From e6d8d72e52f16033990c327c7b005627199bf01c Mon Sep 17 00:00:00 2001
From: Patrick Jentsch
Date: Wed, 14 Jun 2023 14:50:04 +0200
Subject: [PATCH] Add visualization data method to cqi over socketio
---
.../cqi_over_socketio/cqi_corpora_corpus.py | 77 +++++++++++++++++--
app/static/js/CorpusAnalysis/CQiClient.js | 14 ++++
.../js/CorpusAnalysis/CorpusAnalysisApp.js | 9 +++
3 files changed, 95 insertions(+), 5 deletions(-)
diff --git a/app/corpora/cqi_over_socketio/cqi_corpora_corpus.py b/app/corpora/cqi_over_socketio/cqi_corpora_corpus.py
index 9a976dd7..79b1a800 100644
--- a/app/corpora/cqi_over_socketio/cqi_corpora_corpus.py
+++ b/app/corpora/cqi_over_socketio/cqi_corpora_corpus.py
@@ -1,6 +1,8 @@
+from collections import Counter
from flask import session
import cqi
import math
+import random
from app import db, socketio
from app.decorators import socketio_login_required
from app.models import Corpus
@@ -38,10 +40,75 @@ def cqi_corpora_corpus_query(cqi_client: cqi.CQiClient, corpus_name: str, subcor
@cqi_over_socketio
def cqi_corpora_corpus_update_db(cqi_client: cqi.CQiClient, corpus_name: str):
corpus = Corpus.query.get(session['d']['corpus_id'])
- corpus.num_tokens = cqi_client.corpora.get(corpus_name).attrs['size']
+ cqi_corpus = cqi_client.corpora.get(corpus_name)
+ corpus.num_tokens = cqi_corpus.size
db.session.commit()
+@socketio.on('cqi.corpora.corpus.get_visualization_data', namespace=ns)
+@socketio_login_required
+@cqi_over_socketio
+def cqi_corpora_corpus_get_visualization_data(cqi_client: cqi.CQiClient, corpus_name: str):
+ cqi_corpus = cqi_client.corpora.get(corpus_name)
+ payload = {}
+ payload['num_tokens'] = cqi_corpus.size
+ cqi_word_attr = cqi_corpus.positional_attributes.get('word')
+ payload['num_unique_words'] = cqi_word_attr.lexicon_size
+ payload['word_freqs'] = dict(zip(cqi_word_attr.values_by_ids(list(range(0, cqi_word_attr.lexicon_size))), cqi_word_attr.freqs_by_ids(list(range(0, cqi_word_attr.lexicon_size)))))
+ # payload['word_freqs'].sort(key=lambda a: a[1], reverse=True)
+ # payload['word_freqs'] = {k: v for k, v in payload['word_freqs']}
+ cqi_lemma_attr = cqi_corpus.positional_attributes.get('lemma')
+ payload['num_unique_lemmas'] = cqi_lemma_attr.lexicon_size
+ payload['lemma_freqs'] = dict(zip(cqi_lemma_attr.values_by_ids(list(range(0, cqi_lemma_attr.lexicon_size))), cqi_lemma_attr.freqs_by_ids(list(range(0, cqi_lemma_attr.lexicon_size)))))
+ # payload['lemma_freqs'].sort(key=lambda a: a[1], reverse=True)
+ # payload['lemma_freqs'] = {k: v for k, v in payload['lemma_freqs']}
+ cqi_s_attr = cqi_corpus.structural_attributes.get('s')
+ payload['num_sentences'] = cqi_s_attr.size
+ # assuming all tokens are in a sentence
+ payload['average_sentence_length'] = payload['num_tokens'] / payload['num_sentences'] if payload['num_sentences'] != 0 else 0
+ # payload['average_sentence_length'] = 0
+ # for s_id in range(0, cqi_s_attr.size):
+ # s_lbound, s_rbound = cqi_s_attr.cpos_by_id(s_id)
+ # payload['average_sentence_length'] += s_rbound - s_lbound + 1
+ # payload['average_sentence_length'] /= payload['num_sentences']
+ cqi_ent_type_attr = cqi_corpus.structural_attributes.get('ent_type')
+ payload['num_ent_types'] = cqi_ent_type_attr.size
+ payload['ent_type_freqs'] = dict(Counter(cqi_ent_type_attr.values_by_ids(list(range(0, cqi_ent_type_attr.size)))))
+ payload['num_unique_ent_types'] = len(payload['ent_type_freqs'])
+ payload['texts'] = []
+ cqi_text_attr = cqi_corpus.structural_attributes.get('text')
+ for text_id in range(0, cqi_text_attr.size):
+ text_lbound, text_rbound = cqi_text_attr.cpos_by_id(text_id)
+ text_cpos_list = list(range(text_lbound, text_rbound + 1))
+ text_payload = {}
+ text_payload['num_tokens'] = text_rbound - text_lbound + 1
+ text_word_ids = cqi_word_attr.ids_by_cpos(text_cpos_list)
+ print(text_word_ids)
+ text_payload['num_unique_words'] = len(set(text_word_ids))
+ text_payload['word_freqs'] = dict(Counter(cqi_word_attr.values_by_ids(text_word_ids)))
+ text_lemma_ids = cqi_lemma_attr.ids_by_cpos(text_cpos_list)
+ text_payload['num_unique_lemmas'] = len(set(text_lemma_ids))
+ text_payload['lemma_freqs'] = dict(Counter(cqi_word_attr.values_by_ids(text_lemma_ids)))
+ text_s_attr_ids = list(filter(lambda x: x != -1, cqi_s_attr.ids_by_cpos(text_cpos_list)))
+ text_payload['num_sentences'] = len(set(text_s_attr_ids))
+ # assuming all tokens are in a sentence
+ text_payload['average_sentence_length'] = text_payload['num_tokens'] / text_payload['num_sentences'] if text_payload['num_sentences'] != 0 else 0
+ # text_payload['average_sentence_length'] = 0
+ # for text_s_id in range(0, cqi_s_attr.size):
+ # text_s_lbound, text_s_rbound = cqi_s_attr.cpos_by_id(text_s_id)
+ # text_payload['average_sentence_length'] += text_s_rbound - text_s_lbound + 1
+ # text_payload['average_sentence_length'] /= text_payload['num_sentences']
+ text_ent_type_ids = list(filter(lambda x: x != -1, cqi_ent_type_attr.ids_by_cpos(text_cpos_list)))
+ text_payload['num_ent_types'] = len(set(text_ent_type_ids))
+ text_payload['ent_type_freqs'] = dict(Counter(cqi_ent_type_attr.values_by_ids(text_ent_type_ids)))
+ text_payload['num_unique_ent_types'] = len(text_payload['ent_type_freqs'])
+ for text_sub_attr in cqi_corpus.structural_attributes.list(filters={'part_of': cqi_text_attr}):
+ text_payload[text_sub_attr.name[(len(cqi_text_attr.name) + 1):]] = text_sub_attr.values_by_ids([text_id])[0]
+ payload['texts'].append(text_payload)
+ # print(payload)
+ return {'code': 200, 'msg': 'OK', 'payload': payload}
+
+
@socketio.on('cqi.corpora.corpus.paginate', namespace=ns)
@socketio_login_required
@cqi_over_socketio
@@ -52,13 +119,13 @@ def cqi_corpora_corpus_paginate(cqi_client: cqi.CQiClient, corpus_name: str, pag
per_page < 1
or page < 1
or (
- cqi_corpus.attrs['size'] > 0
- and page > math.ceil(cqi_corpus.attrs['size'] / per_page)
+ cqi_corpus.size > 0
+ and page > math.ceil(cqi_corpus.size / per_page)
)
):
return {'code': 416, 'msg': 'Range Not Satisfiable'}
first_cpos = (page - 1) * per_page
- last_cpos = min(cqi_corpus.attrs['size'], first_cpos + per_page)
+ last_cpos = min(cqi_corpus.size, first_cpos + per_page)
cpos_list = [*range(first_cpos, last_cpos)]
lookups = lookups_by_cpos(cqi_corpus, cpos_list)
payload = {}
@@ -67,7 +134,7 @@ def cqi_corpora_corpus_paginate(cqi_client: cqi.CQiClient, corpus_name: str, pag
# the lookups for the items
payload['lookups'] = lookups
# the total number of items matching the query
- payload['total'] = cqi_corpus.attrs['size']
+ payload['total'] = cqi_corpus.size
# the number of items to be displayed on a page.
payload['per_page'] = per_page
# The total number of pages
diff --git a/app/static/js/CorpusAnalysis/CQiClient.js b/app/static/js/CorpusAnalysis/CQiClient.js
index fcc0c87d..5baa59e8 100644
--- a/app/static/js/CorpusAnalysis/CQiClient.js
+++ b/app/static/js/CorpusAnalysis/CQiClient.js
@@ -98,6 +98,20 @@ class CQiCorpus {
this.subcorpora = new CQiSubcorpusCollection(this.socket, this);
}
+ getVisualizationData() {
+ return new Promise((resolve, reject) => {
+ const args = {corpus_name: this.name};
+
+ this.socket.emit('cqi.corpora.corpus.get_visualization_data', args, response => {
+ if (response.code === 200) {
+ resolve(response.payload);
+ } else {
+ reject(response);
+ }
+ });
+ });
+ }
+
getCorpusData() {
return new Promise((resolve, reject) => {
const dummyData = {
diff --git a/app/static/js/CorpusAnalysis/CorpusAnalysisApp.js b/app/static/js/CorpusAnalysis/CorpusAnalysisApp.js
index cb012730..c955117b 100644
--- a/app/static/js/CorpusAnalysis/CorpusAnalysisApp.js
+++ b/app/static/js/CorpusAnalysis/CorpusAnalysisApp.js
@@ -34,6 +34,15 @@ class CorpusAnalysisApp {
.then(
cQiCorpus => {
this.data.corpus = {o: cQiCorpus};
+ // this.data.corpus.o.getVisualizationData()
+ // .then(
+ // (visualizationData) => {
+ // console.log(visualizationData);
+ // this.renderGeneralCorpusInfo(visualizationData);
+ // this.renderTextInfoList(visualizationData);
+ // this.renderTextProportionsGraphic(visualizationData);
+ // }
+ // );
this.data.corpus.o.getCorpusData()
.then(corpusData => {
this.renderGeneralCorpusInfo(corpusData);