement de l'api..."
+en,load_api,"Loading API..."
+
+fr,overload_api,"L'API est surchargé, relancer la requête dans quelques secondes"
+en,overload'api,"The API is overloaded, please retry the request in a few seconds"
+
+fr,nb_doc,"Nombres de documents : "
+en,nb_doc,"Numbers of documents : "
+
+fr,perform1,"Pour des raisons de performence, on limit à "
+fr,perform2," le nombre de document maximum"
+en,perform1,"For performance reasons, we limit to "
+en,perform2," the maximum number of documents"
+
+fr,nb_taken,"Nombres de documents à prendre"
+en,nb_taken,"Number of documents to take"
+
+fr,createTSV,"Création du fichier TSV (Cela peut prendre quelque minutes)"
+en,createTSV,"Creation of the TSV file (It may take a while)"

Streamlit/lang/text_IsidoreToGarganText.csv

@@ -5,8 +5,8 @@ en,title,"**Isidore To GarganText**"
 fr,keyword,"Mots clés"
 en,keyword,"Key word"
 
-fr,lang,"Langue"
-en,lang,"Language"
+fr,lang,"Langue des textes (si possible)"
+en,lang,"Text languages (if possible)"
 
 fr,submit,"Soumettre"
 en,submit,"Submit"

Streamlit/lang/text_IstexToGarganText.csv

+locale,key,value
+fr,title,"# Istex Vers GarganText"
+en,title,"# Istex To GarganText"
+
+fr,file,"Choisir un fichier"
+en,file,"Choose a file"
+
+fr,dup1,"Certains fichiers ("
+fr,dup2,") ont été retirés pour divers raisons (fichier au mauvais format, fichier identique...)"
+en,dup1,"Some file ("
+en,dup2,") have been removed for various reasons (file with wrong format, file already present...)"
+
+fr,new_file,"Télécharge ton fichier TSV :"
+en,new_file,"Download your TSV file:"
+
+fr,error,"Erreur : le fichier n'est pas valide"
+en,error,"Error : the file isn't valid"
\ No newline at end of file

Streamlit/lang/text_Welcome.csv

 locale,key,value
+
+fr,title,"# Bienvenue sur GanganText Tools"
+en,title,"# Welcome to GanganText Tools"
+
 fr,welcome,"Bienvenue sur ces pages rassemblant des outils développés par des utilisateurs de GarganText pour des utilisateurs de GarganText."
 en,welcome,"Welcome to these pages featuring tools developed by GarganText’ users for GarganText’ users."
 

Streamlit/lang/text_ZoteroToGarganText.csv

@@ -5,6 +5,9 @@ en,title,"**Zotero vers GarganText**"
 fr,data,"Type de donnée"
 en,data,"Type of data"
 
+fr,help,"Trouvé votre ID d'utilisateur ici: https://www.zotero.org/settings/keys"
+en,help,"Find your user ID here: https://www.zotero.org/settings/keys"
+
 fr,submit,"Suivant"
 en,submit,"Submit"
 

Streamlit/pages/GarganText_Json_To_TSV.py

+"""
+Streamlit Application
+Loïc Chapron
+"""
+
+import streamlit as st
+import pandas as pd
+
+
+st.image('img/gargantool_banner.jpg')
+
+st.markdown("""
+        <style>
+               .block-container {
+                    padding-top: 2rem;
+                    padding-bottom: 0rem;
+                    padding-left: 1rem;
+                    padding-right: 1rem;
+                }
+        </style>
+        """, unsafe_allow_html=True)
+
+def load_bundle(lang):
+    df = pd.read_csv("lang/text_GarganTextJsonToTSV.csv")
+    df = df.query(f"locale == '{lang}'")
+
+    tmp = {}
+    for i in range(len(df)):
+        tmp[df.key.to_list()[i]] = df.value.to_list()[i]
+    return tmp
+
+def update_lang():
+    st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
+
+if 'general_session_page' not in st.session_state.keys():
+    st.session_state.general_lang_dict = {'Français' : 'fr', 'English': 'en'}
+    st.session_state.general_text_dict = load_bundle('fr')
+    st.session_state.general_language = 'Français'
+    st.session_state.general_session_page = 'GarganTextJsonToTSV'
+elif st.session_state.general_session_page != 'GarganTextJsonToTSV':
+	st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
+	st.session_state.general_session_page = 'GarganTextJsonToTSV'
+	for key in st.session_state.keys():
+		if 'general_' not in key:
+			del st.session_state[key]
+
+
+st.selectbox('Langue', list(st.session_state.general_lang_dict.keys()), list(st.session_state.general_lang_dict.keys()).index(st.session_state.general_language),key='general_language', on_change=update_lang)
+
+
+def getText(corpusJson):
+
+	output = "title\tsource\tpublication_year\tpublication_month\tpublication_day\tabstract\tauthors\tweight\n"
+
+
+	for row in corpusJson['corpus'] :
+		doc = row['document']['hyperdata']
+		abstract = "empty"
+		authors  = "empty"
+		title  = "empty"
+		source = "empty"
+
+		if 'title' in doc.keys() :
+			title = doc['title'].replace('"','').replace('\t','')
+
+		if 'source' in doc.keys() :
+			source = doc['source'].replace('"','').replace('\t','')
+
+		if 'abstract' in doc.keys() :
+			abstract = doc['abstract'].replace('"','').replace('\t','')
+
+
+		if 'authors' in doc.keys() :
+			authors = doc['authors']
+
+
+		output += title + "\t" + source + "\t" + str(doc['publication_year']) + "\t" + str(doc['publication_month']) + "\t" + str(doc['publication_day']) + "\t" + abstract + "\t" + authors + "\t" + str(1) + "\n"
+
+	return output
+
+st.write(st.session_state.general_text_dict['title'])
+st.write(st.session_state.general_text_dict['text'])
+file = st.file_uploader(st.session_state.general_text_dict['file'],type=["json"],key='file')
+
+if file:
+	try:
+		name = file.name.split('.')[0] + '.csv'
+		df = pd.read_json(file) 
+		st.write(st.session_state.general_text_dict['new_file'])
+		
+		st.download_button(name, getText(df), name)
+	except Exception as e:
+		st.write(st.session_state.general_text_dict['error'])
+		print(e)
+	file.close()
+

Streamlit/pages/HAL_To_GarganText.py

+"""
+Streamlit Application
+Loïc Chapron
+"""
+import requests as req
+import json
+import streamlit as st
+from datetime import datetime
+import pandas as pd
+
+limit = 500
+limitItems = 10000
+
+st.image('img/gargantool_banner.jpg')
+
+st.markdown("""
+        <style>
+               .block-container {
+                    padding-top: 2rem;
+                    padding-bottom: 0rem;
+                    padding-left: 1rem;
+                    padding-right: 1rem;
+                }
+        </style>
+        """, unsafe_allow_html=True)
+
+def load_bundle(lang):
+    df = pd.read_csv("lang/text_HALToGarganText.csv")
+    df = df.query(f"locale == '{lang}'")
+
+    tmp = {}
+    for i in range(len(df)):
+        tmp[df.key.to_list()[i]] = df.value.to_list()[i]
+    return tmp
+
+def update_lang():
+    st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
+
+if 'general_session_page' not in st.session_state.keys():
+    st.session_state.general_lang_dict = {'Français' : 'fr', 'English': 'en'}
+    st.session_state.general_text_dict = load_bundle('fr')
+    st.session_state.general_language = 'Français'
+    st.session_state.general_session_page = 'HALToGarganText'
+elif st.session_state.general_session_page != 'HALToGarganText':
+    st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
+    st.session_state.general_session_page = 'HALToGarganText'
+    for key in st.session_state.keys():
+        if 'general_' not in key:
+            del st.session_state[key]
+
+
+st.selectbox('Langue', list(st.session_state.general_lang_dict.keys()), list(st.session_state.general_lang_dict.keys()).index(st.session_state.general_language),key='general_language', on_change=update_lang)
+
+
+def loadApiHALNbFile(search, lang):
+    url = 'http://api.archives-ouvertes.fr/search/?q=' + search + '&rows=5&fl=title_s,' + lang + '_title_s,source_s,publicationDate_s,authFullName_s,' + lang + '_abstract_s,abstract_s&fq=language_s:' + lang
+    resp = req.get(url)
+    print(url)
+    try:
+        docs = json.loads(resp.content)
+    except Exception as e:
+        docs = 0
+
+    return docs['response']['numFound']
+
+
+def loadApiHAL(search, lang, page, nbvalue):
+    url = 'http://api.archives-ouvertes.fr/search/?q=' + search + '&start=' + str(page * limit) + '&rows=' + str(nbvalue) + '&fl=title_s,' + lang + '_title_s,source_s,publicationDate_s,authFullName_s,' + lang + '_abstract_s,abstract_s&fq=language_s:' + lang
+    resp = req.get(url)
+    print(url)
+    try:
+        docs = json.loads(resp.content)
+    except Exception as e:
+        docs = 0
+
+    return docs
+
+
+def getParamFromDoc(docs):
+    output = ''
+    for doc in docs:
+        if 'en_title_s' in doc.keys():
+            title = doc['en_title_s'][0]
+        elif 'fr_title_s' in doc.keys():
+            title = doc['fr_title_s'][0]
+        elif 'title_s' in doc.keys():
+            title = doc['title_s'][0]
+        else:
+            continue
+
+            # Authors
+
+        if 'authFullName_s' in doc.keys():
+            authors = []
+            for author in doc['authFullName_s']:
+                authors.append(author)
+            authors = ';'.join(authors)
+        else:
+            authors = ''
+
+        # Source
+            
+        if 'source_s' in doc.keys():
+            source = doc['source_s']
+        else:
+            source = ''
+
+        # Abstract
+        if 'en_abstract_s' in doc.keys():
+            abstract = doc['en_abstract_s'][0]
+        elif 'fr_abstract_s' in doc.keys():
+            abstract = doc['fr_abstract_s'][0]
+        elif 'abstract_s' in doc.keys():
+            abstract = doc['abstract_s'][0]
+        else:
+            abstract = ''
+
+        # Date
+
+        if 'publicationDate_s' in doc.keys():
+            split = doc['publicationDate_s'].split('-')
+            if len(split) == 3:
+                pdate = datetime.strptime(doc['publicationDate_s'], '%Y-%m-%d').strftime('%Y\t%m\t%d')
+            elif len(split) == 2:
+                pdate = datetime.strptime(doc['publicationDate_s'], '%Y-%m').strftime('%Y\t%m\t1')
+            else:
+                pdate = doc['publicationDate_s'] + '\t1\t1'
+        else:
+            pdate = '1900\t1\t1'
+
+        abstract = abstract.encode(encoding = 'UTF-8', errors = 'ignore').decode("utf-8").replace('\t', '').replace('"', '').replace('\n', '')
+        title = title.encode(encoding = 'UTF-8', errors = 'ignore').decode("utf-8").replace('\t', '').replace('"', '').replace('\n', '')
+        source = source.encode(encoding = 'UTF-8', errors = 'ignore').decode("utf-8").replace('\t', '').replace('"', '').replace('\n', '').replace('\n', '')
+        
+        # Output
+        output += str(title) + "\t" + source + "\t" + str(pdate) + "\t" + abstract + "\t" + authors + "\t" + str(1) + "\n"
+    return output
+
+def create_output(search, lang, nb_value):
+    output = "title\tsource\tpublication_year\tpublication_month\tpublication_day\tabstract\tauthors\tweight\n"
+    for i in range(0, nb_value//limit):
+        response = loadApiHAL(search, lang, i, limit)
+        output += getParamFromDoc(response['response']['docs'])
+    if (nb_value%limit != 0):
+        response = loadApiHAL(search, lang, nb_value//limit, nb_value%limit)
+        output += getParamFromDoc(response['response']['docs'])
+    return output
+
+lang = {
+    'Français' : 'fr',
+    'Anglais' : 'en',
+}
+
+if 'stage_isidore' not in st.session_state:
+    st.session_state.stage_isidore = 0
+    st.session_state.nb_wanted = 1
+
+def set_stage(stage):
+    st.session_state.stage_isidore = stage
+    st.session_state.output = ''
+
+
+# txt
+st.write(st.session_state.general_text_dict['text'])
+
+
+# Form
+form = st.form('api')
+form.write(st.session_state.general_text_dict['title'])
+
+search = form.text_input(st.session_state.general_text_dict['keyword'])
+language = form.selectbox(st.session_state.general_text_dict['lang'], lang.keys())
+
+form.form_submit_button(st.session_state.general_text_dict['submit'], on_click=set_stage, args=(1,))
+
+# API and Slider
+if st.session_state.stage_isidore > 0:
+
+    # Only call first time and after 
+    if 'search' not in st.session_state or 'language' not in st.session_state or search != st.session_state.search or language != st.session_state.language:
+        with st.spinner(st.session_state.general_text_dict['load_api'] ):
+            nb_doc = int(loadApiHALNbFile(search, lang[language]))
+        st.session_state.nb_doc = nb_doc
+        if nb_doc != 0:
+            st.session_state.search = search
+            st.session_state.language = language
+
+    if st.session_state.nb_doc != 0:
+
+        # Form with slider
+        form2 = st.form('my_form2')
+        form2.write(st.session_state.general_text_dict['nb_doc'] + str(st.session_state.nb_doc))
+
+        if st.session_state.nb_doc > limitItems:
+            form2.write(st.session_state.general_text_dict['perform1'] + str(limitItems) + st.session_state.general_text_dict['perform2'])
+            st.session_state.nb_wanted = form2.slider(st.session_state.general_text_dict['nb_taken'], 1, limitItems)
+        else:
+            st.session_state.nb_wanted = form2.slider(st.session_state.general_text_dict['nb_taken'], 1, int(st.session_state.nb_doc))
+        form2.form_submit_button(st.session_state.general_text_dict['submit'], on_click=set_stage, args=(2,))
+
+    else:
+        st.write(st.session_state.general_text_dict['overload_api'] )
+
+    
+# Download
+if st.session_state.stage_isidore > 1:
+
+    with st.spinner(st.session_state.general_text_dict['createTSV'] ):
+        if st.session_state.output == '':
+            print(st.session_state.nb_wanted)
+            st.session_state.output = create_output(st.session_state.search, lang[st.session_state.language], st.session_state.nb_wanted)
+
+    st.download_button('Download TSV', st.session_state.output, 'output.csv')
+
+

Streamlit/pages/Isidore_To_GarganText.py

@@ -3,20 +3,25 @@ Streamlit Application
 Loïc Chapron
 """
 
-#streamlit run Isidore_To_GarganText.py
-
 import streamlit as st
 import requests as req
 import json
 import pandas as pd
 import time
 
-keys = ['search','language','stage_isidore','output','nb_doc','nb_wanted', 'nb_bad_file']
 
-for key in st.session_state.keys():
-    if key not in keys and 'general_' not in key:
-        del st.session_state[key]
+st.image('img/gargantool_banner.jpg')
 
+st.markdown("""
+        <style>
+               .block-container {
+                    padding-top: 2rem;
+                    padding-bottom: 0rem;
+                    padding-left: 1rem;
+                    padding-right: 1rem;
+                }
+        </style>
+        """, unsafe_allow_html=True)
 
 def load_bundle(lang):
     df = pd.read_csv("lang/text_IsidoreToGarganText.csv")
@@ -38,6 +43,9 @@ if 'general_session_page' not in st.session_state.keys():
 elif st.session_state.general_session_page != 'IsidoreToGarganText':
     st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
     st.session_state.general_session_page = 'IsidoreToGarganText'
+    for key in st.session_state.keys():
+        if 'general_' not in key:
+            del st.session_state[key]
 
 
 st.selectbox('Langue', list(st.session_state.general_lang_dict.keys()), list(st.session_state.general_lang_dict.keys()).index(st.session_state.general_language),key='general_language', on_change=update_lang)
@@ -84,7 +92,9 @@ def create_output(search, language, nb_doc):
                 break
             time.sleep(retryTime)
             print('Retry')
-        output += createFile(txt, numberReplies, language)
+        tmp, nb_tmp = createFile(txt, nb_doc%numberReplies, language)
+        output += tmp
+        nb += nb_tmp
     if nb_doc%numberReplies != 0:
         txt = loadApiIsidorePage(search, language, nb_doc//numberReplies + 1)
         tmp, nb_tmp = createFile(txt, nb_doc%numberReplies, language)

Streamlit/pages/Istex_To_GarganText.py

+"""
+Streamlit Application
+Loïc Chapron
+"""
+
+import json
+import pandas as pd
+import datetime
+import zipfile
+import streamlit as st
+
+
+st.image('img/gargantool_banner.jpg')
+
+st.markdown("""
+        <style>
+               .block-container {
+                    padding-top: 2rem;
+                    padding-bottom: 0rem;
+                    padding-left: 1rem;
+                    padding-right: 1rem;
+                }
+        </style>
+        """, unsafe_allow_html=True)
+
+def load_bundle(lang):
+    df = pd.read_csv("lang/text_IstexToGarganText.csv")
+    df = df.query(f"locale == '{lang}'")
+
+    tmp = {}
+    for i in range(len(df)):
+        tmp[df.key.to_list()[i]] = df.value.to_list()[i]
+    return tmp
+
+def update_lang():
+    st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
+
+if 'general_session_page' not in st.session_state.keys():
+    st.session_state.general_lang_dict = {'Français' : 'fr', 'English': 'en'}
+    st.session_state.general_text_dict = load_bundle('fr')
+    st.session_state.general_language = 'Français'
+    st.session_state.general_session_page = 'IstexToGarganText'
+elif st.session_state.general_session_page != 'IstexToGarganText':
+    st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
+    st.session_state.general_session_page = 'IstexToGarganText'
+    for key in st.session_state.keys():
+        if 'general_' not in key:
+            del st.session_state[key]
+
+
+st.selectbox('Langue', list(st.session_state.general_lang_dict.keys()), list(st.session_state.general_lang_dict.keys()).index(st.session_state.general_language),key='general_language', on_change=update_lang)
+
+
+def read_zip(zip_file):
+    output=[]
+    dup = 0
+    with zipfile.ZipFile(zip_file, 'r') as zip_ref:
+        for file in zip_ref.namelist():
+            
+            if file.split('.')[1] != 'json' or file.split('.')[0] == 'manifest':
+                continue
+            try:
+                
+                with zip_ref.open(file) as f:
+                    data = json.load(f)
+                    article=pd.json_normalize(data)
+                    f.close()
+                
+                temp={}
+                temp["title"]=article.get("title", '')[0].encode(encoding = 'UTF-8', errors = 'ignore').decode("utf-8").replace("\t", " ")
+                
+                try:
+                    temp["abstract"]=article.get("abstract","")[0].encode(encoding = 'UTF-8', errors = 'ignore').decode("utf-8").replace("\t", " ")
+                except Exception as e:
+                    temp["abstract"] = '' 
+                
+
+                try:
+                    authors=""
+                    for author in article["author"][0]:
+                        authors+=author["name"]+"; "
+                    authors=authors[:-2]
+                except:
+                    author = ''
+
+
+                temp["code"] = article.get("_id")[0]
+                temp["authors"]=authors.encode(encoding = 'UTF-8', errors = 'ignore').decode("utf-8").replace("\t", " ")
+                try:
+                    temp["source"]=article.get('host.title')[0].encode(encoding = 'UTF-8', errors = 'ignore').decode("utf-8").replace("\t", " ")
+                except:
+                    temp["source"] = ''
+
+                try:
+                    temp["publication_year"]=article["publicationDate"][0]
+                except:
+                    temp["publication_year"]=datetime.date.today().year
+                temp["publication_year"]=article.get("publicationDate", datetime.date.today().year)[0]
+                temp["publication_month"]=1
+                temp["publication_day"]=1
+                
+
+                output.append(temp)
+                
+            except Exception as e:
+                dup += 1
+            
+        zip_ref.close()
+
+    output=pd.DataFrame(output)
+    duplicated = output['title'].str.lower().replace(",", "", regex=True).duplicated()
+
+    if (duplicated.any()):
+        dup += duplicated.sum()
+    output.drop(['code'], axis=1)
+    output = output[~duplicated]
+    df = pd.DataFrame(output)
+    return df.to_csv(index=False, sep='\t'), dup
+
+st.write(st.session_state.general_text_dict['title'])
+file = st.file_uploader(st.session_state.general_text_dict['file'],type=["zip"],key='file')
+
+if file:
+    try:
+        name = file.name.split('.')[0] + '.csv'
+        res, nb_dup = read_zip(file)
+        if nb_dup:
+            st.write(st.session_state.general_text_dict['dup1'] + str(nb_dup) + st.session_state.general_text_dict['dup2'])
+        st.write(st.session_state.general_text_dict['new_file'])
+        st.download_button(name, res, name)
+    except Exception as e:
+        st.write(st.session_state.general_text_dict['error'])
+        print(e)
+    file.close()
\ No newline at end of file

Streamlit/pages/Merge_Term_GarganText.py

-import json
+"""
+Streamlit Application
+Loïc Chapron
+"""
+
 import streamlit as st
 from datetime import datetime
 import pandas as pd
 
-for key in st.session_state.keys():
-    if 'general_' not in key:
-        del st.session_state[key]
 
+st.image('img/gargantool_banner.jpg')
+
+st.markdown("""
+        <style>
+               .block-container {
+                    padding-top: 2rem;
+                    padding-bottom: 0rem;
+                    padding-left: 1rem;
+                    padding-right: 1rem;
+                }
+        </style>
+        """, unsafe_allow_html=True)
 
 def load_bundle(lang):
     df = pd.read_csv("lang/text_MergeTermGarganText.csv")
@@ -28,6 +41,9 @@ if 'general_session_page' not in st.session_state.keys():
 elif st.session_state.general_session_page != 'MergeTermGarganText':
     st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
     st.session_state.general_session_page = 'MergeTermGarganText'
+    for key in st.session_state.keys():
+        if 'general_' not in key:
+            del st.session_state[key]
 
 
 st.selectbox('Langue', list(st.session_state.general_lang_dict.keys()), list(st.session_state.general_lang_dict.keys()).index(st.session_state.general_language),key='general_language', on_change=update_lang)
@@ -36,11 +52,8 @@ st.selectbox('Langue', list(st.session_state.general_lang_dict.keys()), list(st.
 
 def tmp(file1, file2):
 
-	listJson1 = json.load(file1)
-	listJson2 = json.load(file2)
-
-	ngrams1 = listJson1['NgramsTerms']['data']
-	ngrams2 = listJson2['NgramsTerms']['data']
+	ngrams1 = file1['NgramsTerms']['data']
+	ngrams2 = file2['NgramsTerms']['data']
 	merged  = {}
 
 	roots = []
@@ -85,22 +98,36 @@ def tmp(file1, file2):
 			merged[root] = ngrams1[root]
 
 
-	listJson1['NgramsTerms']['data'] = merged
-	listJson1['Authors']['data']   = {}
-	listJson1['Institutes']['data'] = {}
-	listJson1['Sources']['data']   = {}
-	
-	return json.dumps(listJson1, sort_keys=False, indent=4)
+	file1['NgramsTerms']['data'] = merged
+	file1['Authors']['data']   = {}
+	file1['Institutes']['data'] = {}
+	file1['Sources']['data']   = {}
+
+	tmp = file1.to_json(orient='columns',indent=4)
+	return tmp
+
+
+
 
 
 st.subheader(st.session_state.general_text_dict['title'])
 col1, col2 = st.columns(2)
 with col1:
-    file1 = st.file_uploader(st.session_state.general_text_dict['file'],'json',key='file1')
+    file1 = st.file_uploader(st.session_state.general_text_dict['file'],type=["json"],key='file1')
 with col2:
-    file2 = st.file_uploader(st.session_state.general_text_dict['file'],'json',key='file2')
+    file2 = st.file_uploader(st.session_state.general_text_dict['file'],type=["json"],key='file2')
+
+
 if (file1 and file2):
-	st.write(st.session_state.general_text_dict['new_file'])
-	time = datetime.strftime(datetime.now(),"%d-%m-%Y/%H:%M:%S")
-	name = 'output-' +time+ '.json'
-	st.download_button('Download File', tmp(file1, file2), name)
\ No newline at end of file
+	try:
+		df1 = pd.read_json(file1) 
+		df2 = pd.read_json(file2)
+		st.write(st.session_state.general_text_dict['new_file'])
+		time = datetime.strftime(datetime.now(),"%d-%m-%Y/%H:%M:%S")
+		name = 'output-' +time+ '.json'
+		st.download_button('Download File', tmp(df1, df2), name)
+	except Exception as e:
+		st.write("Error : one of the file isn't valid")
+	
+	file1.close()
+	file2.close()
\ No newline at end of file

Streamlit/pages/Zotero_To_GarganText.py

@@ -3,8 +3,6 @@ Streamlit Application
 Loïc Chapron
 """
 
-#streamlit run zotero2ggt.py
-
 import streamlit as st
 import requests as req
 import json
@@ -16,13 +14,19 @@ limit = 50 # This value cannot exceed 100 !
 
 st.set_page_config(page_title="Zotero to GarganText")
 
-keys = ['id','page','docs','docsByKey','result','stage', 'collectionsKey', 'format', 'zotero_search', 'nbdoc', 'select']
+st.image('img/gargantool_banner.jpg')
 
-for key in st.session_state.keys():
-    if key not in keys and 'general_' not in key:
-        del st.session_state[key]
+st.markdown("""
+        <style>
+               .block-container {
+                    padding-top: 2rem;
+                    padding-bottom: 0rem;
+                    padding-left: 1rem;
+                    padding-right: 1rem;
+                }
+        </style>
+        """, unsafe_allow_html=True)
 
-@st.cache_data
 def load_bundle(lang):
     df = pd.read_csv("lang/text_ZoteroToGarganText.csv")
     df = df.query(f"locale == '{lang}'")
@@ -43,6 +47,9 @@ if 'general_session_page' not in st.session_state.keys():
 elif st.session_state.general_session_page != 'ZoteroToGarganText':
     st.session_state.general_text_dict = load_bundle(st.session_state.general_lang_dict[st.session_state.general_language])
     st.session_state.general_session_page = 'ZoteroToGarganText'
+    for key in st.session_state.keys():
+        if 'general_' not in key:
+            del st.session_state[key]
 
 st.selectbox('Langue', list(st.session_state.general_lang_dict.keys()), list(st.session_state.general_lang_dict.keys()).index(st.session_state.general_language),key='general_language', on_change=update_lang)
 
@@ -248,7 +255,7 @@ if st.session_state.stage == 0:
     form.write(st.session_state.general_text_dict['title'])
 
     lst = ['items', 'collections']
-    st.session_state.id = form.text_input('ID',st.session_state.id, key='idForm', help="Trouvé votre ID d'utilisateur ici: https://www.zotero.org/settings/keys")
+    st.session_state.id = form.text_input('ID',st.session_state.id, key='idForm', help=st.session_state.general_text_dict['help'])
     st.session_state.format = form.selectbox(st.session_state.general_text_dict['data'], lst,lst.index(st.session_state.format), key='formatForm')
 
     form.form_submit_button(st.session_state.general_text_dict['submit'], on_click=set_stage, args=(1,))